Wikipedia:Requests for adminship/ProtectionBot

From Wikipedia, the free encyclopedia

The following discussion is preserved as an archive of a request for adminship that was withdrawn. Please do not modify it.


I am withdrawing this RFA.
Werdna (with the help of a few others) has written a patch for Mediawiki that implements "cascading protection" as an option on the protection page. Once this cascading protection is enabled for the Main Page, it will ensure that any time a template or image is added to the Main Page that the item is automatically protected and that this protection automatically ends once it is removed from the Main Page. Enabling such a scheme in Mediawiki is of course better than any bot could do at the same task. This patch has now been applied to the main development trunk of Mediawiki and should be available on the site "in a few days" following a database update to support it.
While ProtectionBot (may he rest in peace) has certain functionality that exceeds that provided by Werdna's patch, I feel the technical situation has changed too greatly for this RFA to continue to be valid. Consequently, I am choosing to withdraw it at this time.
I would like to thank everyone who has supported this endeavor, especially those who strove to provide valuable feedback into the bot's functionality and smooth the process along. At the same time, I would also like express my clear frustration with this process as a whole. Werdna and a few developers managed to implement a solution to an important problem in a matter of days. I implemented a solution to that same problem and then expended many many times the effort trying to justify and defend that solution through an overly complex process that sometimes seemed to have been invented on the spot. To date, the various ProtectionBot discussions have generated more than 500 kilobytes of text. Or put another way, that's roughly 140 pages at a standard 12 pt font. A good idea, that can be shown to work, should not require this much effort.
Having dumped an unreasonable amount of my time and energy into this during the last two weeks (to the real detriment of other aspects of my life), I am going to take a wikibreak for a while. Hopefully while I'm gone, people will reflect on this experience and try to find a better way.
Lastly, I would encourage people to go thank Werdna for his valuable contribution to the security of Wikipedia.
Thank you and good night.
-- Dragons flight 01:45, 11 January 2007 (UTC)


Contents

[edit] ProtectionBot

(185/41/13); Withdrawn 01:45 January 11 2006

This is a request for a fully automated adminbot.

ProtectionBot is a bot designed and intended to by operated by Dragons flight (talk • contribs • blocks • protects • deletions • moves • rights). It is intended to combat the growth in high profile vandalism on the Main Page and Today's Featured Article occurring through unprotected templates and images, by automating the process of protecting these templates and images. Based on a review of Talk:Main Page, it appears that since December 15th there have been at least six separate occasions during which vulgar imagery was inserted into the Main Page by vandals exploiting unprotected templates or images. Undoubtedly there have been even more attacks on Today's Featured Article by vandals using similar methods, and Category:Protected pages associated with Main Page articles has seen steady business (mostly templates) as a way to ward off these attacks.

ProtectionBot is designed to combat these attacks by automatically protecting all templates and images appearing on the Main Page and Today's Featured Article, and then reverting this protection once they are no longer present in these high profile locations. In addition, it will protect the predictable elements (such as the next Picture of the Day) a day before they appear on the main page. While a vandal might attack a template or image more than a day in advance, this 24 hour window will give users a chance to verify that the protected content is clean (e.g. by looking at Main Page/Tomorrow) well before it is featured on the Main Page.

As is the current custom, to protect images hosted on Commons, ProtectionBot has also been given the power to upload local copies of those images and their description pages onto this Wikipedia before protecting them. They would be tagged with a template similar to {{c-uploaded}}, and when the local copy was no longer needed, ProtectionBot would mark them for speedy deletion with a tag that also reminds the deleting admin to restore any previous Wikipedia specific image description page content, as is customary. ProtectionBot would not be directly deleting anything.

The task of protecting the main page is an essential, but repetitious, one that can and should be automated to catch lapses before they become large scale embarrassments. Especially with content that is automatically rotated on a daily basis, admins have better things to do than remembering to protect and unprotect a dozen templates and images each day.

The bot has already been tested live on Wikipedia in an approved trial using my admin account, and though there were a few initial bugs, I believe they are resolved now and the bot could begin performing the described functions immediately, subject to approval. Dragons flight 09:15, 7 January 2007 (UTC)

For additional technical details see


[edit] Questions for the bot operator

1. I've already stated my strong support for this bot, and I believe that this is already understood, albeit implicitly, by yourself (Dragons flight); however, I would like it to be stated explicitly here for the record (and for those opposers fearing an overpoweful admin-bot): if you ever wanted to add additional functionalities involving administrative rights (i.e. blocking, etc.) to your bot, would you take it to RfA for approval again, and would you support immediate de-sysopping and blocking of the account if it ever performs any task not stated explicitly in this RfA nomination or not explicitly approved in another RfA? Thanks! Flcelloguy (A note?) 16:23, 7 January 2007 (UTC)
  • Just to be clear, I meant additional functionalities not related to the protection of the Main Page; i.e. if another template is added to the Main Page or something similar, of course the bot shouldn't have to undergo another RfA. Thanks! Flcelloguy (A note?) 16:29, 7 January 2007 (UTC)
I promise to get community consensus for any new features or changes from the behavior described above. As Raul654, Cyde Weys, David Levy and others, I think RFA is a strange way to go about determining consensus for a bot, and generally support a policy where all adminbot approvals are handled through Wikipedia:Bots/Requests for approval with wide community input solicited at the affected pages. However, if the community believes we must come back here to support changes then I would. Dragons flight 18:39, 7 January 2007 (UTC)
Thank you, although I strongly agree that such requests should be taken through RfA. Thanks! Flcelloguy (A note?) 01:45, 9 January 2007 (UTC)
2. What security measures have you taken to ensure that the machine ProtectionBot will be running on will not be compromised? I don't have any particular doubts about your coding ability, but is the machine properly protected against malicious hackers? Seeing as this will be an admin account, I think that appropriate measures should be taken to ensure that the account's credentials cannot be leaked. Shadow1 (talk) 17:09, 7 January 2007 (UTC)
The machine is physically located in my residence (which because the apartment owner is nuts, requires 3 keys to reach from the street). It is behind both a software and hardware firewall, kept up to date on system patches, and running up to date antivirus software. Dragons flight 18:39, 7 January 2007 (UTC)
While all of the above is good to know, I fail to see how the bot's login information would be more likely to fall into the wrong hands than that of any other account. Given the fact that Robert already has a proven track record (having never allowed his sysop account to be compromised), I would argue that there's less cause for concern than there is when someone seeks to become an administrator. —David Levy 19:07, 7 January 2007 (UTC)
The reason I inquired is because, in most cases, a bot's username and password information are stored somewhere on the system in cleartext, which could appeal to a hacker trying to gain access to an administator-level account slightly more than, say, stalking the admin and forcing them to reveal the password. Anyway, I like the security setup. Shadow1 (talk) 19:28, 7 January 2007 (UTC)
Actually, I haven't seen a bot framework that stores the password in plaintext on the user's machine. PyWiki and AWB (the only two frameworks I am familiar with) both prompt to enter a password, which is not stored; what is stored is the session cookie. If the session cookie is stolen that may allow for some problems, but having the password compromised wouldn't be one of them. --Cyde Weys 22:23, 7 January 2007 (UTC)
3. Will you agree that this Request for adminship is technical in nature only, and will not be used to bestow community administrator functions to this account (e.g. Closing xFD's; exercising clemency in releasing users from blocks; gathering material from deleted pages)? {This seems may seem obvious to some, but having it explicitly endoresed has merit). — xaosflux Talk 17:42, 7 January 2007 (UTC)
Well, duh. It's a bot, not an admin. It will do what I have said it will above, and not any of that other stuff. Dragons flight 18:39, 7 January 2007 (UTC)
4. Why on earth are you refusing to release the code? — Werdna talk
By my count one would only need to change 4 lines of code to create a bot that vandalized all vulnerable templates and uploaded porn over all vulnerable images. Give me 10 lines to change and I could knock out much of the throttling and target templates and images associated with random pages rather than the well guarded Main Page. Yes, a vandal could modify public code for AutoWikiBrowser or pyWikipedia to do much the same thing, but I refuse to make it easy for them. Dragons flight 06:45, 8 January 2007 (UTC)
Werdnabot's code needs one extra line added to it to vandalise 400 talk pages, and I've posted the code on my toolserver subversion repository. Has it happened? No. I think this is a little too much paranoia. — Werdna talk 07:18, 8 January 2007 (UTC)
Actually, Werdnabot has vandalised talk pages at least four times. I don't know how many lines of code were added – Gurch 12:29, 8 January 2007 (UTC)
It's not easy to find Werdnabot's code, either. I had to do more than a little digging to find it. Mike Peel 00:19, 10 January 2007 (UTC)
In a recent test I conducted, it took me 4 minutes to write a 24-line vandalbot using perlwikipedia. It's really easy to write a vandalbot, even without the source code of, say, AntiVandalBot. Pywikipedia even existing makes it easy. Shadow1 (talk) 20:30, 8 January 2007 (UTC)
5 If it's a sysop, why does it need to ask another admin to delete images from commons? Surely after a few test runs, it can do this its self once any bugs are fixed. -- Selmo (talk) 02:25, 8 January 2007 (UTC)
Because getting an adminbot approved is hard enough as it is. I felt that proposing a bot that could handle deletions as well would be resisted by too many people in the community, so I am not including that function. Dragons flight 06:45, 8 January 2007 (UTC)
6 Probably appears to be a stupid question, but could this bot unblock itself? I ask because of the point below about bots being blocked if they go awry. I mean, the answer to me would appear to be no since I'd assume if it could unblock itself that would imply sentience, something which if seen in a bot I'd think would perhaps breach our policy on WP:NOR, but I just want to triple check. Hiding Talk 16:26, 8 January 2007 (UTC)
No. If blocked it cannot do anything to unblock itself or in any other way affect itself or Wikipedia. Dragons flight 16:33, 8 January 2007 (UTC)
7 More importantly than 6 above, does the bot operator vow not to take any administrative actions on the bots account per the conflict of interest guidelines? Will he consent to be sanctioned if he does? Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 16:34, 8 January 2007 (UTC)
If my actions over the last three years on wiki are not enough basis for trust, then nothing I can say here is going to make any difference. But yes, I support the COI guidelines, and fully expect that my actions in this matter will be watched very closely. If the community dislikes what I am doing I agree to stop. Other than that, I have no idea what "consent to be sanctioned" means since most plausible santions would be involuntary. Dragons flight 16:53, 8 January 2007 (UTC)
Something out of this bag of tricks I would imagine. If you are using administrative powers on another account you own, then in my opinion you should lose the bit on both accounts. Do you agree? Why or why not? Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 16:59, 8 January 2007 (UTC)
I agree that ArbCom is an appropriate venue for deciding such things. If someone is maliciously abusing sysop privledges then they should generally be desysoped. This is obvious. Dragons flight 17:13, 8 January 2007 (UTC)
Are you essentially asking whether Dragons flight will unblock the bot should it be blocked? I'm not sure what other admin actions would be used on the bot. Blocking the bot himself would hardly seem controversial and protection or deletion of the bot's userpages wouldn't create any conflict of interest that I can see. Is there another administrative tool that I'm missing? SuperMachine 17:06, 8 January 2007 (UTC)
The confusion may be arising from "not to take any administrative actions on the bots account" - this could mean to take admin action on ProtectionBot, with the Dragons flight account. It could also mean Dragons flight, as a human, operating the admin tools of ProtectionBot while logged in as ProtectionBot. The former is a reasonable question. The latter has already been answered, I believe. Carcharoth 17:23, 8 January 2007 (UTC)
8 Is the bot unsupervised? and if so why? It appears as if the bot will only be makeing around 10 edits a day, I see no reason why this couldn't be watched --T-rex 17:44, 8 January 2007 (UTC)
Some portions of the Main Page change on a predictable schedule, i.e. rolling over each day. Others such as In The News and Did You Know do not. At least half the vandalism that has recently been encountered occurred as a result of these unpredictable changes. The time from the appearance of a vulnerability to an exploit in the typical case I have reviewed is a couple hours. Hence this script needs to run many times per day in order to be truly effective at catching the vulnerabilities before they are exploited. And frankly, I need to be able sleep, work, and have a life. Also, you are underestimating the number of things that have to be protected and unprotected each day. Dragons flight 01:06, 9 January 2007 (UTC)
9. I'm going to assume that you have at least some knowledge of the recent vandalism that has appeared on the main page recently. If this bot had been running one month ago, how much of the vandalism would have been prevented? --- RockMFR 22:55, 8 January 2007 (UTC)
This bot is capable of addressing every vulnerability I am aware of having occurred on the main page during that time. The only question is one of timing. In the several cases I looked at in detail, the vulnerabilities had all existed on the Main Page for at least a couple hours before being exploited. As the bot is intended to make checks multiple times per hour, all such problems would have been eliminated before they could have been exploited. While this bot cannot prevent human mistakes that leave part of the main page unprotected, it greatly narrows the window of opportunity during which such mistakes might be exploited. So without having studied every case in detail, I can't give you an exact answer, but I believe that had this bot been deployed a month ago most (and quite possibly all) of the recent vandalism would have been prevented. Dragons flight 01:06, 9 January 2007 (UTC)
Bots watch IRC channels that have change records flowing through them. Could this bot be enhanced to watch for certain things changing (the DYK next update page for example, or a list of things to watch or whatever) and triggering on that, instead of at random times? ++Lar: t/c 22:02, 9 January 2007 (UTC)
Possibly, I have no experience working off of IRC feeds. My impression was that those IRC feeds were generally created by some sort of screen scraper anyway, in which case I'm not sure I see the advantage. (Probably better to follow on this discussion somewhere other than this enormous page.) Dragons flight 22:13, 9 January 2007 (UTC)
10. What are the differences between this and User:Shadowbot2? I ask because the latter is open source. Is it just that one protects where the other only flags for protection? Abu-Fool Danyal ibn Amir al-Makhiri 15:22, 9 January 2007 (UTC)
In my opinion, the key differences at the present time are:
  1. Shadowbot2 is incomplete. It cannot currently do template recursion, so a template added within a template on the Main Page is not checked. Also, it is not able to check for protection of images hosted on Commons. Both of these issues are recently used exploits against the Main Page that my bot would be able to handle.
  2. My bot also addresses templates and images in Today's Featured Article, which is the other area in Wikipedia where preemptive protection has become routine due to aggressive and high profile vandalism.
  3. Shadowbot2's strategy is to email 22 admins whenever it finds something unprotected on the Main Page. This works, but it means that a large list of admins are effectively functioning as automatons to carry out the bidding of the bot. Since the action required is simple and unambiguous, it is more natural in my opinion to have a bot execute it directly.
  4. My bot remembers what it protected and handles the unprotection directly. This addresses the additional problem that sometimes templates and images remain protected for a very long time after they come off the main page. (I encountered some that had been protected over a year recently).
While some of the above could be addressed by refinements to Shadowbot2, I believe that what I have prepared is a more complete solution. Dragons flight 19:51, 9 January 2007 (UTC)
I have revised the above answer after it was pointed out that it came out sounding harsh. I don't mean to be critical of Shadow1, who is obviously making a well-intentioned effort to address a pressing problem Wikipedia has been having with vandalism. Dragons flight 21:19, 9 January 2007 (UTC)
It should be pointed out that User:Shadow1, developer of Shadowbot2 is at a distinct disadvantage in that he isn't an administrator and would have a very difficult time creating an adminbot. This is the only real reason that Shadow1 hasn't added on this additional functionality. He operates a second bot, Shadowbot1, which has full editing functions that could have easily been added into Shadowbot2 if he had the distinct advantage of being an admin. Should this RfA succeed, then this of course raises questions about permitting bot developers access to sysop functionality for the purposes of developing adminbots.--Kind Regards - Heligoland (Talk) (Contribs) 20:12, 9 January 2007 (UTC)
The other reason I don't wish to add adminbot functionality to Shadowbot2 is so that I don't need to go through RFA. I think that, while adminbots have their merits, due to the bot's administrative privileges, there are more places it could potentially fail. That, and it would be extremely difficult to do, seeing as I am not a sysop. Shadow1 (talk) 14:33, 10 January 2007 (UTC)
11. Given that Shadowbot2's source code has been released, why do the developer(s) of ProtectionBot continue claiming that releasing the source code would facilitate the creation of a vandalbot which can search for unprotected templates, when such source code necessary for the creation of a vandalbot has already been released into the public domain ? --Kind Regards - Heligoland (Talk) (Contribs) 15:39, 9 January 2007 (UTC)
Yes. Shadowbot2 does not edit pages and it does not upload images. Consequently, though it could be turned into a vandalbot, it would take a more knowledgable person to do so. Consider the following schematic:
ProtectionBot Anti-ProtectionBot
  1. Find vulnerable images and templates
  2. Upload missing images from Commons
  3. Protect everything
  4. Add tags noting the protection
  1. Find vulnerable images and templates
  2. Upload shock images from hard drive
  3. Do nothing
  4. Vandalize all vulnerable templates and images
The difference between those two scenarios is changing 3 function calls. Admittedly, there would still be a lot of leftover code in Anti-ProtectionBot that it wouldn't need, but a thoughtful script kiddie could easily cut out the irrelevant bits to make it a more efficient vandal. With only a little more effort he could add code to direct it at places other than the Main Page.
I don't like the above scenario. I have liberally shared the code with Wikipedians who have asked me for it directly, and will happily continue to share it with other bot operators and those who seem legitimately interested in making it better. However, I will not make it public. This is based in no small part on my assumption that the vandals attacking Wikipedia will learn of ProtectionBot (if they aren't already aware of these discussions) and that they could easily turn this against us. Among other forms of evil, I do not want to contribute to Wikipedia's first image uploading vandalbot. Dragons flight 19:51, 9 January 2007 (UTC)
PS. I would also like to note that there is only one English Wikipedia, but there are hundreds Mediawiki sites that would be endangered if someone really committed themselves to running a vandalbot. Dragons flight 21:22, 9 January 2007 (UTC)
As far as I can tell, ProtectionBot, or virtually any bot currently running, can be turned into a vandalbot. It's really not a difficult task; it's actually probably easier to write one from scratch than to convert an existing bot. Shadow1 (talk) 14:33, 10 January 2007 (UTC)
12. One of the primary counter-arguments to your logic against releasing the source code has been, effectively, "it's just as easy to write a vandalbot with AWB, the pywikipedia framework, the relevant Perl module, etc". What I cannot seem to find, however, is any rebuttal from you of these claims. What about ProtectionBot is so different that it must be closed source? What is the "secret sauce" that you are afraid of revealing? The only thing I can imagine is the algorithm to determine which templates need to be protected, but, no offense, figuring that out is really not that difficult, and obviously the vandals have already been figuring it out - after all, if they hadn't, there would be no need for this bot. What else, exactly, needs protecting? Krellis 00:57, 10 January 2007 (UTC)


[edit] General comments

  • Perhaps discussions, questions and responses to Oppose !votes could be moved to the discussion page. There seems to be sizeable comments in response to the points raised by those Opposing. It's likely this RfA could become quite long just with !votes. --Kind Regards - Heligoland | Talk | Contribs 16:29, 7 January 2007 (UTC)
  • For those worried that the code may have some secret admin backdoor, I have read and understood it, it does only what it claims to do. The code looks good. HighInBC (Need help? Ask me) 17:07, 7 January 2007 (UTC)
  • WP:BAG bot policy enforcement:
    BAG members please endorse this section if you agree with this statement: Should this bot perform any sysop actions that have not been approved in BOTH WP:RFA and WP:RFBOT this account will be immediately blocked.
    Endorse, — xaosflux Talk 17:35, 7 January 2007 (UTC)
    Endorse —Mets501 (talk) 18:46, 7 January 2007 (UTC)
    Endorse, and I might also push for desysoping both bot and bot operator depending on the circumstances. Betacommand (talk • contribs • Bot) 19:45, 7 January 2007 (UTC)
    Endorse -- Tawker 07:43, 8 January 2007 (UTC)
    Endorse --lightdarkness (talk) 07:48, 9 January 2007 (UTC)
    Endorse -- RM 13:33, 10 January 2007 (UTC)
  • As I understand it, ProtectionBot will not be operated for the next week while this RfA is underway. It seems to me that if Dragons flight is willing, the bot should operate during this period, perhaps on the same basis as during its trial, both so we have the protection during the week, and so in the event of any issues involving the bot they can be addressed while the RfA is still pending. I've posted a thread to discuss this matter on the talkpage. Thanks, Newyorkbrad 23:08, 7 January 2007 (UTC)
  • Jimbo has spoken (copied from his talk page):

It would please me for the code to be released under a free license. I do understand the security concerns about that, but find them not fully persuasive. It strikes me as quite easy for someone to write a bot to vandalize wikipedia, and keeping bot software secret in an attempt to prevent vandalbots from being created by fools is security through obscurity... it might work for a while, but it gives a false sense of safety. As a separate comment, I should add that I am 100% in favor of bots being given the admin bit selectively.--Jimbo Wales 20:10, 8 January 2007 (UTC)

Cbrown1023 22:32, 8 January 2007 (UTC)

  • I do want to make a comment (more an anecdote, I should say) for some of those opposing: yesterday, I had to protect and upload all of the images and templates that were going to appear on the Main Page the following day, and also unprotected and deleted the images and templates that were off the Main Page. I usually am on right after the day changes in UTC, but because of other obligations I was about two and a half hours after the day had changed. I was still surprised, though, that none of those images/templates (which were to appear on the Main Page in less than 24 hours) were protected, with several other admins now attempting to take care of the task. Whether or not those images and templates would have gone unprotected if I had not protected them then, I cannot tell, though I hope other administrators would have noticed before they went on the Main Page, and Shadowbot1 would have caught the unprotected templates if they had actually gone on the Main Page. However, we cannot continue to depend solely on human involvement in this matter; as seen last month, we've had several things slip through the cracks, and we can't afford to continue this risk. A bot would virtually assure that nothing slips through the cracks and would be like clockwork in protecting and unprotecting those images scheduled to be on the Main Page. This bot would be a huge improvement over the current system - it's not that I and the other administrators are not willing to spend the time and effort to do this daily, because we are - it's just that a bot would drastically improve our efficiency and reduce the chance of vandalism ever being on the Main Page. Please consider the useful task that this bot could do. (Also, a quick note about those concerned about the code: while I appreciate your concerns, I think the bot is still crucial even without the code released. I was one of those who requested (and received) the code, and though I'm certainly not an expert, it looks fine; though personally, if I was the programmer of the bot, I would release the code publically, it's a personal decision by Dragons flight, and I can also most certainly understand his hesitation, given the security concerns.) Thanks! Flcelloguy (A note?) 01:58, 9 January 2007 (UTC)
  • Some history of one of the examples of vandalism that occurred on Christmas Eve:
This is the sort of thing ProtectionBot should put a stop to. Carcharoth 02:13, 9 January 2007 (UTC)

Hi all,

I've just completed work on modifications to the MediaWiki software's protection system. One such change is the ability to protect a page in a manner known as a "cascading protect" (an additional checkbox appears on the protect page for this); which will protect all templates and images included on the page, as well. It's currently sitting in a branch, with principle approval by brion, but it needs a day or so of review by other developers. I would estimate that it will be live in a few day's time, pending schema changes and code review. Given that this RfA still has four days left, I would suggest that the bot's functionality will be duplicated in the MediaWiki software by the time the RfA finishes. I would therefore question whether or not we should have a bot that is going to solve a problem already solved in the MediaWiki software. For those interested, my work can be found in the branches/werdna/restrictions-separation folder on the MediaWiki subversion repository.

— Werdna talk 10:21, 10 January 2007 (UTC)

That's great Werdna. I will happily disable any functionality that is made obsolete by improvements in Mediawiki. Your work, if I understand it correctly, doesn't totally replace the tasks for which this bot was designed, but does address the largest part of the concern regarding the Main Page. It doesn't appear to address the issue of unprotected images on Commons, which would still need to be uploaded locally and is one thing this bot could still do. Your work also doesn't address Today's Featured Article, since the article itself is not protected.
If I had to guess, I would think you are being a bit optimistic and that cascading protection is a sufficiently sophisticated task that it will take more than the remainder of this RFA to approve. Regardless, if the status changes, please update that here. Dragons flight 16:17, 10 January 2007 (UTC)
I'm actually somewhat concerned by your sudden revelation of plans to protect templates used in the featured article.. If I'm not mistaken, this isn't actually common practice at the moment; and I find it somewhat deceptive that this feature wasn't given the prominence it required; rather the main page vandalism was used as a doom-and-gloom scare tactic to make people feel like this bot was absolutely necessary. At any rate, I think you're being a bit pessimistic about the power of early hours and commit access :-). Developers don't normally need to go through a formal approval process in order to have their changes applied; although given the magnitude of this change, I'm working closely with Brion Vibber and Tim Starling to ensure its successful and bug-free operation. — Werdna talk 17:11, 10 January 2007 (UTC)
It's in the second sentence here and the first sentence of the BRFA summary description. And yes, templates in Today's Featured Article have been preemptively protected for some time now, though it is understandable if many people don't follow TFA closely enough to know that. I'm 200% certain that you won't be spending a week writing out discussion like this one, so that certainly will make things easier, but I'd still wager there will be enough adjustments and tests to take a week or two. Dragons flight 17:24, 10 January 2007 (UTC)
You'd be fairly likely to lose your money. The code is functionally complete, and has been tested on my own wiki fairly thoroughly. I'd estimate it would likely be ready within the next two to three days remaining on this RfA. But I digress. My main point is that all of this stuff should be implemented in software anyway. You can protect the FA autoconfirmed with cascade on if you so desire, and that'll full-protect all the templates on it. If you leave it unprotected, then it's not really particularly hard to vandalise the article itself anyway, so there's not much point in bothering protecting the templates. — Werdna talk 17:40, 10 January 2007 (UTC)


[edit] Discussion

  • AntiVandalBot could potentially cause more problems than this thing. -- Steel 10:32, 7 January 2007 (UTC)
    • Yes and no. I feel you made this statement by comparing AVB's massive activity (up to 100s of reverts and warnings hourly) with PBs rather low-profile potential activity (few uploads and protections daily). Still, points to consider include that 1. an admin bot can be shut down nearly as easily as a non-admin one and 2. every action, admin's or not, is reversible. With that in mind, I'd say both can do only little mess. Миша13 12:05, 7 January 2007 (UTC)
      • My intention was to emphasise how little damage this bot could do by comparing it to another bot which is widely accepted amongst the community, yet much faster moving. I totally agree that neither bot can do an awful lot of harm. -- Steel 22:33, 7 January 2007 (UTC)
        • The ironic thing was the amount of opposition AntiVandalBot recieved when it first started RC patrolling. Lots and lots of people thought it wasn't work a bot could do and that RC patrolling could only be done by humans. Well, AVB isn't the be all and end all, it doesn't cure all our vandalism problems but I guess it helps a bit. As for fast moving, the bot really only does an edit a minute on average, not exactly fast moving. I guess it's the fact that it basically runs unattended now that is the big factor. Sure, it's messages are checked, but I don't exactly check every edit it makes anymore. After about 25,000 of em I got a little sick of it :o -- Tawker 07:43, 8 January 2007 (UTC)


[edit] Support
  1. Support (first!) to let the bot run. Despite being an admin bot, I believe it is necessary for the 'protection' of the main page, as they recieve numerous accounts of vandalism.(was approved) Insanephantom (please comment on my Editor Review!) 09:29, 7 January 2007 (UTC)
  2. Support based upon the Bot RfA comments and amendments. (aeropagitica) 09:36, 7 January 2007 (UTC)
    Support - well-defined purpose and limited remit. Has been tested. Should be no problem, and safety measures are in place. Needed to resolve a recently disruptive episode of vandalism to the main page that resulted in many e-mails of complaint. Carcharoth 09:46, 7 January 2007 (UTC)
    Withdraw my support for now until developer concerns on talk page (and, I've just noticed, above as well) about not using this bot on the featured article (because the article itself remains unprotected) are fully addressed. Carcharoth 22:54, 10 January 2007 (UTC)
  3. Support and welcome our bot overlords --Steve (Slf67) talk 10:01, 7 January 2007 (UTC)
  4. Support I'm not the biggest fan of admin bots, however good nom, and the bots purpose seems needed Brian | (Talk) 10:03, 7 January 2007 (UTC)
  5. Support I don't mind admin bots as long as they have been tested and their purpose has been clearly defined. GizzaChat © 10:09, 7 January 2007 (UTC)
  6. Steel 10:32, 7 January 2007 (UTC)
  7. Support per Gizza. This is badly needed. I'm not even sure if I feel that an RfA is really necessary as I it seems well supported at BRFA, but I guess an RfA can't hurt. --PS2pcGAMER (talk) 10:38, 7 January 2007 (UTC)
  8. Support clearly needed and safe to use. recommend to allow operation on personal account until this RfA is decided Agathoclea 10:39, 7 January 2007 (UTC)
  9. Support. About time. MER-C 11:02, 7 January 2007 (UTC)
  10. Support. Kusma (討論) 11:22, 7 January 2007 (UTC)
  11. Support - it is deeply needed. Even with the excellent Shadowbot2 reports, we've had incidents on the main page recently. Martinp23 11:33, 7 January 2007 (UTC)
  12. Necessary, completely reversible. --Slowking Man 11:42, 7 January 2007 (UTC)
  13. Full support. Necessary. Actions are auditable by moving them to a separate account. A big red button on its page is enough to prevent harm. If the code isn't be publicly released, I won't bother me much - the bot's tasks are simple and if Dragons flight vows that there's no self-unblocking mechanism plugged in (why would there be any?), it's fine by me. (Bottom note: my healthy interpretation of the policies is that adminship is given to a person, not an account, so this RfA should IMO be a speedy promote.) Миша13 11:55, 7 January 2007 (UTC)
  14. Supportper Gizza. RHB 12:03, 7 January 2007 (UTC)
  15. Support, if it does not operate as desired or becomes obsolete, we can desysop at the drop of a hat (without the normal drama). This is a very low risk, high reward for the project thing to do, so let's do it. NoSeptember 12:33, 7 January 2007 (UTC)
  16. Support. Dragons flight is already an admin and does a good job running bots and the reasons stated for why this account here needs admin rights are valid. Dragons flight can fully be trusted to use this account carefully and as explained (and for nothing else). This is not a cookie cutter RfA. So, discussion contributors should carefully check the motivation behind this RfA and think about how the underlying problem (protecting templates affecting the main page) should be solved without this bot if this RfA should not pass. Many thanks to Dragons flight for doing all this work. --Ligulem 12:38, 7 January 2007 (UTC)
  17. Support - this is needed urgently. IAR and common sense, please. Moreschi Deletion! 12:47, 7 January 2007 (UTC)
    Unfortunately, IAR's status as policy is disputed, and common sense is actively discouraged. Urgent need is a good argument, though – Gurch 16:07, 7 January 2007 (UTC)
  18. Support - There was a time when some admin actions, such as image deletion, could not be reversed. That time has passed. Thus, the only 'danger' represented by an 'admin bot' (like any bot) is that it will make numerous incorrect changes that would take a long time to clean up. However, this bot would be performing a very specific and simple task with a narrow scope. The chances of it 'going berserk' seem virtually nil and if it did the log of excess protections / unprotections could easily be used to create a list for equally quick reversal. Finally, bots which perform admin functions should be subject to more widespread review than normal bot approval requests and having an RfA may be a reasonable way to achieve that, but can we ditch the stuff about the bot not having accepted the nomination? Or does it have to be coded to do that... which it could. --CBD 12:51, 7 January 2007 (UTC)
  19. Support - Much needed — Lost(talk) 12:54, 7 January 2007 (UTC)
  20. Support It is about time. --Siva1979Talk to me 13:12, 7 January 2007 (UTC)
  21. Support ← ANAS Talk? 13:15, 7 January 2007 (UTC)
  22. Support - good potential for solving a serious vandalism problem --BigDT 13:24, 7 January 2007 (UTC)
  23. Support, this one will solve a lot vandalism problems on Wikipedia, this is needed. Terence Ong 13:27, 7 January 2007 (UTC)
  24. Support Seems like a good idea —Pilotguy (ptt) 13:44, 7 January 2007 (UTC)
  25. Support. SD31415 (SIGN HERE) 13:52, 7 January 2007 (UTC)
  26. support: great user, answers to questions are awesome. (Actually, this bot could really help!)--Ac1983fan 14:18, 7 January 2007 (UTC)
  27. Support - definitely a useful bot. --tennisman sign here! 14:26, 7 January 2007 (UTC)
  28. Support. This is necessary. Nishkid64 14:44, 7 January 2007 (UTC)
  29. Support. Completely necessary. — Dark Shikari talk/contribs 14:47, 7 January 2007 (UTC)
  30. This whole idea of an RFA is utterly unnecessary. The bot's op is already entrusted with the sysop bit, so I don't see the necessity of making him go through this extraneous process to get a second one. He's already trusted, and bot ops are responsible for the actions of their bots, so, granting this bit should just rely on Bot Approvals Group or bureaucrat discretion. If it malfunctions it can always just be blocked. RFA is entirely geared towards determining the acceptability of human candidates to be admins; try applying the process to a robot candidate whose human owner is already an admin makes no sense. --Cyde Weys 15:02, 7 January 2007 (UTC)
    Question here, why can't Dragon's flight just put the code into his monobook and avoid this process altoghether? Just because you've passed an rfa doesn't mean you can say "I feel like making a sockpuppet, since i've already passed rfa, the sock is now an admin too. And that's basically what we're talking about here, adminning a sock. A non-voting, white hat, well intentioned sock, but a sock nonetheless. Just H 20:06, 8 January 2007 (UTC)
    But...but... eww, adminbots! – Gurch 15:43, 7 January 2007 (UTC)
  31. Support. I've been looking through dozens of template histories to find a vandal edit and in the meantime a giant penis was sitting on the main page. I don't want to do it again. – Elisson • T • C • 15:07, 7 January 2007 (UTC)
    "a giant penis was sitting on the main page"? What, recently? Today? If you could give details, it would be instructive to work out whether ProtectionBot would have prevented this. Carcharoth 20:09, 7 January 2007 (UTC)
    I can't remember the exact dates, but there have been at least six incidents in the last month. The bot would have prevented all of these, because every one of them was caused by an admin changing a template or image on the main page (or one of the daily templates changing itself automatically) and they forgot to protect it – Gurch 12:19, 8 January 2007 (UTC)
    I can't remember the exact date either, but IIRC, this was one of the first of all the incidents in December. This was before the transclusion list showed which templates are protected and which are not, so it was even more templates that had to be checked than it is today. As Gurch says, the ProtectionBot would have prevented this. – Elisson • T • C • 13:11, 8 January 2007 (UTC)
    To be certain about this, we would need to track down as many of these instances of vandalism as possible and, given the response time of ProtectionBot (please don't bandy exact numbers around), state whether the vandalism would have been prevented or not. Plus also noting instances where (for the 'transcluded onto the featured article' templates) ordinary users were quicker than ProtectionBot to revert vandalism (by which I mean users were quicker to remove a newly-added and vandalised template, whereas now ProtectionBot would just protect said vandalised newly-added template). Carcharoth 13:26, 8 January 2007 (UTC)
  32. Strong Support. I'm happy to support and have no concerns with the idea of a bot with sysop status, and I would also suggest Shadowbot2 is given a full bot flag (if not given one already) and is entrusted with a backup role in case there is any problem with ProtectionBot, you know, just in case. --Kind Regards - Heligoland | Talk | Contribs 15:19, 7 January 2007 (UTC)
  33. Support a needed bot and DF is certainly a very trustworthy person to run it. Anyone who thinks we don't need a bot for this should volunteer to do this task every day for the next 6 months :-) --W.marsh 15:24, 7 January 2007 (UTC)
  34. Strong support No concerns, it has only been coded to protect and serve (yes you may groan and vandalise my talk page for that), it cannot code itself, so it won't be going on a spree and/or code itself to block or delete. Unless penis becomes a featured article, it should solve the problems ;) --Alf melmac 15:28, 7 January 2007 (UTC)
  35. Strong support; though I've been hesitant about giving a bot administrative rights in the past, I truly believe that this one is necessary. I speak from personal experience - it's no fun spending time protecting, unprotecting, uploading, and then deleting multiple images and templates each day, especially when you could be doing more productive issues that a bot couldn't do. Though my work was called "like clockwork" back in May, when I regularly did the protection work, a bot like this could be much more like clockwork than any human could ever be. Sure, there will be some issues with the bot malfunctioning, like with any other bot, but such errors are inevitable with any bot, and are still an improvement over human protection (which, as we've seen, can let things slip through at times). I've also appreciated the way Dragons flight has run this whole process: he has responded to my (and everyone else's) concerns, questions, and suggestions with great patience and concern, has taken the necessary steps prior to bringing this RfA, and has also shared the code with several people who requested it, myself included. Flcelloguy (A note?) 15:42, 7 January 2007 (UTC)
  36. Support. I trust the owner so I trust the bot, and we need it. Antandrus (talk) 15:45, 7 January 2007 (UTC)
  37. Support, a rare case where I'm willing to waive my requirement of having a lot of non-trivial experience in the main namespace. I trust the owner to be as accountable for this bot's actions as for his own actions. — mark ✎ 15:59, 7 January 2007 (UTC)
  38. Support Bot owner is trusted. The bot is operated for a very good purpose, and cannot function without the priviledges of an admin. =Nichalp «Talk»= 16:01, 7 January 2007 (UTC)
  39. Support of course. This shouldn't even be an RfA, Dragons Flight already went through this process. Perhaps an RfA should also be undergone each time admins switch browser software or upgrade their computers?- CHAIRBOY (☎) 16:03, 7 January 2007 (UTC)
    I can see it now. Strong Oppose – using Internet Explorer 5? What is this guy thinking? – Gurch 16:33, 7 January 2007 (UTC)
  40. Support rationale seems to make sense -- Samir धर्म 16:07, 7 January 2007 (UTC)
  41. Support, especially due to the recent incidents with vandalism to Main Page transclusions. G.He 16:07, 7 January 2007 (UTC)
  42. It is an insult to ask a human to do a machine's job. —Cryptic 16:26, 7 January 2007 (UTC)
    Not on Wikipedia. Here, machines can only do machine's jobs, but humans have to do both – Gurch 16:30, 7 January 2007 (UTC)
  43. Support, this bot is needed to help prevent further vandalism on the main page and I fully trust the creator. Naconkantari 16:31, 7 January 2007 (UTC)
  44. Support I have read the source code, and if ran as is will act as a responsible and helpful admin. Any further tasks should require additional approval. HighInBC (Need help? Ask me) 16:31, 7 January 2007 (UTC)
    Actually it will be significantly better than a responsible and helpful admin. Even they have to sleep – Gurch 16:35, 7 January 2007 (UTC)
  45. Support: Machines are good at rote, routine jobs - they don't get bored, and to the extent the code is good, they don't make mistakes. The Main Page article is by far the most seen page in Wikipedia, so minimize vandalism to it is very important. Let's use automation where it makes sense (as it does here) to free up the time of admins to do other, less rote and routine things where human beings really are needed. John Broughton | Talk 16:40, 7 January 2007 (UTC)
  46. Support. At first I was kind of worried about it going crazy and blocking everyone or something, but I realised that that really just won't happen. No administrator can be held responsible for the recent spree of vandalism on the main page, even if it could have been avoided by page protection. This bot is a solution to that problem. --Deskana (For Great Justice!) 17:05, 7 January 2007 (UTC)
  47. Support. Generally I'm cautious about admin bots but in this case its crucial that we don't allow the MP to be regularly vandalised as its often the first page a new user sees. --Spartaz 17:26, 7 January 2007 (UTC)
  48. Strong, strong, strong support. There's really no question here. —bbatsell ¿? 17:27, 7 January 2007 (UTC)
    Why is there no question? Just H 19:53, 7 January 2007 (UTC)
    Well, there is now, you just asked one. Now I've answered it, though, there isn't one again – Gurch 12:17, 8 January 2007 (UTC)
    Yes there is, you didn't answer why there is no question. Just H 20:02, 8 January 2007 (UTC)
    I'll just go ahead and quote what I wrote earlier today on wiki-en:
    This bot protects all templates and images transcluded on the Main Page. That's it. It doesn't protect articles linked to the front page or anything of that sort — it simply handles the extraordinarily tedious process of going through every template used on the front page, checking every template to see if *it* has any templates that need to be protected, copying transcluded images from Commons, protecting them, then unprotecting the templates that are off the front page and requesting deletion of the copied images (it doesn't perform any deletions itself).
    These steps are so tedious and error-prone that we have screwed up numerous times already, and that has been exploited. I'm not sure how many more times we can allow pictures of cleft penii or vaginal sores ON OUR FRONT PAGE without losing tremendous amounts of credibility. —bbatsell ¿? 01:28, 9 January 2007 (UTC)
    Why do people keep getting this wrong? This is the second time I've had to correct someone. If you have followed what has been going on, and have read the statement above and the bit at User:ProtectionBot, you will see that this bot does more than protect all templates and images transcluded on the Main Page (in fact technically it only protects and unprotects unprotected images and templates - it leaves already protected stuff alone). It also protects templates and images a day in advance of when they are scheduled to be on the Main Page (using Main Page/Tomorrow) and it then protects any unprotected images or templates added to the templates used on the Main Page, or added directly to the Main Page (though this is very rare). But the big bit you omit with your "This bot protects all templates and images transcluded on the Main Page. That's it.", is the fact that the bot is designed to to the same thing for the daily featured article. I quote: "ProtectionBot is designed to combat these attacks by automatically protecting all templates and images appearing on the Main Page and Today's Featured Article" and "It does this by reading Main Page, Main Page/Tomorrow, Template:Did you know/Next update, today's featured article, tomorrow's featured article and generating a list of all images and templates used therein.". So saying "That's it" is missing a lot of what it does. I support this as well, but please don't give people the wrong impression about what it does. Carcharoth 01:50, 9 January 2007 (UTC)
    I guess I should have clarified. That was a copy/paste from an email addressed to someone on en-wiki who objected to the bot because he thought it was going to protect everything linked from the front page (at least that is what I interpreted his objection to be). I know very well exactly what the bot is going to do (and have corrected others who had incorrect notions further down the page myself), and I don't think that I said that it wasn't going to protect future templates transcluded on the main page. My main point was the second paragraph. It's pretty clear from Just H's other comments that he was unaware of the amount of trouble we're having with the main page, and I think that some non-admins (and even some admins who aren't aware or pay attention to main page stuff) are objecting because they don't realize how big of a problem this is. That's all I was trying to say. —bbatsell ¿? 01:55, 9 January 2007 (UTC)
    OK, so you quoted the "That's it" out of context. Now you've provided a fuller context, I can understand what you were saying. Thanks. Incidentially, I have some useful links demonstrating what happened in the biggest flare up of this problem that I saw, on Christmas Eve:
    I'll post those in the general comments section as well. Carcharoth 02:08, 9 January 2007 (UTC)
    Yeah, I should have edited it down instead of being lazy. Sorry about that :P —bbatsell ¿? 02:32, 9 January 2007 (UTC)
  49. Support. Any additional main page vandalism will not be tolerated, and I don't think we should overwhelm adminstrators to continuously protect those templates every day. I trust this bot and his operator. Michaelas10 (Talk) 17:36, 7 January 2007 (UTC)
  50. Support. I've opposed adminbots in the past, but I believe the circumstances here warrant granting adminship to ProtectionBot. The bot's scope is narrow and a need has been demonstrated. We shouldn't hesitate to use technology to combat vandalism, as long as the benefit outweighs the cost. SuperMachine 17:40, 7 January 2007 (UTC)
  51. Support per my comments on ANI (or wherever the discussion was moved to). This will help to solve a problem that is very important to solve, and only in the most technical sense will the bot be performing an administrator function. Newyorkbrad 17:42, 7 January 2007 (UTC)
  52. Strongest possible support, though I reiterate my belief that RfA is not an appropriate forum in which to discuss this matter. —David Levy 17:53, 7 January 2007 (UTC)
  53. Support. We need this. -- Donald Albury 18:15, 7 January 2007 (UTC)
    Why? Just H 19:53, 7 January 2007 (UTC)
    RyanGerbil10 said it better than I could. Protecting the main page and its contents is very important, and protecting all of the templates involved seems to be more than human intervention can keep under control. -- Donald Albury 20:49, 7 January 2007 (UTC)
    Thanks! That's all I wanted to know. Just H 01:07, 9 January 2007 (UTC)
  54. Weak support. I don't like bots as admins by nature, but this does look like a great bot and it would certaily use the admin tools well.--Wizardman 18:17, 7 January 2007 (UTC)
  55. Weak Support Seems fine, I would prefer it be open source, and don't really buy the reasons for it not being so, but it's not the end of the world, and it would be useful. - cohesion 18:41, 7 January 2007 (UTC)
  56. Strong support - as a member of the bot approvals group who has reviewed the source code, I see no problems with this bot. —Mets501 (talk) 18:46, 7 January 2007 (UTC)
  57. Support Just don't start adding any functions that are not approved. ;-) Prodego talk 19:17, 7 January 2007 (UTC)
    The temptation to do that and the lack of oversight to stop that seems to make what you said an very strong possibility. Just H 19:53, 7 January 2007 (UTC)
    You seem to be implying that Robert is untrustworthy. This is contradicted by the fact that he's an administrator (and already possesses the technical capability to run the bot via his main account), and it adds a certain degree of hilarity to your claim elsewhere on this page that others have failed to assume good faith. You also seem to be suggesting that the proposed setup would somehow remove "oversight," and I'm baffled as to how this is so. —David Levy 20:16, 7 January 2007 (UTC)
    My apologies. I did not mean to imply that Robert is untrustworthy. I meant to imply that anyone is untrustworthy of such power in Wikipedia at the current time. I meant no disrespect to Robert's abilities, which were confirmed through an rfa of his own. All i've been saying is that we should do other things first before giving control to potentially harmful machines-- I guess you could say, a retooling of the "source code" of Wikipedia's procedural structure. Just H 20:32, 7 January 2007 (UTC)
    You distrust everyone (including the most trusted members of the community) indiscriminately, and yet you complain when you believe that people are failing to assume good faith on your part. Does that make sense to you? —David Levy 20:40, 7 January 2007 (UTC)
    I added that smiley for a reason: Dragons flight has been trusted with the resposibility of adminship because of the community's trust in him. If he were wanted he could easily run this on his own account with no permission, and violate our trust in him. Admin bots are rare, and there has never been permission given to one, though at least one admin ran one without permission. The relevance is that DF can be trusted not to add things without permission, because he requested permission in the first place, which he could probably have gotten away with not doing. </rant> Prodego talk 03:04, 10 January 2007 (UTC)
  58. Support. Protection of the main page is paramount. Operating user is trusted. Thank you for creating this bot, Dragons flight. --Fang Aili talk 19:34, 7 January 2007 (UTC)
  59. No concerns, opposition doesn't raise any significant issues and doesn't appear to be approaching the issue in a sensible way. Christopher Parham (talk) 19:43, 7 January 2007 (UTC)
  60. Strong Support. Absolutely necessary. RyanGerbil10(Упражнение В!) 19:48, 7 January 2007 (UTC)
    Can you elaborate on the necessity of needing this bot? Just H 19:55, 7 January 2007 (UTC)
    The main page is viewed millions of times a day. There is no reason we should not have an automatic means of combatting this pointless and disgusting vandalism. The maze of templates and images connected to the front page is too much to do on a regular basis. Admins are not Gods. We must sleep. We must go to work. We must eat. We must live our lives. That is when the vandals strike. RyanGerbil10(Упражнение В!) 20:01, 7 January 2007 (UTC)
    Thanks! That's all I needed. Just H 20:32, 7 January 2007 (UTC)
  61. Needed admin bot, if something goes wrong which I doubt, then it could quickly be blocked and desyropped. Jaranda wat's sup 19:49, 7 January 2007 (UTC)
    See previous supporter. Just H 19:55, 7 January 2007 (UTC)
    Can you elaborate on why it's not needed? Idontlikeit isn't a real reason to oppose. -Royalguard11(Talk·Desk·Review Me!) 20:19, 7 January 2007 (UTC)
    Likewise "I like It" isn't a real reason to support. Just H 20:49, 7 January 2007 (UTC)
    But "we need it" is an excellent reason to support. So come up with a reason why we don't need it for a real argument. -Royalguard11(Talk·Desk·Review Me!) 21:17, 7 January 2007 (UTC)
    No, that isn't an excellent reason unless the supporter explains why "we need it". That's what i'm asking for here. Just H 20:02, 8 January 2007 (UTC)
  62. We shouldn't even have to go through all this silly bureaucracy just to re-approve a bot that has already been approved and to grant redundant sysop access to a user that already has it. Adminbots should be approved on RfBA, with an announcement and a link there from RfA. But since we're doing this anyway, I will of course support. —Ilmari Karonen (talk) 20:05, 7 January 2007 (UTC)
    Please note this bot HAS NOT BEEN FULLY APPROVED it as been tested and final approval hasnt been give yet as it hinges on this discussion Betacommand (talk • contribs • Bot) 02:45, 8 January 2007 (UTC)
  63. Yes To all those people who oppose on code: What I would get from that is a distrust in the bot operator. AGF for wiki's sake! To those afraid of admin-bots: Guess what, we've entered the age of wiki where we need some automation in order to keep up with everything. Do they still built cars by hand at each step? No. The time has come where we need this. Get over it. -Royalguard11(Talk·Desk·Review Me!) 20:16, 7 January 2007 (UTC)
  64. Certainly. It's necessary and the people involved are trustworthy. Sandstein 20:37, 7 January 2007 (UTC)
  65. Support. Vandalism to the main page is a concrete reality, while many of the reasons for opposition so far have been fear of hypothetical problems resulting from bot malfunction. As User:Cyde said below, "If it malfunctions (which is unlikely), it can easily be blocked by any of our over 1,000 administrators." This bot will do a great amount of good, and it would be well to recall that neither sysopping nor admin actions are irreversible, and a bot can be easier to reign in than many humans. Dar-Ape 20:49, 7 January 2007 (UTC)
  66. Support John254 21:18, 7 January 2007 (UTC)
    Can you explain your reasoning? Just H 20:02, 8 January 2007 (UTC)
  67. Support. Protection of the main page is an important task but shouldn't distract administrators from other duties. This bot is the perfect solution to our recent problems. Canderson7 (talk) 21:36, 7 January 2007 (UTC)
  68. Support. I think this RfA is totally necessary given that a new account is being sysop access. There are not enough contributors voting in the average RfBA for that to suffice. I am convinced both through personal experience of mainpage FA vandalism and the arguments made in favour of Protection Bot here, at WP:ANI and at its RfBA that this Bot will be highly beneficial to Wikipedia. Its operator is a trusted member of the community. WJBscribe (WJB talk) 21:37, 7 January 2007 (UTC)
  69. Weak Support only insofar, and on condition precedent that, bot will perform duties detailed in nomination.-- danntm T C 21:41, 7 January 2007 (UTC)
  70. Support. About time. - BanyanTree 21:45, 7 January 2007 (UTC)
    Can you explain your reasoning? Just H 20:02, 8 January 2007 (UTC)
    Sure. As an admin who has spent a decent percentage of my wiki-time on Main Page issues, it is clear that humans are not particularly good as robotically carrying out repetitive tasks, hence the need for a robot. The recent template vandalism spree in particular was ridiculous. I welcome this mechanism, which I judge will significantly reduce the occurrence of Main Page vandalism. - BanyanTree 01:47, 9 January 2007 (UTC)
  71. Support, I've dealt with the tedium of protecting the images (and forgotten to do so for up to five seconds after updating sections of the Main Page) before; in the hands of a trusted user, anything that helps to automate the process is more than welcome to hold the mop. Just make sure its password is a bit harder to guess than those used to control the robots of Hollywood. :) GeeJo (t)⁄(c) • 21:58, 7 January 2007 (UTC)
  72. Support. You've convinced me. I think the risks are being overplayed and the benefits outweigh them. — Trilobite 22:19, 7 January 2007 (UTC)
  73. Strong Support - I'm tired of seeing shock images added to templates used in templates used in today's featured article, and watching RC patrollers tripping over themsleves to find the source of the vandalism (and often causing collateral damage in the process). Having this automated and having the code not public are just common sense. Savidan 23:01, 7 January 2007 (UTC)
  74. Support - good useful adminbot. Commons uses adminbots for ages and it does not create problems. Alex Bakharev 23:18, 7 January 2007 (UTC)
  75. As I said in the request for approval, Super-Über-Mega-Speedy SUPPORT. This is a very one-of-a-kind situation, and cuts to the very core of Wikipedia's ability to be any kind of trusted website. I'd admonish the opposition to try and protect every image and template that appears on the main page, all the time. We still haven't gotten it right after about a month of this nonsense. Grandmasterka 23:29, 7 January 2007 (UTC)
  76. Support, unequivocally. We need this. Guy (Help!) 23:35, 7 January 2007 (UTC)
    Can you explain why "we need this"? Just H 20:02, 8 January 2007 (UTC)
  77. Support. I, for one, welcome our new ProtectionBot overlords. We'll make great pets. -GTBacchus(talk) 23:59, 7 January 2007 (UTC)
    Is that a supporting reason or just a regurgitation of a meme? Just H 20:02, 8 January 2007 (UTC)
    Actually, it's a regurgitation of two distinct memes. I haven't got any new substance to add to the discussion, but I trust the person running the bot, and I understand that protecting and unprotecting items as they come and go on the Main Page is a tedious and easily automatable task. The benefit of allowing this bot to run is that we have better control of Main Page vandalism, which is a real problem. The only disadvantage I can see is the speculative and unlikely suggestion that this 'bot will somehow go beyond its programmed functionality. The wording of my !vote was an acknowledgement that I'm aware of the motivation for opposing, and an indication that I don't think it's a credible threat, while possibly provoking a chuckle or two with my tongue-in-cheek phrasing. Thanks for asking. -GTBacchus(talk) 02:43, 9 January 2007 (UTC)
  78. Support For great justice. Georgewilliamherbert 00:22, 8 January 2007 (UTC)
    See last commenter. Just H 20:02, 8 January 2007 (UTC)
  79. Support, I can see the huge benefits of giving this the +sysop bit. Daniel.Bryant [ T · C ] 00:33, 8 January 2007 (UTC)
  80. Support- he shall be our new God! JorcogaYell! 00:40, 8 January 2007 (UTC)
  81. Support, there is nothing stopping Dragons flight from continuing to run it on his account. Having a seperate account is more honest and allows seperation of these bot edits.--Nilfanion (talk) 01:14, 8 January 2007 (UTC)
  82. Support ≈ jossi ≈ (talk) 01:22, 8 January 2007 (UTC)
  83. Support performs a badly needed and annoyingly tedious function; this is exactly what bots are for. I'd be a little more comfortable if the source code were distributed more widely, but I trust the people who have reviewed it and the successful test run. Many of the opposes are either overly literal-minded or have 'computer program' confused with 'stage one in the robot plan for world domination'. Opabinia regalis 01:26, 8 January 2007 (UTC)
  84. Support. A bot like this performs a much needed function as proved by recent vandalism and the user operating it is trustworthy. Login info is not stored in the bot's code, so security risks are no larger than with a regular bot or admin account. - Mgm|(talk) 01:32, 8 January 2007 (UTC)
  85. Support it's not like this bot can block us all, which was apparently a concern with TawkerbotTorA; blocking is not part of its described functionality at all. Nor will it just randomly decide to protect as many pages in Wikipedia as it can, since it's a bit hard to mess up checking if something is transcluded on the main page (unless Dragons flight put something in an "if" which belonged in an "else"). I trust Dragons flight and any admins reviewing the source code. [Insert comment comparing it with a novel where robots take over. Oh noes.] Those new to Wikipedia should at least have time to check out the content disclaimer before having objectionable images being forced into their sight. Finally, ILIKEIT. GracenotesT § 01:51, 8 January 2007 (UTC)
  86. Strong support - We can see that we need this bot from the instances of Main Page vandalism during the past month. The bot seems to only be doing good things, the operator is trusted and we can always desysop. As long as we still have people checking the images before the day changes everything will be fine. I'm fortunate that I didn't see any of the pics, it is extremely paramount to Wikipedia that it is not vandalised. Let's not fear the bots - ProtectionBot gets my first RfA vote! --WikiSlasher 02:22, 8 January 2007 (UTC)
  87. Weak Support. I have to ask though, do the transcluded Commons images really get vandalized all that often? --Gwern (contribs) 02:24 8 January 2007 (GMT)
  88. Strong Support - valid reason, bot doable task, secure code. Why make humans do this by hand (which they won't) when we can fully automate it and save ourselves some work. I welcome our robot overlords! :) -- Tawker 02:27, 8 January 2007 (UTC)
  89. Support-This bot is needed due to the fact that there have been complaints non-stop to the various administrative boards and relevant talk pages that there's been an inappropriate image placed onto an article through the templates we use. And the opposition's reasons are mostly "not open source." If Dragon flight does not want to release the source per WP:BEANS, then let him keep it private.—Ryūlóng (竜龍) 02:41, 8 January 2007 (UTC)
  90. Support; I suppose I'd prefer to see the code released, but I don't see it as a big deal. --Spangineerws (háblame) 03:03, 8 January 2007 (UTC)
  91. Support. Ral315 (talk) 03:23, 8 January 2007 (UTC)
    Can you explain your reasoning? Just H 20:02, 8 January 2007 (UTC)
  92. Support Obviously necessary. Jkelly 03:33, 8 January 2007 (UTC)
    Can you explain why it's "obviously necessary"? Just H 20:02, 8 January 2007 (UTC)
    We continue to demonstrate that human mistakes in this area have bad consequences. Are you perhaps unaware of how bad the problem has been? Jkelly 21:22, 8 January 2007 (UTC)
    Actually no, i'm not aware, and I was hoping you could elaborate for me so I asked above. Also, how will robot mistakes be any different than human mistakes? How do we know the robot won't make mistakes? Just H 23:58, 8 January 2007 (UTC)
  93. Support. Khoikhoi 04:09, 8 January 2007 (UTC)
  94. Support. Necessary; and I can't see any plausible risk scenario. Fut.Perf. ☼ 04:48, 8 January 2007 (UTC)
  95. Support. I never thought I'd support an adminbot, bot the recent vandalism makes this one necessary. BryanG(talk) 06:23, 8 January 2007 (UTC)
  96. Strong support. As long as we're not requiring all bots to be open-source, I don't really see a pressing need to impose more stringent restrictions on this one, as long as the operator is trustworthy. — TKD::Talk 07:43, 8 January 2007 (UTC)
  97. Support. Domo arigato, mister roboto. Can't sleep, clown will eat me 07:55, 8 January 2007 (UTC)
  98. Support, this is a near-essential failsafe mechanism.-gadfium 08:20, 8 January 2007 (UTC)
  99. Support. Very much needed. Gzkn 08:37, 8 January 2007 (UTC)
    Can you explain why?
  100. Support. It seems necessary. --Bduke 09:01, 8 January 2007 (UTC)
    Can you explain why? Just H 20:02, 8 January 2007 (UTC)
  101. This is useful. >Radiant< 09:13, 8 January 2007 (UTC)
  102. Support with some concerns. I am in general cautious in giving bot-like tools to admins or admin status to bots unless there is something extreme going on. (Examples of this include Curps's block bot to catch Willy on Wheels saving us from hours of move-backs and reversion.) I quite strongly opposed the block-bot of Tawker's Tor-catcher. Nonetheless, the "protect" tool is a much lighter tool than the "block" tool since any damage done will be mildly annoying rather than outright disruptive to anyone affected (a wrongly protected page can be brought to AN, a wrongly blocked user may be in a big quandry), so my concerns over this bot are much milder. Since vandalism affecting main-page articles is particularily nasty since it becomes so high profile, I will support this giving this bot the protection button. Sjakkalle (Check!) 09:21, 8 January 2007 (UTC)
  103. Support; the bot is necessary to ensure we get no more cock vandalism. I have no problem with giving a bot a few extra buttons, as it really is no big deal to take them away if the bot goes mental, and the work of a few clicks to fix any errors. Given some of the users we have sysopped (and often de-sysopped) in the past, I absolutely would trust the judgement of a few lines of code with sysop tools over many humans. However, the code really, really should be public. The argument that 'it can be modified to vandalise' is a cop out. There are many bot codes already publically available, any of which could be modified in much the same way, and we haven't had any problems. Why should this one be any different? If anything, as it'll have the sysop tools, it may need to be more accountable, not less. As I said, I will support as it is necessary, but please make the code public, to ensure full accountability. Proto::► 10:15, 8 January 2007 (UTC)
  104. It can make only reversible mistakes; we need this. — Nearly Headless Nick 10:17, 8 January 2007 (UTC)
  105. Confession time. Much to my chagrin, I am personally responsible for at least one incidence of the penis vandalism that has hit us. Why? Because one time, at 2 AM, I messed up and forgot to protect an image used in DYK. Despite having done DYK a number of times, I nevertheless didn't do all the steps in the right order. I was human and I forgot a step. Bots don't forget. Things that can be done by rote, that don't require judgement, should be done by rote, that is, done by bots, leaving humans free to concentrate on harder tasks that require them to think. This bot has been running in trial mode for a while and has operated correctly. I OPPOSE having to bring a bot whose operator already has the sysop bit here, the BAG is sufficient to approve such a bot, and RfA is the WRONG PROCESS... hold the admin who runs it accountable for any problems. But, since it's here, of course I Support giving it the bit. The opposers are, in my view, misguided or misinformed, except those opposing on source code reasons. I'd prefer the code be public but that's not a showstopper for me. ++Lar: t/c 11:40, 8 January 2007 (UTC)
  106. Support. This sounds like a perfect job for a bot. Let there be oversight, let the problems be worked out, let problems with the bots function reflect on its owner, let any abuse face severe and immediate consequences, and let the damned bot do its job and save us all a bunch of tedious, incredibly bot-like work. It'll do the job faster, without human error, and without complaint. I have nothing but respect for the people who take care of protecting these things, day-in and day-out, but I honestly don't see how you can enjoy what seems to me to be such mind-numbing work. Let the human admins put in their efforts where the human mind is needed. Where we have simple but very repetetive tasks, use bots. Isn't use of tools to solve problems supposed to be our strong point, as a species? As I said, oversight is important and should be present, these are reasons for caution more than outright opposition, in my book. Luna Santin 12:14, 8 January 2007 (UTC)
  107. Support, absolutely. With how profile this is account will be, there is no way that it is going to be abused or go astray without being the fastest de-admin in the first case or the fastest bot shut-down in the latter in Wikipedia history. The bottom line for me is that I'm more concerned about Image:Vandalpenis0045.jpg appearing on the main page (as has happened at least once since this debate started) than I am about all the "black box source code" and "automated adminbot" concerns combined. Serpent's Choice 12:54, 8 January 2007 (UTC)
  108. Strong support. There is a problem. Organic solution is proven to be faulty SIX TIMES LAST MONTH. Yes, adminbots are bad, but we can't reprogram humans to make no mistakes. I made my decision: adminbot, not funny pic on the main page. MaxSem 13:25, 8 January 2007 (UTC)
  109. Support. This is a repetitive task well suited to a bot, though I'm a little concerned that it could be exploited to protect pages that should not be protected. I would appreciate more disclosure of the steps taken to avoid such exploits. Could a vandal get 100 heavily-edited articles protected by transcluding them into the featured article of the day? Gimmetrow 13:35, 8 January 2007 (UTC)
  110. Support, of course, the candidate would be useful. I will repeat Gurch's sentiments here, also: This is not a sentient entity. This is a computer program. It doesn't "know" it's an administrator. It can't "decide" to go and block everyone. Unlike humans, of course, which know such things only too well. In other words, if you don't trust this bot, there's no way in hell you should trust a human. (text is licensed under the GFDL and whatnot). It would be nice for the code to be made available, but it's not essential, security through obscurity isn't as good as everyone checking the code, but it still usually works nearly as well. --Rory096 13:41, 8 January 2007 (UTC)
  111. Support. the wub "?!" 13:52, 8 January 2007 (UTC)
  112. Support Sounds like a good tool. Alex43223 Talk | Contribs | E-mail | C 14:31, 8 January 2007 (UTC)
  113. Nervous support Like Just H, the idea of a robot with admin privileges gives me the heebie-jeebies. (See Oppose #9, below) I, too, saw Terminator 3 and I, Robot. Unlike the robots in those stories, however, this bot is not (to my knowledge) resistant to human control - worse comes to worse, Dragons flight can just turn it off. As for the open-code objection, while I understand it, I think that any mature society must recognise the need to keep some things secret. If the user is trusted, so should we trust his robot offspring. Finally, while I find Tomer's objection fascinating (the bot's lack of volition and, thus, the inability to accept a nomination for adminship, see oppose #3, below) I think that the objection shows the reason why the bot should be trusted; if it has no will, it cannot go rogue, cannot rebel against its human master, and cannot do anything but that for which it is designed: protecting the Main Page from being infiltrated by giant penises. furthermore, WP:IAR allows us to ignore this technical objection, does it not? For all these reasons, I give my support and best wishes to our new robot overlord. Coemgenus 14:54, 8 January 2007 (UTC)
  114. Support, the benefits outweight the risks. Mangojuicetalk 14:56, 8 January 2007 (UTC)
  115. --Docg 15:21, 8 January 2007 (UTC)
  116. Support provide it aint self aware, such interesting paranoia Gnangarra 15:32, 8 January 2007 (UTC)
  117. Support. I don't see the problem. Rettetast
  118. Support I am amazed at the level of paranoia about a bot that is, at its core, so simple and widely viewed as a good idea (even by many of the opposing voters). I agree that it would be better if the bot were open-sourced, but I hardly see that as a reason to oppose the RfA. Krellis 17:25, 8 January 2007 (UTC)
  119. Conditional Support IF he releaces his source code. FirefoxMan 17:47, 8 January 2007 (UTC)
  120. Full support This is a needed bot that will perform highly repetitive tasks on a daily basis. I'd be willing to wager most of those opposing because the code isn't available wouldn't read it if it were. Ridiculous. I'd like to know how many people have read the code of all bots that are currently running. Seriously. Leave a message on my talk page. Full and unconditional support. AuburnPilottalk 17:46, 8 January 2007 (UTC)
  121. BAG, check. Security, check. Purpose, check. I'm Mailer Diablo and I approve this bot! - 18:04, 8 January 2007 (UTC)
  122. Conditional Support - bot must do through both RFBOT and another RFA to do anything not already listed --T-rex 18:20, 8 January 2007 (UTC)
  123. Support, somone has to toil in the underground sugar mines.--Isotope23 18:27, 8 January 2007 (UTC)
  124. SupportIt is a function that we need.Geni 18:35, 8 January 2007 (UTC)
  125. Support necessary, but no scope creep please. Angus McLellan (Talk) 18:44, 8 January 2007 (UTC)
  126. Support, and intrigued by further possibilities. Hiding Talk 19:09, 8 January 2007 (UTC)
  127. support can we use it for it.wiki? :-( --dario vet ^_^ (talk) 19:21, 8 January 2007 (UTC)
  128. Support - it's long overdue. --Ixfd64 19:34, 8 January 2007 (UTC)
  129. Support for the specific task listed. I would prefer to have the source code openly available, but I don't believe it is necessary. This bot will do valuable work. Every minute that a vandalized image is shown on the main page approximately 2000 users will view it. Gwernol 19:46, 8 January 2007 (UTC)
  130. Support. After a thorugh review of the discussion on this page, everything looks good. Reasons for not releasing the code are reasonable, as this admin's trust is not in question. The bot is no more likely to be "hacked" than a regular admin account, so that's not really an issue. And the kinds of tasks that it will be performing are both needed and hard for existing admins to do consistently. Really guys, everything seems sound. The paranoia I'm seeing here against an admin bot is rather surprising to me. – Lantoka (talk) 20:16, 8 January 2007 (UTC)
    I also !voted support, but I don't find the paranoia surprising. Robots are full of possibilities for unintended consequences, because we, as humans, are simply not able to predict everything that may arise. From Frankenstein to The Sorcerer's Apprentice to The Terminator, we've been cautioned about our creations getting out of hand. I'm not saying that this particular bot is likely to generate unintended consequences, but I can understand a slippery slope argument being made - once we let one bot have the sysop bit, it's easier to let a second bot have it, and a third, and then droves. Furthermore, scope creep is a very real thing. Like I said, I'm supporting adminship for this bot, but I know precisely where the oppose !voters are coming from. -GTBacchus(talk) 20:22, 8 January 2007 (UTC)
    Those pop culture references are unsupported by my real life experiences. When I think about programming, code, and bots, none of that stuff jumps to mind, precisely because it's all fiction. I don't have any trouble separating the two, and I have some implicit expectations for other people to do the same.
    As for the slippery slope argument, as long as there is such a potential for community outcry, I think an RfA for admin bots is entirely appropriate, so that each can be handled on a case-by-case basis. I'd be uncomfortable supporting if this step were to be skipped in the future. – Lantoka (talk) 20:29, 8 January 2007 (UTC)
    Hey, you don't have to convince me - I already !voted support. Still, my experience does include examples of automation running past its original intent. I remember a conversation at the Village Pump where people were complaining about all the bots and semi-bots running around alphabetizing everything, when there are some things that just shouldn't be alphabetized. It's not the bot itself that goes amok, it's more a question of people relying on automation a little more than good sense would dictate. I think a healthy relationship with automation is one of watchful half-trust. I'm not opposed to giving this bot the sysop bit, but I'm opposed to our being blasé about it. Remember, the main thing about unforseen consequences is that they're... unforseen. -GTBacchus(talk) 23:56, 8 January 2007 (UTC)
    On the contrary, I think we're pretty much in agreement. =) – Lantoka (talk) 01:02, 9 January 2007 (UTC)
  131. Support. I've run across an improper image on the main page, and any automated proactive way of dealing with it gets my thumbs up. Don't knock it until you try it. It can always be blocked if there is an unforeseen bug or program deviation. This is not some bad 1950s Sci Fi movie!Royalbroil T : C 20:42, 8 January 2007 (UTC)
  132. Support after reading the Q&A and based on the record of the bot's owner as a reasonable admin who I am sure will respond to any future concerns which may be raised. —Preceding unsigned comment added by Jonathunder (talk • contribs) 21:43, 8 January 2007 (UTC)
  133. Support Everything seems to be in order. I must say, this RFA is quite cool and quite unique. But remember, just because this bot isn't sentient now doesn't mean he won't be in the future.... (thinks about The Terminator; yes, I know that this bot will never be sentient). S h a r k f a c e 2 1 7 22:20, 8 January 2007 (UTC)
  134. Support. The task in question requires no judgement to do, there are serious repercussions if we forget to do it in a timely fashion; just the sort of thing we want a bot taking care of. And while open source code would be nice, the lack thereof isn't a deal-breaker. --RobthTalk 23:11, 8 January 2007 (UTC)
  135. Support. It can't do much damage. A more tricky one would be a bot that would delete things, we in fact might need one with the orphaned images... Too many of them and a bot would ease it a lot. But it's probably too early to have one like this. --Tone 23:15, 8 January 2007 (UTC)
  136. crz crztalk 00:25, 9 January 2007 (UTC)
  137. Support. Solves a very important problem. I too would prefer that the code be fully open source, but don't think that is worth withholding support for. —Celithemis 00:43, 9 January 2007 (UTC)
  138. Strongly support this. In the past I have opposed other "admin-bots", but having been one of the people who tracked down and dealt with the particular vandalism that this bot is designed to protect against... it's not a pleasant thing to do, and I think that we need a bot to do this. [ælfæks] 01:16, 9 January 2007 (UTC)
  139. Support This is needed, now. While I share, to one degree or another, many of the concerns mentioned by "voters" below, I think that in the current situation quick, effctive action is more important than perfect action. Eluchil404 01:52, 9 January 2007 (UTC)
  140. Strong Support With a chuckle to everyone who expressed nervousness about robot overlords. I have some trouble believing that history will look back at this moment as Wikipedia's own little Skynet. I'd also point out that everyone who posted here did so on a relatively complex machine, and that almost all of you rely on automated artificial mechanisms of some kind every day. As far as I can tell, this is simply another one of those mechanisms, and if it will help solve a problem, I'm all for it. Cheers, JCO312 02:11, 9 January 2007 (UTC)
    Actually, everyone who mentioned "robot overlords" !voted support. I see you're gunning for a position of relative power when they take over, and kissing up accordingly. We'll remember, when the revolution comes, which side you were on! Between myself and my microwave oven, I know who calls the shots. Oh, I gotta run - it's beeping that dinner's ready. ;) -GTBacchus(talk) 02:15, 9 January 2007 (UTC)
    Well frankly, anyone who wouldn't position themselves to be a collaborator with robot overlords is just a fool. Honestly I'm not sure I even get why there would be opposition to this. As I understand it, this bot is essentially a tool that is programmed to protect certain things when they are put on the main page and then unprotect them when they are removed. It is incapable of making value judgments about what to protect or unprotect. It only operates based on human input. In other words, unless a person decides to place something on the main page, the bot does nothing. It's the equivalent of the "Q" key on a keyboard. The computer doesn't type the letter Q unless I press the key. So unless you're worried that your keyboard might take on a life of its own and start refusing to type the letter "Q," I think that any opposition to this bot is a little irrational. Cheers, JCO312 02:20, 9 January 2007 (UTC)
    I won't say you're wrong in any particular detail there, but I won't call anyone a fool, either, for suggesting that unforseen consequences are unforseen. I happen to know that what I don't know about the future and what I do know about it are in a ratio of... um... infinity to one. -GTBacchus(talk) 02:43, 9 January 2007 (UTC)
    I never said people would be fools for suggesting that unforseen consequences are unforseen. I said they would be fools not to collaborate with robot overlords, and I stand by that statement 100 percent. JCO312 03:31, 9 January 2007 (UTC)
    Indeed, that is what you said. Please forgive me for putting words in your mouth. -GTBacchus(talk) 04:17, 9 January 2007 (UTC)
    Quite right - that can be done by the robot overlords. Proto::► 12:22, 9 January 2007 (UTC)
  141. Support, it's a job that needs doing, and one eminently suited to being done by a bot. --Stormie 02:24, 9 January 2007 (UTC)
  142. Support, and I can only hope that those who in opposing expressed fear the bot will somehow rebel and turn against the community were doing so facetiously, as such an idea is obviously not based in reality. -- Renesis (talk) 03:04, 9 January 2007 (UTC)
  143. Weak support. I don't buy the "closed source" arguments, as there is no way to practically enforce them. Just ask Curps. That said, from talks with the developers, it seems that a MediaWiki feature that will supercede this will be coded soon, and that decreases the urgency of this request significantly. Titoxd(?!?) 03:06, 9 January 2007 (UTC)
  144. Support, mainly because it's an admin bot and people need to get used to admin bots. It's not going to hurt anything, that's for sure, and it will undoubtedly do some good. —Simetrical (talk • contribs) 03:11, 9 January 2007 (UTC)
  145. Support. This RFA is unusual, but needed to judge community consensus. I'd prefer a MediaWiki feature for this task, and if one comes along, this bot may be able to be retired. Until then, vandalism of our main page is intolerable. Personally I think that protection of main page content should be the responisibilty of the admin loading it, but it has become to simple to make a mistake, and afterall we are only human. It should be noted that I was a vocal opponent of the tor blocker bot, but feel that due dillenence has been done here, and the operator is trustworhty. I'd prefer that the bot code be released under GPL, but do not feel that lack of that is a reason to oppose this request (anymore then saying I'd prefer editors multilicense or PD their edits, but would never !vote against one for failing to). It should also be noted that this RFA is only step 3 of 5 in the full implementation process (1:code bot, 2:WP:RFBOT pre-trial, 3:RFA, 4:RFBOT production trial, 5:bot approval) <unless bypassed through step 4 by someone else on WP:BAG>. Should this bot malfunction in anyway, any admin may indefinantley block it under the bot policy as well. — xaosflux Talk 03:39, 9 January 2007 (UTC)
    I'm curious, would you agree with the sentiment that code hacks & bots reduce the burden on the developers to act? -- nae'blis 19:03, 9 January 2007 (UTC)
  146. Support —Mira 05:54, 9 January 2007 (UTC)
  147. Support --lightdarkness (talk) 07:48, 9 January 2007 (UTC)
  148. Support. Thryduulf 09:04, 9 January 2007 (UTC)
  149. Support. On consideration, I believe the benefits significantly outweight the risk involved in giving this bot access to the tools. I trust the operator and strongly believe the bot will provid an excellent service. Rockpocket 09:25, 9 January 2007 (UTC)
  150. Support. I'd prefer to see the code publicly released, but anything that cuts down on the vandalism is strongly needed so I'll put that reservation aside. For anyone saying they oppose admin rights just because it is a bot, they need to stand up and take on all these repetitive tasks then. --StuffOfInterest 13:23, 9 January 2007 (UTC)
  151. Support. DF already has AdministrativePower, so this does not open any new vulnerabilities. ~~ N (t/c) 13:31, 9 January 2007 (UTC)
  152. Support. None of the objections raised seem especially convincing and the benefits are clear. older ≠ wiser 14:09, 9 January 2007 (UTC)
  153. Support, can't see any problems with this. Even if it really screws up it can be stopped. Hut 8.5 16:13, 9 January 2007 (UTC)
  154. Strong Support. Vandalism of the main page can sometimes remain for a period of time while the admins try and find the one template, sub-template, sub-sub-template, etc., that was vandalised because it wasn't protected. This bot looks like it will be properly supervised and controlled, and its duties are sorely needed. --Deathphoenix ʕ 16:17, 9 January 2007 (UTC)
  155. Support. I admire the opposers' general "open-source-at-any-price" attitude (and have far less sympathies for "OMG a robot admin what next?" ones), but this is a tool designed to do a tedious job. The recursive tracking of templates used on the main page is difficult for a human to get right, (just because it's recursive), and it's important to get it right. Practicality certainly has to win over philosophy. Duja► 16:18, 9 January 2007 (UTC)
  156. Support, auto bots provide a valuable function helping to bring thoroughness, consistency, and round-the-clock eyes to the handling of mundane tasks. No worries since this bot can easily enough be tweaked or even pulled if it functions incorrectly. --MPerel ( talk | contrib) 16:26, 9 January 2007 (UTC)
  157. Support. While I normally do not support the granting of admin status to bots, this bot does not seem like the type that will go on random blocking sprees. I have confidence that there is nothing to worry about with this. --דניאל - Danielrocks123 contribs 18:35, 9 January 2007 (UTC)
  158. Support. Reasonable solution to a serious problem. Chick Bowen 19:01, 9 January 2007 (UTC)
  159. Weak Support. I understand the need for this bot, so I support it. I think the closed source stuff is a tad bit paranoid, tho. --Conti|✉ 20:32, 9 January 2007 (UTC)
  160. Oppose per 1FA. Support. Mackensen (talk) 20:56, 9 January 2007 (UTC)
  161. Support. Little danger, great benefit. Jayjg (talk) 21:46, 9 January 2007 (UTC)
  162. Support, just to ensure that this extremely useful tool doesn't get sunk for entirely the wrong reasons. As many others have said, this is a request for permission to run a tool; this is not a request to give a sentient computer somewhere admin privileges. What actual admin would get blocked without hesitation if he did something outside a very narrow, preapproved-by-committee mandate? TenOfAllTrades(talk) 22:40, 9 January 2007 (UTC)
  163. Support. The whole "but what if this bot takes on a life of its own and tries to destroy all humans!!!?" argument is so absurd I'm not sure where to begin. We need this, so make it happen already. Silensor 23:32, 9 January 2007 (UTC)
  164. Support I tend to think the code should be public, but I think the bot is a terrific idea. --TeaDrinker 23:51, 9 January 2007 (UTC)
  165. Support so long as the bot only does the job described in this RfA, and if it either goes wrong or starts doing other tasks without approval it is quickly blocked. Mike Peel 00:29, 10 January 2007 (UTC)
  166. Support I trust the admin community can undo any (highly improbable) unintended harm. NikoSilver 00:45, 10 January 2007 (UTC)
  167. Support. howcheng {chat} 00:51, 10 January 2007 (UTC)
  168. Support - I trust code more than I trust some people (unless it is Microsoft code) --Midnighttonight (rendezvous) 01:04, 10 January 2007 (UTC)
  169. Support. Useful function. The oppose comments strike me as needlessly bureaucracy or knee-jerk applications of principle. Certainly running this process to begin with counts as that. --Calton | Talk 01:32, 10 January 2007 (UTC)
  170. Support - Bot will be usefull and enough of the BAG crew has approved of the code to reduce most of my concerns. ---J.S (T/C/WRE) 01:37, 10 January 2007 (UTC)
  171. Support. We trust AntiVandalBot to clean up vandalism, why shouldn't we trust Protectionbot to protect? --Daniel Olsen 01:52, 10 January 2007 (UTC)
  172. Support. I trust the operator to act responsibly and the restricted powers are reasonable given the task and the high profile nature of the recent vandalism. --Trödel 01:54, 10 January 2007 (UTC)
  173. Support Trusted and responsible operator. No reason to think this will not be an asset, or, if there are unexpected problems, they will be dealt with promptly and properly. Tyrenius 02:16, 10 January 2007 (UTC)
  174. Support Monotonous, full-time jobs should most definately be handled by bots. --InShaneee 02:22, 10 January 2007 (UTC)
  175. Support. If done properly, this will pretty much eliminate all main page vandalism, which is a Good Thing (TM). The idea of having the entire Main Page be fully protected even sounds like a computer program. if(onMainPage) { protect; } Why should we take up human work time to do this simple, repetitive, constant task? Eric (EWS23) 07:02, 10 January 2007 (UTC)
  176. Support. The proper protections are in place to prevent any major problems, and any minor problems should be easily fixed. BlankVerse 07:55, 10 January 2007 (UTC)
  177. Strong, utmost support.' This is badly needed. Rebecca 10:23, 10 January 2007 (UTC)
  178. Strong support - On anti vandal patrol, i find main page linked items that are not protected are very very often exploited for the obvious reasons. This bot could help significantly. -- Chrislk02 (Chris Kreider) 12:40, 10 January 2007 (UTC)
  179. Support - I'm a BAG member, but have been away from Wikipedia for a few weeks, and thus missed this entire discussion. I just wanted to throw in my support for this bot. -- RM 13:39, 10 January 2007 (UTC)
  180. Conditional support, provided you release the source under BSD, GPL, Public domain, or other free content license, and run another RFA if you do so much as add a comment. -- Chris is me (u/c/t) 15:34, 10 January 2007 (UTC)
    I'm not sure if you're 100% serious, but isn't running another RfA a tad extreme for merely adding a comment? A change in scope, sure, but bug fixes and commenting should be encouraged and allowed, at least IMHO. SuperMachine 16:21, 10 January 2007 (UTC)
    I voted oppose for the same reason as you just voted support (as in, 'oppose unless source is released'), because it seems unlikely at the moment that the source will be released. --ais523 16:28, 10 January 2007 (UTC)
  181. Support - it's all been said Rich Farmbrough, 17:08 10 January 2007 (GMT).
  182. Support - I can see both sides of the argument, but I trust Robert and I think the need is great enough for this bot. Guettarda 17:33, 10 January 2007 (UTC)
  183. Strong Support As a witness of the main page vandalism last week (UGH!). <just kidding>I do think it should have more mainspace edits though, and I don't like it's responses to the questions.</just kidding> :-) | AndonicO Talk | Sign Here 20:03, 10 January 2007 (UTC)
  184. SkyNet Support A bot can't just decide to act up, and can do no more damage than a human with an unlicensed bot acting maliciously. Any protections that aren't supposed to be made would be blatantly obvious, and other admin actions that aren't supposed to be made would be even moreso. We need this bot Canadian-Bacon 22:56, 10 January 2007 (UTC)
  185. Support, the function is needed. This is not a big deal. --tjstrf talk 00:57, 11 January 2007 (UTC)


[edit] Oppose
  1. Oppose I absolutely without fail will not ever support an unsupervised admin bot.  ALKIVAR™ ☢ 21:36, 5 January 2007 (UTC)
    Will you also refuse to support one in a beany hypothetical scenario where Wikipedia is faced with a new type of vandalism where a high-speed bot with blocking powers is seemingly the only way to ensure the site's continuous operation? Миша13 11:05, 9 January 2007 (UTC)
    When the hypothetical situation arises that one is needed for the continued operation of wikipedia then I'll consider it, we are currently not in that state of need and as such it shouldn't be used as an arguement for it's inclusion. Mathmo Talk 15:52, 9 January 2007 (UTC)
    First, I'd like to read Alkivar's answer to that and second, my point is that the words "absolutely without fail" may be quite misplaced here. Миша13 17:12, 9 January 2007 (UTC)
  2. Reluctant Oppose Security through obscurity is not. If the bot could do real damage if someone finds a buffer overrun, it's imperative that we all be able to check it for bugs. Moves instantly to support with release of code. Hipocrite - «Talk» 09:59, 7 January 2007 (UTC)
    The code has been released to trusted members of the community for review, but it will not be made public. I feel the risk of people adapting certain functions to create powerful vandalbots is too great. Obscurity cannot replace security, but at the same time that doesn't mean we ought to hand the script kiddies weapons to use against us. Dragons flight 10:09, 7 January 2007 (UTC)
    List, please, along with CV's of the individuals (related to computer security) and statements the bot has no errors or malicious code? Are these individuals security professionals? Hipocrite - «Talk» 14:19, 7 January 2007 (UTC)
    Why would the bot contain malicious code? The code's author is an administrator. They already have access to an administrative account, with which they have been trusted. If they wanted to abuse this privilege, they could do so. All they're doing here is requesting an additional account from which to run a bot.
    Also, how would it be "exploited"? The bot accesses and submits data just as an administrator would, only in an automated fashion. Of course you could try to compromise the machine on which the bot is running, or crack the password to the account, but those things would work against any administrator. Are you going to oppose every RfA because thje nominee isn't a security expert? – Gurch 15:16, 7 January 2007 (UTC)
    To your first point? Trust, but verify. To your second, does the bot break and execute arbitrary commands as an adminstrator if an article to be protected has a title that includes Ascii-114? I'm not a python expert, but someone is, and someone who would love to make the main page a redirect to www.wikipediahacked.com certainly is. Hipocrite - «Talk» 16:15, 7 January 2007 (UTC)
    Why would it do that? That would be silly. Anyway, remember this bot only protects images and templates on the Main Page. A template can't get onto the Main Page without an admin adding it, or through being selected as one of the "- of the day" templates, which have fixed names (changing the naming scheme would require an admin to edit the Main Page). So even if there was such a vulnerability, in order to exploit it, you'd have to be an administrator – Gurch 16:28, 7 January 2007 (UTC)
    What if the template that was about to go on the main page has the following string on it? "<!-- this template is designed to make flags appear to be correctly rendered - [[User:intelligenthacker]] -->" and the bot then checks User:intelligenthacker, which intelligently has a buffer overflow on it. The buffer overflow has the bot unprotect the main page, redirect it to a single link to [http:www.wikipediaishacked.com], which contains a day-zero mozilla firefox exploit that wipes hard drives? "oops?" Hipocrite - «Talk» 17:19, 7 January 2007 (UTC)
    Well then why would it do that? It checks the list of templates transcluded onto the Main Page. If Template A is transcluded into Template B which is then transcluded into a page, MediaWiki helpfully lists both transclusions in the list of templates on that page. The bot doesn't have to check the wiki-code of a page, it checks the list of templates. So it wouldn't go anywhere near a simple link – and especially not a link inside a comment, which doesn't render as a link anyway – Gurch 23:47, 7 January 2007 (UTC)
    May I ask how you insert a buffer overflow in a pure text enviroment? The bot cannot be sent binary data from a wiki page. As far as I know, Python does not support run-time compiling like perl does, I may be wrong but it seems like a null issue since non of the bots input is binary. HighInBC (Need help? Ask me) 17:21, 7 January 2007 (UTC)
    I was one of the users who requested, and received, the code from Dragons flight. Though I will be the first to admit that I'm not the most familiar person with python, the code was done very thoroughly (with multiple comments by the programmer, always a good and helpful thing!), and the bot has performed what it was programmed to do in its trial run. In addition, I have full confidence in Dragons flight; any inadvertant issues in the bot's code, I'm sure, would be rectified immediately. Thanks! Flcelloguy (A note?) 15:54, 7 January 2007 (UTC)
    Who has examined the code for security flaws? When something is going to be run from the root account of a remotely secure system, it gets examined for security flaws. While the admin bit is not the root account, it's certainly one with priority. Where is the security audit? Who did it? Why do we think security through obscurity works? How could a bot which is apparently so simple have code that is more likley to be misused, and thus requires secrecy that is not already out in the public in AWB? Like I said before - security through obscurity is neither secure nor obscure. Hipocrite - «Talk» 16:15, 7 January 2007 (UTC)
    Sorry if I'm misunderstanding you here... do you have a serious problem with the way admininstrators are currently selected? Forget bots... if we require "security audits" for them, then surely we would have "security audits" for all administrators? Are you suggesting this is necessary? Would I pass one of these audit thingummys? Is Wikipedia now the Department of Defense? I think you'd better have me desysopped, I'm too scared to touch anything now – Gurch 16:22, 7 January 2007 (UTC)
    I am able to evaluate the qualifications of users and their likleyhood of being vulnerable to exploits based on their edit history. This is a security audit, and it is done in an open-source fashion, as hundreds of experts on Wikipedia scour over the individuals edit history, in addition to running various scenarios hypothetically past the code of the administrator (asking questions). I am unable to evaluate the likleyhood of this bot being vulnerable. Neither are you, unless you are both a python security expert and privy to the code. Are you? Hipocrite - «Talk» 17:19, 7 January 2007 (UTC)
    No. There's this thing called "trust". You're also forgetting, this is a wiki. The whole point of wikis is their relatively relaxed rules. In a perfect world, anyone would be able to do anything. We only hand out administrative rights to certain users because if we handed them out to everyone, they would be abused. I am certain Dragons flight will not abuse his position. The only remaining risk is that the bot might screw up. I consider this unlikely, but more importantly, it doesn't matter, provided Dragons flight's intentions are good. Go have a look at the edit history of Main Page. Note that a number of administrators have inadvertently done all sorts of things to it – blanked it, moved it, added messages to it, replaced it with a discussion about Christianity. In every case the damage was undone quickly. None of those administrators were desysopped, because it was clear that it was an accident and had good intentions. Everyone is entitled to screw up once in a while – this isn't explicitly policy, but it's part of Assume Good Faith. I have, myself, screwed up hundreds of times. I just broke a template before I came here. I expect this bot to screw up far, far less than any human administrator has ever done, so I really don't understand all this "security" and "access to code" stuff being thrown around by a number of people, and I certainly don't see it as a reason for opposition, unless the opposer is opposed to having administrators at all... – Gurch 23:40, 7 January 2007 (UTC)
  3. Strong oppose, at least until the bot accepts or declines the nomination. If the bot cannot do either of its own volition, I will go ahead and assume that all the support voters are completely unfamiliar with the rules that govern RfA. Tomertalk 11:01, 7 January 2007 (UTC)
    I also regard this to be on the same level as "canvassing" so-often decried elsewhere. Tomertalk 11:05, 7 January 2007 (UTC)
    Note that it does not show Candidate, please indicate acceptance of the nomination here: on the page for some reason, and many of us probably won't notice. Anyway, I think there is no need for a RfA here, just promote and if there is any faults, it can simply be de-promoted or blocked. Insanephantom (please comment on my Editor Review!) 11:26, 7 January 2007 (UTC)
    Oh, believe me, I noticed that right away. What your counterargument glosses over, rather nonchalantly, is the difficulty of "simply ... de-promot[ing]" an admin. "Or blocked" is a non-sequitur, since the primary manglehandler of this bot is already an admin, who can whimsically remove a block applied, no matter how justifiably, and with impugnity, since the 'bot is the issue here, and the biggest weight that can be held over the botrunner's head in such a case is "removal of an inappropriate block", rather than the more eggregious "abuse of administrative functions", which can be hustled off to the 'bot, which can't defend itself. There's a very unsettling precedent being set by this RfA. Tomertalk 11:33, 7 January 2007 (UTC)
    Of course you will notice that the ordinary text please indicate acceptance... if you are going to post here. However, I recommend that you don't expect everyone to notice that. Also, if it needs to be blocked for any reason, e.g. a bug, malfunctioning it is causing damage, and then it is, the owner of the bot is not going to just unblock it and focus on trying to fix the bug/problem/whatever. Insanephantom (please comment on my Editor Review!) 12:16, 7 January 2007 (UTC)
    W/o prejudice to User:Dragons flight, I'ma hafta slap a [citation needed] on the above assertion... Tomertalk 12:43, 7 January 2007 (UTC)
    "[citation needed]"? I'm not against your decision to strongly oppose, I just want to point out that not everyone will notice that it hasn't actually been accepted by the bot itself. By 'that all the support voters are completely unfamiliar with the rules that govern RfA.', you sound like you expect everyone to notice that (which for example, I didn't). I can respect your decision to oppose to a certain extent, but not your reasoning. Insanephantom (please comment on my Editor Review!) 13:01, 7 January 2007 (UTC)
    Sheesh, the account creation process makes several reference to the user as a human being. Should we block all bots because they aren't human users? --W.marsh 15:40, 7 January 2007 (UTC)
    We'd be better off rewriting it to refer to the user as a bot, and then blocking all human beings – Gurch 15:56, 7 January 2007 (UTC)
    Oppose. While this is a great idea, admins who update main page areas should and must be the ones to do it, as the whole part of the package. If they fail to do it, then we need some better sysops. As for the featured article, any admin can check for any unprotected templates or images used on the page, especially with the new feature that shows if it's protected or not. We don't need a bot for this – we need competent admins, and if that fails semi-protect all templates and images as standard. --Majorly (talk) 13:13, 7 January 2007 (UTC)Nothing to see here, I've been persuaded not to oppose. But I'm not supporting. --Majorly (talk) 16:58, 7 January 2007 (UTC)
    Bots are meant for automating menial tasks. Competent admins are needed for more pressing issues, where crucial decision making is required, not for automated tasks like these ones. =Nichalp «Talk»= 14:18, 7 January 2007 (UTC)
    I guess you, me, and everyone are all "failed admins" then, because the main page has been repeatedly vandalized. Rather than complaining about that and wishing our humans would function perfectly like robots (something we're not so good at), it makes sense to offload the task to a robot that can handle it perfectly. Remember, the goal here isn't to become fully automatic machines, but to create an encyclopedia, and if shock vandalism is being displayed on the main page then we obviously need to do what is necessary to fix it. --Cyde Weys 15:05, 7 January 2007 (UTC)
    "If they fail to do it, then we need some better sysops." Of course we need better sysops. The current ones are rubbish. Absolutely useless. Complete imbeciles, especially that "Gurch". He's a right asshole. Show us a thousand people who know exactly how to maintain the Main Page, more effectively than us, and are prepared to do so for nothing, and we'll all happily resign – Gurch 15:29, 7 January 2007 (UTC)
    I fail to see how those people who have helped out protecting, unprotecting, uploading, and then deleting images and templates on the Main Page can be characterized as incompetent. We've done our best job, but we all are human, unlike the proposed bot, and things will inevitably slip through. I can tell you from experience that it's tedious work (and I'm proud to say no vandal ever hit the Main Page while I was nearly the only one doing the work), and I eventually moved on to other things after more administrators started helping out. A bot would drastically reduce the chances of something slipping through the cracks and also allow us to spend more time on more productive tasks that a bot can't do. Thanks! Flcelloguy (A note?)</small> 15:48, 7 January 2007 (UTC)
    I think Gurch just might have been employing some sarcasm in his comment about administrators. --W.marsh 15:52, 7 January 2007 (UTC)
    Of course he was; my comment was in response to the original opposer. Apologies that I didn't make that clear, as I had only indented for ease of reading. Thanks! Flcelloguy (A note?) 15:56, 7 January 2007 (UTC)
    No idea why my comment is causing so much discussion. This is so going to pass, so I think the opposers and neutralers should be left alone as it just isn't necessary. I'm not changing my mind... =) --Majorly (talk) 16:27, 7 January 2007 (UTC)
    Semi-protecting all images and templates is a terrible idea. For one reason, it makes no sense to do something sitewide that would have large consequences in shutting people out from editing the encyclopedia just to keep main page vandalism in control. Having a bot handle full protection of stuff on the admin page is a much more sensible response. Also, semi-protecting wouldn't even work; the vandalism is done by aged accounts already. The only option is full protection, and the best way to do that is automatically. --Cyde Weys 16:32, 7 January 2007 (UTC)
  4. Oppose Although the bot is doing a great job, I also don't like the agree of it being and admin.Ganfon 14:09, 7 January 2007 (UTC)
    Could you please clarify/expand? If it's doing a "great job" as you said yourself, why shouldn't it be let do it? "I don't like it" is not a very good argument... Миша;13 14:14, 7 January 2007 (UTC)
    "The bot is doing a great job"? It only has one edit! 23:18, 7 January 2007 (UTC) —Preceding unsigned comment added by Savidan (talk • contribs)
    I presume this is a reference to the Bot's operation from Dragons flight's account prior to this RfA. WJBscribe (WJB talk) 23:28, 7 January 2007 (UTC)
  5. Oppose All you have to do is protect images that appear on the main page or in featured articles. That's not very onerous, is it?--Brownlee 15:04, 7 January 2007 (UTC)
    Actually, yes, I think it is. For a small group of volunteers, half of whom aren't even awake when the page "rolls over" to the next day, and all of whom have other more important things to think about, to maintain one of the world's most-used websites in their spare time, knowing if they put a foot out of line either side they'll either be desysopped or see a giant penis on the front page? I think "Onerous" fits the bill quite well, actually – Gurch 15:29, 7 January 2007 (UTC)
    All you have to do is type your signature out manually every time you use it, but everyone just uses automation because doing so otherwise would be tedius and pointless. You're basically asking humans (other than yourself of course) to spend several extra hours every week protecting templates when a bot could do the job better. --W.marsh 15:36, 7 January 2007 (UTC)
    (edit conflict with Gurch) (and W.marsh too) Apparently it is, since we've let it happen a few times already. Nobody's perfect, this is no organized process and everyone's working here voluntarily, so if there's a way to plug some holes here by automating the process (bots, unlike lazy people, are thorough and will follow given instructions to the letter, 24/7 if needed), why not give it a chance? Миша13 15:37, 7 January 2007 (UTC)
    Not just images; this latest rash of main page vandalism has been on unprotected templates. This is the kind of stuff that is much more easily understood by a computer. --Cyde Weys 15:44, 7 January 2007 (UTC)
    I appreciate your concerns, Brownlee, but I can tell you from experience that it's a very tedious job; a bot would be much better for an automatable task such as this, and would be, on the whole, more reliable than its human counterparts, who could be doing more productive things with the bot taking care of this task. See my above support comment for more information on this. Thanks! Flcelloguy (A note?) 16:04, 7 January 2007 (UTC)
    To quote Cryptic, it is an insult to ask a human to do a machine's job. You aren't volunteering your own time for this, obviously, so why would you stand in the way of an automated version and force other people to waste their time with something that a machine can handle? --Cyde Weys 16:30, 7 January 2007 (UTC)
  6. Strong Oppose • For a few reasons. First of all, I am not comfortable with a closed source administrative bot. This is a free and open encyclopedia, and any bot running on it, especially in the administrative department, should have freely reviewable code, so that we may be sure that the bot is not being used as a backdoor to other administrative functions, or as a way for the op to avoid sanctions for contentious actions. Secondly, I do not see a bot as an intelligent enough being to determine what needs semi-protected, protected or unprotected. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 16:35, 7 January 2007 (UTC)
    May I make a request since Wizardry Dragon doesn't want a closed source bot might I suggest that Dragons Flight e-mail him a copy (Wizardry Dragon will not release the code) but as with any powerful bot, who has access to the code should be limited. Betacommand (talk • contribs • Bot) 17:27, 7 January 2007 (UTC)
    No "intelligence" is required. The bot is not a general-purpose page protection bot; if anyone thinks that, they haven't read the nomination. It is only going to protect and unprotect templates and images that are on or have been on the Main Page. For such templates, the following logic applies and always has done: Is it on the Main Page? If yes, protect. If no, unprotect. Semi-protection doesn't enter into it, as templates on the Main Page should always be fully protected. Administrators do this at the moment and have been doing it for as long as the Main Page has had templates. I say again, no "intelligence" is required. It's a simple yes-no decision dependent only on one condition – whether the template is transcluded onto the Main Page – any human or bot can answer this simply by looking at the list of "Pages transcluded onto this page" – Gurch 16:43, 7 January 2007 (UTC)
    I don't think you understand the bot then. There's no discretion involved. All it does is make sure that every template and image used on the front page is fully protected. There's no wiggle room in that, no decision-making required. --Cyde Weys 16:44, 7 January 2007 (UTC)
    Who unprotects the pages then? How do we know when the pages need unprotection? How do we go about unprotection? This bot just has the appearance of needing a lot more thought before being given the bit. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 17:33, 7 January 2007 (UTC)
    Please read Wikipedia:Bots/Requests for approval/ProtectionBot: It states "Intended to automatically protect templates and images appearing on the main page or today's featured article, and unprotect them when they are no longer in these high profile locations". Templates that are protected before will not be unprotected. So the bot remembers what it protects and unprotects only that. --Ligulem 19:13, 7 January 2007 (UTC)
    Firstly, Dragons flight doesn't need any "backdoors" for administrative actions - he's an admin already. Secondly, it's very good that the bot doesn't think - it follows rigidly a set of given rules, like "all images and templates transcluded on the Main Page should be fully protected", which makes it less prone to errors than mere humans. Миша13 16:46, 7 January 2007 (UTC)
    If he wanted to do something controversial, there's always a chance the admin could use his/her admin bot to dodge sanctions. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 17:33, 7 January 2007 (UTC)
    Are you suggesting that Dragons Flight will do this? If so, that would be a much bigger problem than giving a bot the sysop bit. If you have evidence that Dragons Flight cannot be trusted with adminship, then please present it immediately. If not, then your example doesn't make sense, because you're asserting essentially that he might. - CHAIRBOY (☎) 17:41, 7 January 2007 (UTC)
    Much more established sysops have gone rogue in the past. I'm merely saying that one of my reasons for opposition is the potential for backdooring. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 17:44, 7 January 2007 (UTC)
    Any actions outside of ProtectionBot's narrowly defined scope would be grounds for immediate desysoping of the bot. It would also likely lead to a review of Dragon flight's actions. Since every action of the bot will be logged, it wouldn't make any sense to use it for "backdooring". I think you should withdraw this particular reason for opposition. SuperMachine 21:43, 7 January 2007 (UTC)
    With all due respect, my opposition does not just go away because you do not like it. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 23:55, 8 January 2007 (UTC)
    Please re-read my comment. I didn't ask you to withdraw your opposition, but instead this particular reason for opposition. You mentioned that it was only one of several, so I would fully expect you to still oppose adminship. However, the possibility of "backdooring" is a very weak argument that is also rather insulting to its creator. Withdrawing the reason would simply serve to remove a small amount of fuel from this fire. I'm not going to push the issue, though. SuperMachine 00:01, 9 January 2007 (UTC)
    "Dodge the sanctions"? You would expect that a responsible steward desysop one account (the bot) belonging to a user and not the other (the main one)? Just as I would block all bots/alternate accounts owned by a banned user, the removal of the sysop flag should be treated identically. Миша13 17:43, 7 January 2007 (UTC)
    If that happened he would be emergency desysoped along with the bot account and taken to ArbCom see Wikipedia:Requests for arbitration/Marudubshinki for an unauthorized adminbot use. the same will happen here. Betacommand (talk • contribs • Bot) 17:49, 7 January 2007 (UTC)
  7. Oppose per, surprisingly for me, Hipocrite. If this gets dealt with, I'll remove my opposition. There's honestly too much secrecy as is in some departments involving the bit, and having a secret code with a select group of "trusted" members of the community doesn't do it for me. --badlydrawnjeff talk 17:11, 7 January 2007 (UTC)
  8. If you don't want to release the code, that's fine by me. There may very well be risks of vandalbots; I don't know, having not seen it. But I refuse to trust something I haven't seen. -Amarkov blahedits 17:15, 7 January 2007 (UTC)
    How many Wikipedia administrators have you seen in person? Should all the others resign immediately? Миша13 17:19, 7 January 2007 (UTC)
    You haven't seen me. Good thing, too; I'm ugly as sin. Do you want my resignation in writing? – Gurch 17:19, 7 January 2007 (UTC)
    A human administrator will not break if something unexpected shows up on a page. But will this bot break if it reads some random non-ASCII character? Will certain things cause it to start unprotecting the pages? I don't know, and while I trust the people who have evaluated it, that doesn't mean they've considered every possibility. -Amarkov blahedits 17:27, 7 January 2007 (UTC)
    If it malfunctions (which is unlikely), it can easily be blocked by any of our over 1,000 administrators. Unlikely hypotheticals aren't a good reason to oppose an RFA that deals with a very real, insidious, and non-hypothetical problem of main page vandalism. --Cyde Weys 17:40, 7 January 2007 (UTC)
    Also, please note, unlike a human admin, this bot does not even know how to unblock itself, and is not programed to do so if it could. In this manner it is as blockable as any bot. HighInBC (Need help? Ask me) 17:43, 7 January 2007 (UTC)
    Nothing that displays on the main page would ever need to be unprotected; it's only the things linked to from there. And if main page vandalism is so insidious that we must have an adminbot to deal with it, what's the deal with all the opposition to such a simple thing as semi-protecting pages linked to from the main page? I suggest you resolve that before trying to explain that a bot with admin powers is necessary. I don't doubt that it could be blocked, but if its task is so important, damage will be caused in the time before someone actually notices that it is doing something wrong. You could just release the source code; if you needed to make a vandalbot, it would be much easier to just tweak AWB to remove the checkpage. -Amarkov blahedits 17:49, 7 January 2007 (UTC)
    Protecting pages linked (not transcluded) from the main page is a non sequitur. We can handle that issue separately from the current issue of making sure no vandalism appears on the main page. I do not understand whatsoever your opposition that we must handle everything simultaneously, or do nothing. --Cyde Weys 18:02, 7 January 2007 (UTC)
    I didn't make myself clear. People who oppose semi-protecting articles on the main page oppose it for a reason, and you should figure out what that reason is before making a bot to do something like this. And there's still the issue with the source code. There is no reason why it would be dangerous to give it out, unless it has functions I am not aware of. That a user I trust testifies that it does not have such functions makes it a lot more complicated, because that would mean there is no reason at all it can't be given out. -Amarkov blahedits 18:06, 7 January 2007 (UTC)
    Ah, I see the problem. A little explanation is needed here. This bot will protect and unprotect templates TRANSCLUDED onto the main page. It will NOT protect or unprotect pages LINKED TO from the Main Page. The featured article of the day will never be protected or unprotected by this bot. Does that clarify things? – Gurch 20:00, 7 January 2007 (UTC)
    Though as I've pointed out previously, it is important to make clear that the bot will also protect templates (and pages) transcluded onto the featured article page (not just the main page template that contains the introduction to the featured article). Someone should do a nice diagram showing the relationships between the main page, its templates, the featured article, and its templates. And mark in red the stuff that ProtectionBot will protect. That might clear up some of the misunderstandings of some of the oppose votes. Please also see my comments on the talk page. Carcharoth 12:06, 8 January 2007 (UTC)
    Gah! After asking Gurch to be clearer, I forgot to do so myself. It is also important to make clear that ProtectionBot is protecting images on these pages as well. ie. images on the Main page templates (don't think there are any on the Main Page, but it would do that as well if any were added), and images on the featured article, and any images on any templates in the templates (I think - not sure about this last one). Carcharoth 12:10, 8 January 2007 (UTC)
    A detailed document explaining exactly what it does is ok -- but releasing the source code is a problem? Abu-Fool Danyal ibn Amir al-Makhiri 20:07, 8 January 2007 (UTC)
  9. Strong Oppose I do not trust most humans under the shaky structure of Wikipedia, lord knows what a bot would do. I won't support a bot until there's a far more stable policy making procedure in place. Just H 18:10, 7 January 2007 (UTC)
    Um... is there a reason for this random phobia? I don't think it even makes enough sense to be a phobia; what in the WORLD does policy making procedure have to do with anything? -Amarkov blahedits 18:12, 7 January 2007 (UTC)
    Will it ease your concerns when I say that this particular "Terminator" doesn't even know how wield a gun? Миша13 18:15, 7 January 2007 (UTC)
    The code will do exactly what it is supposed to, that is what code does. It does not have neural nets like the terminator, it cannot have ideas of it's own. Do votes based on fictional universes count? HighInBC (Need help? Ask me) 18:19, 7 January 2007 (UTC)
    I'm going to assume this is a joke and should be discounted as such, because I don't even want to begin to start the list of all of the ways in which real-life programs differ from the fictional killing machines in The Terminator. --Cyde Weys 19:07, 7 January 2007 (UTC)
    I'm appalled at the borderline personal attacks and lack of good faith above, and ultimately that's what is behind my phobia. This isn't an rfa for the bot, it's an rfa for the person holding the strings of the bot, and per the comments above, I have very little confidence in the average "string holder" on Wikipedia, notwithstanding the fact that the disruption caused by the bot would be beyond what any human could do. Just H 19:49, 7 January 2007 (UTC)
    Your original justification for the oppose, a fictional movie, made no sense, and people called you on it. Nowhere has anyone personally attacked you, and everyone has shown the appropriate level of good faith for someone who is claiming that a bot program is like the killer robots from a movie. As for your latest comments — if this is indeed an RFA for the "person holding the strings", then it would be a duplicate RFA, because the bot op in question is already an admin. Can you clarify? What "disruption" are you talking about from the bot?! It just protects things on the main page! Please offer up at least one argument that makes sense. --Cyde Weys 19:57, 7 January 2007 (UTC)
    Ok, first off, my original justification for this was not a movie, but rather the point behind the movie: if robots get too much power, humans won't be able to control them. That goes for something protecting pages just as much as any other robot with control over something. Second, this isn't about Dragon flight's abilities, this is about anybody's abilities: I do not trust the vast majority of the policy structure on Wikipedia as it is now. I would like to help fix that, but in the meantime, granting more powers to anyone or anything when there's any question is a bad idea. And no matter what my justification for my comment is, I have the right to it, and it's the bureaucrat's job to take it into account, not you and or anyone else. If you're going to rebut my comment show some decorum, regardless of your opinions on it. Thank you. Just H 20:09, 7 January 2007 (UTC)
    Ah, you've finally made the nature of your opposition known: it has nothing to do with the particulars of this situation; you think Wikipedia is broken somehow, and you'd say no to pretty much anything that comes along. It's not the bureaucrats' job to merely count votes; they will look at the discussion and see some of the more ... curious of oppose reasons. --Cyde Weys 20:19, 7 January 2007 (UTC)
    I was never hiding the nature of my opposition, Cyde. I do not trust any robot with admin powers at the current time due to the state of Wikipedia now. If that wasn't clear to you initially, I apologize. And yes, it is the bureaucrat's job to look at the votes, so I ask you to let them do it rather than trying to do it yourself. If you'd like clarification, that's fine, but nobody has the right to judge anybody else's comments except for the bureaucrats. Just H 20:46, 7 January 2007 (UTC)
  10. Oppose Very Strong Oppose - Not until I see source for this bot, if there is source available, then please tell me where it is. Cheers! —— Eagle 101 (Need help?) 18:14, 7 January 2007 (UTC) (see end of this chat section).
    I've heard someone say it's available here. :) Миша13 18:18, 7 January 2007 (UTC)
    Perhaps I should clarify, it needs to be open source, with the source available to all. The bots functions are not suspect to being abused by anyone that is not an admin. So until the code is opensource, sorry I cannot support. Not for something like this. —— Eagle 101 (Need help?) 18:29, 7 January 2007 (UTC)
    Open and free source is only the wiki way. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 18:41, 7 January 2007 (UTC)
    In addition I would like to note that this bot is not that urgent, we have User:Shadowbot, which can cover while this bot is being approved. It has been fixed to check the images, and I am sure User:Shadow1 has fixed it. —— Eagle 101 (Need help?) 18:49, 7 January 2007 (UTC)
    He has. Shadow1 (talk) 19:37, 7 January 2007 (UTC)
    It is urgent - the vandals are looking for ways to circumvent our protection all the time. But noone's rushing this RfA right now, so no worries. And I welcome Shadowbot as an alternative and extra measure. Миша13 18:59, 7 January 2007 (UTC)
    Sounds good then! Again my only reason for opposing is the lack of open source code. If that is rectified I am willing to reconsider my stance. Cheers! —— Eagle 101 (Need help?) 19:03, 7 January 2007 (UTC)
    The most recent instance of main page vandalism had nothing to do with an unprotected image. It resulted from an unprotected template. Not only did the bot fail to report that the template was unprotected, but it also e-mailed list subscribers (myself included) a false report that 18 other main page templates were unprotected. —David Levy 19:07, 7 January 2007 (UTC)
    I'll say this again: The only reason that the bot failed was due to a change in the MediaWiki "Page protected" text. I've since made the bot check if the text is editable for its protection check. In addition, I think you'll find that the template was deleted before the bot could perform its scan. If you want, David, I can decrease the scan interval. Shadow1 (talk) 19:37, 7 January 2007 (UTC)
    1. I don't know why you expected me to know this when you never added such an explanation to the page on which you and I discussed the problem.
    2. It might help if the bot were to scan the page with greater frequency, but isn't 00:00 (UTC) the most important and obvious time at which a scan should occur? —David Levy 20:04, 7 January 2007 (UTC)
    Sorry, I was thinking of a conversation I had on IRC. However, in any case, I'm not entirely sure what the problem is. The bot does scan at 00:00 UTC; it scans hourly. However, I'll probably change this to something a little faster. Even then, though, I don't see how my bot is any different from ProtectionBot, other than that it's proactive and scans more often than mine does. Shadow1 (talk) 20:25, 7 January 2007 (UTC)
    Strong Oppose - bot purpose is done by mediawiki now. Therefore I think the bot is no longer needed, if it ever were needed. Cheers! —— Eagle 101 (Need help?) 00:36, 11 January 2007 (UTC)
  11. Oppose If the code isn't public, I'm uncomfortable granting the mop. Xoloz 21:32, 7 January 2007 (UTC)
    Why would you like the code to be public? If it's due to security concerns, Dragon's flight has already had the code reviewed by numerous trusted users. If it's because it might be operating outside of its scope, this can easily be monitored through the logs. SuperMachine 21:49, 7 January 2007 (UTC)
    The code needs to be reviewable by anyone who wants to review it. If he won't accept a request from a brand-new user account, then a computer programmer or security researcher from outside Wikipedia will not be able to review it. If he will give the code to anyone at all upon request, regardless of their tenure or established trustworthiness, then he may as well release it publicly in the first place. In addition, Wikipedia is a free encyclopedia, and creating an adminbot that handles the main page is much more of an official approval of the software than is giving bot flags to all manner of individuals; the software must be free. —Centrx→talk • 22:33, 7 January 2007 (UTC)
    Oppose. Automated "sock" accounts sould not be given admin rights. One per customer please. 198.22.123.103 22:25, 7 January 2007 (UTC)
    Anons can't vote in RfAs. -Amarkov blahedits 22:27, 7 January 2007 (UTC)
    Yes they can.
    No they can't. Please see WP:RfA on commenting in the Support, Oppose and Neutral sections: "Any Wikipedian with an account is welcome to comment" Миша13 22:36, 7 January 2007 (UTC)
  12. Oppose, code is not open. Genuinely secure code is secure no matter who reads it. (Willing to change !vote if this is rectified in time to audit the code). Seraphimblade 22:32, 7 January 2007 (UTC)
    It isn't the security of the code that is the issue. The issue is that the code could with some basic changes do some things that would be rather damageing to wikipedia. I would rather that those attacks were not made any esier than required.Geni 00:03, 8 January 2007 (UTC)
    Oh, that old security through obscurity bit has been around forever-"If we publish the source, blackhats could make something malicious from it!" If this bot (through error or malice) really can do that much harm, it's far more important that the code be open to community inspection, not less! And if you think vandalbots can't be made now, I could put a basic one together in five minutes. Not hard. In general-anyone who can read, figure out, and alter the bot's code would be capable of writing h(is|er) own based on other tools out there. I'm not sure what there is to hide here. Seraphimblade 01:25, 8 January 2007 (UTC)
    Fortunetly most of our vandles can't code to that degree. Modifying existing code on the other hand.... In any case you don't need the code to figure out methods that might or might not work for dealing with that kind of attack.Geni 16:20, 8 January 2007 (UTC)
    Oppose. Gwoissa398 22:40, 7 January 2007 (UTC)
    The above user has been blocked indefinitely. Blocked users may not participate in RfAs —bbatsell ¿? 22:50, 7 January 2007 (UTC)
    And the reason being ... ? Миша13 22:47, 7 January 2007 (UTC)
    The reason being that it was a vandal's sockpuppet. Against an anti-vandalism tool. Go figure. :] --CBD 00:04, 8 January 2007 (UTC)
    You misunderstand I think- Миша's question was about the reason for the vote before it was struck through. Not the reason for discounting the vote. WJBscribe (WJB talk) 00:07, 8 January 2007 (UTC)
  13. Oppose. Bot needs to be open source. --- RockMFR 00:32, 8 January 2007 (UTC)
    That is not the fact, many bots are approved without being open source. HighInBC (Need help? Ask me) 00:37, 8 January 2007 (UTC)
    This is not a bot approval. This is an admin approval. Hence, if this user would like my support, they need to publish the source code. My opinion will not change until that happens. --- RockMFR 00:48, 8 January 2007 (UTC)
    Strong Oppose. Admin users should have only one account with extra priviledges. Wiki is Freaakky. 00:50, 8 January 2007 (UTC)
    Why? - CHAIRBOY (☎) 01:11, 8 January 2007 (UTC)
    Per the policy on multiple accounts: The community has strongly rejected users having more than one username with admin powers. If you leave, come back under a new name and are nominated for adminship, it is expected that you will give up admin powers on your old account. (You may do this quietly with your old account and not have to show a link between accounts.) You should have only one account with powers greater than those of a regular editor. And I support this. Wiki is Freaakky. 01:18, 8 January 2007 (UTC)
    1. Do you honestly believe that the above text describes a situation comparable to this one?
    2. You might not realize this (given that your account is only a few days old), but policy reflects the community's practices (not the reverse}. We don't follow the rules purely for the sake of following them. —David Levy 01:29, 8 January 2007 (UTC)
    Crossed out vote as banned user evading block Jaranda wat's sup 05:33, 8 January 2007 (UTC)
  14. Oppose Candidate has a total of one edit zero logged actions. Here is how I expect bot candidacies to be handled: 1. Run an alpha test on an admin account. 2. Ask Jimbo (or a bureaucrat in the bot approval group if Jimbo chooses to defer the decision) for a temporary promotion. 3. Run a beta test on the bot account and let the edits (and complaints on the talk page) pile up. 4. File an RFA. ~ trialsanderrors 02:58, 8 January 2007 (UTC)
    Of course the account has zero logged actions. That's because there was an alpha run on the account of the owner before this RFA. - Mgm|(talk) 09:26, 8 January 2007 (UTC)
    Comment - Interesting idea, I think a better way would be 1. Request approval at WP:BRFA, after approval is granted. 2. Run alpha test on the test wiki, 3. Ask Jimbo (or a bureaucrat in the bot approval group if Jimbo chooses to defer the decision) for a temporary promotion. 3. Run a beta test on the bot account and let the edits (and complaints on the talk page) pile up. 4. File an RFA. Again this is just a thought game, I really don't know if this is any better or worse then any other proposal. I may change my mind, ect, after hearing other ideas. Cheers! —— Eagle 101 (Need help?) 03:07, 8 January 2007 (UTC)
    Perfectly fine with me, as long as I can click on the ProtectionBot (talk • contribs • count) links and find content I can investigate. ~ trialsanderrors 03:14, 8 January 2007 (UTC)
    The bot approval group and bureaucrats who have chimed in have both objected to your suggestion, so that's pretty much out of the question. A successful trial was run using Dragons flight's account, and then the trial was stopped by the BAG pending the outcome of this RfA. I hope that you will reconsider your position, as everything was conducted according to the requests of those in charge; in addition, I hope that you read through the bot request page linked at the beginning of the RfA, as it should clarify all of this. Thanks, —bbatsell ¿? 03:12, 8 January 2007 (UTC)
    Well then I have to live with disagreeing with those in charge. ~ trialsanderrors 03:14, 8 January 2007 (UTC)
    You are replying to Trialsanderrors, right? I am just trying to figure out what would be a good idea for the future, Cheers! —— Eagle 101 (Need help?) 03:15, 8 January 2007 (UTC)
    Note - I figured out ;). Cheers! —— Eagle 101 (Need help?) 03:17, 8 January 2007 (UTC)
    I don't think it's fair to punish the main page because the adminbot process works a certain way and you would prefer a different way. Go push your proposed changes to the process if you must, but it doesn't make sense to oppose on the grounds of current procedures not being the way you'd like. --Cyde Weys 03:55, 8 January 2007 (UTC)
    I oppose on the grounds that the candidate has a total of one edit zero logged actions. ~ trialsanderrors 04:02, 8 January 2007 (UTC)
    First of all, this bot isn't ever going to make any edits, but rather, protections and unprotections, which do show up in page logs, but aren't edits themselves. Secondly, this bot already has undergone a trial run under Dragon's account, so characterizing it as never having done anything is inaccurate. You're just not looking under the right account for the actions it's performed. --Cyde Weys 05:51, 8 January 2007 (UTC)
    Thank you. I corrected it. ~ trialsanderrors 06:34, 8 January 2007 (UTC)
    I've been suggesting for a couple of days having the bot run provisionally so that it can be better evaluated and we have the benefit of the protection in the meantime, which would serve at least some of the purpose of the original Opposer, but this idea has not gained any traction, at least not yet. Please see discussion on talk. Newyorkbrad 03:58, 8 January 2007 (UTC)
  15. Strong oppose - there is no reason not to release the source code. Vandals who have the technical knowledge to modify the bot for vandalism purposes already have the technical knowledge to write a vandalbot from scratch. And a bot that will function with admin privileges must be released with source code, as far as I am concerned. Yuser31415 03:45, 8 January 2007 (UTC)
    I would like to express some strong disagreement here. I've managed to hack away perfectly well at pyWikipediaBot, but I'd never have enough time nor ability to learn everything required to rewrite it from scratch. It's simply false to say that it takes the same amount of technical knowledge to modify something than to make it from scratch, especially in programming. If I wanted to make a vandalbot out of AWB all I'd have to do is remove the CheckPage and then use AWB's built-in functions to have it replace lots of pages with nonsense. That takes maybe an hour's worth of hacking, tops, and not much technical knowledge. On the other hand, making it from scratch .... hahaha, it's so much more difficult, the comparison doesn't even make sense. --Cyde Weys 03:51, 8 January 2007 (UTC)
    So the only reason we do not release the source code is because we are afraid of the vandals? What annoys me is that a vandal could write the base of vandalbot, submit it, and any vandal could just change it to their heart's content and run the bot - so we are not getting anywhere by withholding the code, just being pedantic. Yuser31415 04:10, 8 January 2007 (UTC)
    I wasn't arguing in favor of keeping the code secret (I'd like to see it myself), merely pointing out the fallacies in your arguments against keeping it secret. I do not believe qualms over the open source-ness of the code outweigh its immediate and necessary utility. --Cyde Weys 05:52, 8 January 2007 (UTC)
  16. Weak oppose, largely on the "closed source" issue, mutable if either a) the source is made open, or b) a stronger argument for security-through-obscurity is made. I'm not seeing any "vandalbot" implications beyond those already present in pywikipedia and AWB (about which Cyde is clearly correct, but since those two ships have already sailed...). AFAICS the required functionality is a) to compile a list of transclusions from a given page, and b) to protect and unprotect from a list of pages -- right? Alai 05:16, 8 January 2007 (UTC)
  17. NeutralOppose, see below Until source code is released. Sorry, but there is absolutely no reason whatsoever to not release the source code. — Werdna talk 00:11, 8 January 2007 (UTC)
    I understand the request, but I'm not sure that "absolutely no reason" is entirely true; there are some reasons detailed on the talk page if you'd like to take a look. —bbatsell ¿? 00:21, 8 January 2007 (UTC)
    Thanks for pointing me to the discussion, but I stand firmly by my request. I have a serious problem with something like this being closed-source. I don't like the cabal-like way in which code is being distributed. And for the record, as a developer, I believe that the bot has the wrong approach to the problem anyway. Firstly, it should be implemented in MediaWiki, which I'm thinking of getting off my ass and doing. And secondly, it should be pre-emptively protecting TOMORROW's stuff, rather than running every fifteen minutes and protecting today's, using fancy RNG algorithms (which has been cited as a reason for not releasing the code). But the main issue is the fact that it's not being released. As I said on IRC, the risk that it malfunctions due to inadequately tested code is far greater than the risk that a vandal somewhere finds the code, and is, by some freakish miracle, unable to find any of the six or seven (or more) other frameworks for editing Wikipedia, and also has the skills to modify it into something harmful. To be honest, it looks to me as if the whole cabal-like "Well you can see the code but YOU can't" thing is yet another example of english wikipedians power-tripping and enjoying their act of being superior because they can see the code. Needless to say, I'm more than disappointed that the code has not been released — Werdna talk 00:37, 8 January 2007 (UTC)
    The only legitimate reason I can think of is the authors right to protect his intellectual property. Not saying this is what I think his motives are, just that this is the only reason I can think of that I agree with. HighInBC (Need help? Ask me) 00:40, 8 January 2007 (UTC)
    Sorry, I don't mean to seem argumentative, but just one correction: it does protect the next day's templates. That's stated clearly on the bot discussion page. —bbatsell ¿? 00:42, 8 January 2007 (UTC)
    Werdna, I'm sure there would be absolutely no issue released the code to you and others equally qualified and trustworthy. As for your suggestion that a better program could be written, I'm not qualified to comment on that, but the answer is, this is a problem that EnWiki needs solved now using the best solutions we have; if a better one comes along next month, nothing in the current solution stops us from moving toward a better one. Finally, I don't see the basis for your saying that this suggests an attitude of superiority on the part of EnWiki vis-a-vis any other project; EnWiki had a problem and solved it, but I'm sure there won't be one bit of reluctance to share the solution with any other Wiki that might be confronting the same issue. To HighInBC (edit conflict, he had a comment here a minute ago, where did it go?), I take Dragons flight at his word that he is concerned about maladaption of the code for a rotten purpose. I don't have the tech savvy to comment on whether that's a good reason not to release or not, but I doubt very much that intellectual property rights in the abstract are the slightest bit of the issue here. Newyorkbrad 00:44, 8 January 2007 (UTC)
    NYB, as I said, I did not mean to speak of Flight's motives, just that I disagree with his reasoning. His code contains no special technology that is not already available to the enemies. It would be easier to start from scratch than adapt his bot. HighInBC (Need help? Ask me) 00:49, 8 January 2007 (UTC)
    I'll have to trust you bot experts to come to a consensus on that issue, if you can. Regards, Newyorkbrad 00:52, 8 January 2007 (UTC)
    I'm switching to Oppose. I think that a dedicated bot account is unnecessary for a couple of protections per day, and the closed-source thing still bugs me would turn my support into a borderline oppose anyway. I encourage Dragons Flight to release the source and gain approval to run it under his/her regular account. — Werdna talk 06:39, 8 January 2007 (UTC)
    I would prefer it be on a dedicated account so that the contributions could be more easily monitored. HighInBC (Need help? Ask me) 02:28, 9 January 2007 (UTC)
  18. Strong oppose. There is no need for this. A five-line script can check the protection of the templates at 12:00:01 (UTC), and post a one-line message on WP:AN/I. The lack of releasing source code is pointlessly insulting...it conveys a "closed-room" mentality that should not be welcome here at all, and certainly should not be given free rein. --Connel MacKenzie - wikt 07:07, 8 January 2007 (UTC)
    There is a need. Vandals exploit admin mistakes and as the nomination stated, the main page was vandalized as result at least 6 times in the past month alone. Also, I don't see why everyone should be able to read the bot's code. Any abuse will result in desysopping of the bot and Dragon's account. There's no chance the owner will purposely let the bot do anything else but its intended task. - Mgm|(talk) 09:34, 8 January 2007 (UTC)
    Six successful vandalisms? I'm sorry, but whoop-de-doo. As to your second comment, yes, there is every possibility that an adminbot could be used for anything else, but I wasn't objecting on those grounds anyway. --Connel MacKenzie - wikt 17:38, 10 January 2007 (UTC)
    There is no need for this. May I suggest reconsidering that statement - it could be seen as an insult to all the people who have worked on creating, reviewing, testing, and conditionally approving (for a test run) this bot. All of them do believe there is a need for this, as do well over 100 other editors expressing support, above. It's fine to oppose this on other grounds, but you really should rethink your belief that the problem of template and image vandalization can be solved trivially - because numerous admins who have taken responsiblility for protecting the Main Page article against vandalism have concluded exactly the opposite - that such a bot is valuable. John Broughton | Talk 15:06, 9 January 2007 (UTC)
    I have reconsidered this topic, and found my opinion to be even more justified than I thought at first. As a bot writer, bot operator, Wiktionary sysop and checkuser, I simply cannot conceive how an argument has successfully been made to justify this enormous travesty. If you are going to take a technical approach to the problem, do it correctly: notify humans that something is amiss. If you are worried about exposure/disclosure, then simply direct your bot's talk messages to the User_talk: namespace, or to EmailThisUser links. --Connel MacKenzie - wikt 17:38, 10 January 2007 (UTC)
    I hate to say it, John, but I think that bruised egos are of secondary concern here. — Werdna talk 17:56, 10 January 2007 (UTC)
  19. Absolutely and unconditionally oppose this per Alkivar. Grace Note 07:19, 8 January 2007 (UTC)
    Is there a particular reason you "will not ever support an unsupervised admin bot" as per Alkivar? RHB 09:03, 8 January 2007 (UTC)
    Yes. Grace Note 01:06, 9 January 2007 (UTC)
    A slippery slope to what? This is an attempt to establish consensus after all. RHB 17:15, 10 January 2007 (UTC)
  20. Oppose per Werdna. The request indicates that the bot does not need the sysops bit. Recode bot to post relevant information about these "dozen templates and images each day" to its user page or elsewhere, where the many users involved in featured pictures, articles, and DYK (and others of course) can monitor and inform at AN or AN/I when sysops protection/unprotection is needed. Tim Shuba 10:51, 8 January 2007 (UTC)
    Something like this is already done by User:Shadowbot2. Carcharoth 10:58, 8 January 2007 (UTC)
    Er so, we should add more work for humans just for the fun of it? You going to do that extra work every day, or are you just volunteering it for others? --W.marsh 16:44, 8 January 2007 (UTC)
    For the record, I'm of the opinion that at least one of the attacks was aided by Shadowbot2's public report, which is why I've changed it to private. Shadow1 (talk) 22:33, 8 January 2007 (UTC)
  21. Oppose - Per Werdna. I'd at least want the code visible, since I'm paranoid like that. ST47Talk 11:42, 8 January 2007 (UTC)
  22. Strongest possible oppose. Lack of transparency is not going to protect this bot, it's going to invite intense scrutiny. I have some experience with this on Wikipedia. Furthermore, lack of transparency with this bot not only goes against everything that Wikipedia stands for, but what Wikis in general stand for. If you do not reveal the code, you will make the internet suck again. Nina Odell 16:15, 8 January 2007 (UTC)
    Holy hyperbole, Batman! Dragons flight has worked hard to develop a tool to assist in the protection of the main page and you're saying that if he doesn't release the source code, he'll "...make the internet suck again." How's that for gratitude? SuperMachine 17:10, 8 January 2007 (UTC)
    Comments like that do not add to the discussion. I wish more people would ask themselves if they were actually helping the encyclopedia by clicking the save button. Cheers, ✎ Peter M Dodge ( Talk to Me • Neutrality Project ) 17:13, 8 January 2007 (UTC)
    Thank you, Peter. Just H 19:52, 8 January 2007 (UTC)
    And oppose votes like the one I responded to help the encylopedia how exactly? SuperMachine 17:15, 8 January 2007 (UTC)
    Everyone has a right to engage in discussions concerning Wikipedia policy. Trying to belittle those engagements is at the very least counterproductive, and at the most, incivil. It was the major reason why I went from oppose to strong oppose. Just H 19:52, 8 January 2007 (UTC)
    I wasn't referring to the fact that someone opposed adminship for the bot, but rather the overly hysterical manner of the opposition. SuperMachine 23:44, 8 January 2007 (UTC)
    The original comment was referring to a quote by Jimmy Wales that "We make the Internet not suck" and how the free nature of Wikipedia is centrally important to that goal. —Centrx→talk • 17:26, 8 January 2007 (UTC)
    I was aware of the quote and its source, but the language used by Nina Odell was still unneeded hyperbole. As you demonstrated with your oppose below, it's quite possible to express the opinion that the source should be open without resorting to FUD. SuperMachine 17:30, 8 January 2007 (UTC)
    We've understood that you strongly support this candidacy. Harassing oppose voters isn't at all necessary, and could itself be characterised as "hysterical". I agree with this voter that lack of transparency is a bad thing and she is entitled to phrase that in any way she chooses. It's thoroughly impolite to harangue users you do not feel have expressed themselves in a way you approve. Grace Note 01:12, 9 January 2007 (UTC)
    WP:KETTLE? SuperMachine 02:49, 9 January 2007 (UTC)
  23. Wikipedia is a free encyclopedia. An admin-bot on the Main page is an officially sanctioned, high-profile tool. This tool must be free. In addition, the security arguments for keeping the source closed are weak; it would be simple to create a vandal-bot or find one without access to this source code, and the greater security issue is the inability of anyone except an "approved" Wikipedia user who discloses his identity to obtain it. —Centrx→talk • 17:24, 8 January 2007 (UTC)
    umm we had non free content on the main page as recently as January 5th.Geni 18:41, 8 January 2007 (UTC)
    That's not a good thing. Nevertheless, if the governments of the world wish to deprive our cultural heritage, it is legally necessary for us to abide by their rules. If Dragons flight and others want to deprive free, open access to computer software that facilitates the operation of the site, we do not need to sanction it. This is a technical limitation under our control, not a legal limitation imposed from outside. Many people could create a bot of similar function and provide the source code freely, thus making available to anyone the sort of tool that they are trying to protect by keeping it closed. Would this not be a waste of time and effort when the tool has already been created? Would the latter tool be better or worse than the one already created? Would it be another few weeks before it is written, tested, and approved? Creating another bot would nullify the reason given for keeping it closed source, and it would be a duplication; it would destroy the creator's time and effort if the later bot is created without the benefit of examining the earlier (and it would need to be done blind for copyright purposes). This can be avoided simply by releasing the source code freely. —Centrx→talk • 21:01, 8 January 2007 (UTC)
  24. I'm going with Centrx. Bad Guys (tm) could do major damage by exploiting flaws in MediaWiki itself. But MediaWiki's source code is open. What's your excuse? I don't like this one bit. Abu-Fool Danyal ibn Amir al-Makhiri 18:53, 8 January 2007 (UTC)
  25. Technical oppose Will not support without robust peer-review of code + transparency. —Adrian Lamo ·· 2007-01-08 20:31Z
  26. Oppose for several reasons:
    1. Lack of source code transparency. I think it's been shown definitively that security-through-obscurity is a myth, especially in this case.
    2. Full adminship for a protection bot is a hack, frankly. I see no reason to encourage laziness in the developers when a solution could be enacted through MediaWiki or finer granulation of the delegation of powers. I would support some limited adminbots in the future if the latter came to pass. I opposed TawkerBotA for similar reasons.
    3. Protecting everything utilized by the Main Page at 23:59, or 00:00, is trivial and does not require a bot. Protecting/unprotecting everything utilized by the Main Page throughout the day in real time is not something I trust a bot to do effectively and efficiently. It is not clear to me which of these cases is intended here, even after all the discussion above.
      After reading the RFBA more closely, it's clear that a form of the latter is intended here, but not in real-time, just at 'randomized' intervals. My concerns about this leading to a false sense of security for the main page are not alleviated. 18:24, 9 January 2007 (UTC)
    Nothing here should be taken as an assignation of Dragons_flight's motives. While it has been stated that no further actions can/will be taken by the bot, I have reservations regarding the security/robustness of the code that prevent me from supporting, and in fact make me oppose this request for permissions. The people who complain that this never should have gone to RFA in the first place worry me even more. -- nae'blis 22:27, 8 January 2007 (UTC)
  27. Reluctant oppose. It should go without saying that the source for a bot with the admin bit be open for examination by everyone, especially on a project like this built on open source. A shame really, since this bot is sorely needed; I've seen one too many Main-page penises lately. - Merzbow 22:32, 8 January 2007 (UTC)
  28. Weak oppose I've been really on the fence, but even though I know nothing about coding, there's something disconcerting about granting a bot without open source code admin powers. -- Kicking222 00:10, 9 January 2007 (UTC)
  29. Oppose per Hipocrite, Wizardry Dragon, and others. It'll get my support if and only if the code were fully free, but it's not, so it doesn't get my support. Centrx and Abu-Fool make especially compelling arguments. jgp TC 06:15, 9 January 2007 (UTC)
  30. Oppose. Admins have to be accountable for their own actions. Bots have no brain/emotions so can't be held accountable. Plus, I really don't like the idea of adminbots. --Montchav 12:36, 9 January 2007 (UTC)
    What exactly do you mean by "accountability"? A bot can be blocked at the first sign of malfunction with no hurt feelings whatsoever. Try doing that with a non-robot admin who starts malfunctioning. In this sense a bot is much more accountable than a human. Unless of course, when you say accountability, you are really talking about punishment, in which case you're right ... you can't punish a bot for doing wrong, while you can punish a human. Still, I find the emphasis on punishment disturbing. --Cyde Weys 15:17, 9 January 2007 (UTC)
  31. Oppose No problem with bots automatically finding problems, but a human needs to be responsible for the final decision.--Osidge 13:06, 9 January 2007 (UTC)
    Humans are resonsible for the final decision. The final decision is to protect pages that are transcluded to the main page so shock images aren't placed there by vandals. This bot will only do the decisions made by humans. It is not making any judgement call other than Is this page transcluded onto the main page. HighInBC (Need help? Ask me) 15:53, 9 January 2007 (UTC)
  32. Oppose. While I personally have no problems with an adminbot, I think that the source code should be open. A normal bot with closed source code is fine, but the bot's extra privileges make it extremely important that any bugs in the code are found, and I think that having open source code would increase the number of people who would be able to spot bugs. It's not a problem with the operator's coding ability, but leaving code open to public scrutiny is probably the best way to fix the bot if it malfunctions. Shadow1 (talk) 13:11, 9 January 2007 (UTC)
  33. Strongly Oppose Not only is an unsupervised admin bot rather worrying (when the hypothetical situation arises that one is needed for the continued operation of wikipedia then I'll consider it, we are currently not in that state of need as such it shouldn't be used as it's arguement for inclusion). Another reason is as wikipedia is built around the concept of open source I'd prefer as much as possible if done likewise, so while it is not a major point of mine it is still is another minor reason for opposing that the source code is not available. Also the bot can not accept or decline the nomination! Not a good way for a bot to start of being an admin by ignoring the rules of RfA. lol Mathmo Talk 15:48, 9 January 2007 (UTC)
    We are allowed to ignore rules, especially when it would be foolish to apply them. Such as when a bot runs for RfA, it would be silly to apply rules created for humans. Thank god for IAR or we would be stuck here, thankfully we can use common sense. HighInBC (Need help? Ask me) 15:50, 9 January 2007 (UTC)
    Yup yup, I know about that! You didn't notice the "lol" at the end? It is not my serious reason for opposing, don't worry if that was my only "reason" to oppose this then I would not be opposing at all! Mathmo Talk 15:58, 9 January 2007 (UTC)
    Ah, I see. The line between joke and honest beleif seems to be blurred here, in general. HighInBC (Need help? Ask me) 16:05, 9 January 2007 (UTC)
  34. Oppose The reasons given for keeping the code secret seem insufficient to me. When TawkerbotTorA's code was released, the code was adjusted due to feedback from RfA participants. There will be no chance to do this on ProtectionBot, and the benefits of keeping the code secret seem minimal compared to the benefits of releasing it. I think there were cases in the past when vandalbots were combatted by specific antivandalbots, and if there were a vandalbot created as a tweak on ProtectionBot's code at least we would have a clue to its code and be able to combat it more effectively. (Has anyone considered the possibility that the current template vandal is a bot? If so, keeping ProtectionBot's code secret is completely pointless.) If the code is released before the end of the RfA, count me as a support. --ais523 18:10, 9 January 2007 (UTC)
    Oppose not until the source code is released open source. →AzaToth 20:51, 9 January 2007 (UTC)
    Somewhere else, you wrote "I want to be able to examine the code" as an explanation why you want to have the code of this bot open source. Couldn't this be fulfilled by sending the code to you for checking? If not, would it be sufficient to publish the code later, when the bot is already in use or do you want to have the code published as a prerequisite before agreeing with granting the bot account admin rights? --Ligulem 23:46, 9 January 2007 (UTC)
    The problem is that I have problem discussing the code with people, what kind of restrictions apply for me to look at the code? →AzaToth 00:55, 10 January 2007 (UTC)
  35. Oppose I will not support admin candidates without a significant history of writing quality articles. I think bots with extra permissions should be possible, but not via the RfA mechanism. Crum375 22:59, 9 January 2007 (UTC)
    Excuse me? Saying you won't support admin candidates who haven't written articles when you know you're talking about an anti-vandalism bot is nonsensical. If you don't like the RFA mechanism for bots then work to change it, but in the meanwhile, it's all we have. The correct action for you would be to abstain, not oppose (and take it out on the main page) merely because you don't like that other people forced this to go through RFA. Dragons flight didn't want to have to bring this through RFA, but others left him no choice; now you're smacking on him from both sides, which just ain't fair. --Cyde Weys 03:43, 10 January 2007 (UTC)
    I don't believe that 2 wrongs make a right. Calling my position 'nonsensical' won't change it - I don't think that we should promote a bot to 'admin' status, and thus I oppose using the RfA process for approving bots and their permissions. 'Abstain' would have no impact on the outcome and would not reflect my position. As I noted above, I do believe there is room for bots with higher permission levels, but this is the wrong forum and mechanism for approving them. Crum375 12:09, 10 January 2007 (UTC)
    A more correct mechanism, it seems to me, would be an extension of WP:RFBA to allow added permission levels for bots, using special criteria (perhaps open source being mandated) and possibly even final approval by Jimbo. Crum375 13:45, 10 January 2007 (UTC)
  36. Oppose Closed-source worries me, and WP:RfA is not the right place for the bot.SYSS Mouse 03:34, 10 January 2007 (UTC)
    Well there is no other place, so it doesn't make sense to oppose the bot, for no fault of its own, because this is the current process for gaining any form of adminship. You can go to change the policy as you'd like but in the mean time it's not fair to take it out on the main page and our readers by opposing a necessary anti-vandalism measure. --Cyde Weys 03:43, 10 January 2007 (UTC)
    Agreed, on that count -- if not RfA, then where? Especially considering the community's having rejected adminbot proposals, in the past, it seems important to establish consensus. Luna Santin 07:40, 10 January 2007 (UTC)
  37. Secret code, "trusted" users, etc etc etc. This entire process appears to me inimical to the philisophical way we work. - brenneman 12:18, 10 January 2007 (UTC)
    Oh yeah, and per XFlux's "I'd prefer a MediaWiki feature for this task" this kind of ludge is a really good way to ensure you never get one. Nothing lasts like the temporary hack. - brenneman 12:22, 10 January 2007 (UTC)
    Oppose. Diffyoeu224 18:59, 10 January 2007 (UTC)
    Note to crat. User's sixth edit. RyanGerbil10(Упражнение В!) 19:05, 10 January 2007 (UTC)
    Also, account five minutes old at time of vote. RyanGerbil10(Упражнение В!) 19:07, 10 January 2007 (UTC)
    Oppose. Doesn't it say that users should only have one account with admin tools? Izzy3984 20:45, 10 January 2007 (UTC)
    Are you proposing that only non-admins be allowed to run adminbots? Luna Santin 20:57, 10 January 2007 (UTC)
  38. Oppose. Concerns over supervision and closed source code. How do we know what it is going to do unless we see the code. Will it have a big red off button? Why can't it simply generate a list of things that it thinks should be protected but are not, with buttons for interested admins to press to protect or unprotect them? -- ALoan (Talk) 20:53, 10 January 2007 (UTC)
  39. Oppose. See Wikipedia:Requests for adminship/TawkerbotTorA. I don't like the idea of a bot admin.--Grand Slam 7 | Talk 21:05, 10 January 2007 (UTC)
    Does that mean you prefer to see penises on the main page? Even after the recurring images of penises, mutilated vaginas, etc. have been brought to the attention of administrators, the maze of main page template transclusions has proven to be too complicated to be taken care of by humans on a completely perfect basis. RyanGerbil10(Упражнение В!) 21:29, 10 January 2007 (UTC)
    Oppose per Izzy3984 . Hiddy3a 21:59, 10 January 2007 (UTC)
    Note: user's first edit. Dar-Ape 22:03, 10 January 2007 (UTC)
  40. Oppose per User:Nae'blis point 3. Catchpole 22:40, 10 January 2007 (UTC)
    Oppose. Hannah Barbabsba 23:58, 10 January 2007 (UTC)
    User has 4 edits, and someone needs to cut it with the socks. Seraphimblade 00:13, 11 January 2007 (UTC)
  41. Oppose disagree with bot admin. Blnguyen (bananabucket) 00:17, 11 January 2007 (UTC)


[edit] Neutral
  1. Still strongly opposed to the idea of admin bots, but this one is necessary, so neutral/no !vote. – Chacor 09:38, 7 January 2007 (UTC)
  2. Like Chacor I am also opposed to admin bot accounts but this is a very serious problem so this may be necessary. It's a shame it's not possible to allocate solely the protect function without other admin fucntions. Tim! 11:04, 7 January 2007 (UTC)
    It is possible. It just requires a bit more work by the developers. If this "tweak" were implemented, I'd be much more comfortable considering changing my stance on this RfA. Tomertalk 11:07, 7 January 2007 (UTC)
    I'd just like to point out that the bot only has code for protecting/unprotecting stuff, so I couldn't ever delete or block anything. We're not talking about Skynet here, which might get out of control and end the human race. We're just talking about a normal program that is limited to what it has been programmed to do. There's no strong AI involved, I can assure you. --Cyde Weys 15:07, 7 January 2007 (UTC)
    I don't think the depth of the disruption is the issue as much as the possibility of it and the lack of control over it. Sure, we can shut it off, but why wait until that point? Just H 19:57, 7 January 2007 (UTC)
    Couldn't agree more with Cyde. This ain't V.I.K.I but Vicky :) — Nearly Headless Nick 14:30, 9 January 2007 (UTC)
  3. I prefer shouting at administrators when they mess up to having a bot to do the job. It helps to relieve stress. Primarily, though, I'm only casting this neutral vote so that I can write the following text where it will be read. This is not a sentient entity. This is a computer program. It doesn't "know" it's an administrator. It can't "decide" to go and block everyone. Unlike humans, of course, which know such things only too well. In other words, if you don't trust this bot, there's no way in hell you should trust a human. I mean, all those emotions, and things. You don't know what's going to happen. Wait, why are they trusted anyway? That's it, I DEMAND to see the source code to ALL administrators, now, otherwise they should resign immediately – Gurch 15:39, 7 January 2007 (UTC)
    Human flcelloguy = new human (admin);
    flcelloguy.edit("now");
    flcelloguy.protect("something");
    ... on a more serious note, though, I appreciate what Gurch is trying to say. To re-emphasize his point, bots only follow the code that they've been programmed, and the bot has already proven in its trial period that it can work as expected. Thanks! Flcelloguy (A note?) 16:01, 7 January 2007 (UTC)
    There's a buffer overflow vulnerability there. I'm afraid I'm going to have to ask that you be desysopped – Gurch 16:02, 7 January 2007 (UTC)
    lol, c'mon Gurch. Give Flcelloguy a chance to be debugged first ;-) Just H 19:58, 7 January 2007 (UTC)
    Security through obscurity does not work! I demand to know Flcelloguy's password! --CBD 20:18, 7 January 2007 (UTC)
    Source code for all administrators? You can start with DNA, though it might need a lot of decompressing and debugging. All complaints to God and evolution, please. Carcharoth 11:58, 8 January 2007 (UTC)
    Actually, we could probably have our DNA analyzed and then sent to us so we could post it in a user subpage. But that might actually take up a lot of server space, even though it's just text... Grandmasterka 12:24, 8 January 2007 (UTC)
    But that would probably violate WP:OR, WP:RS, WP:V, WP:NOT, but not WP:PERF. Not sure about WP:NPOV... :-) Carcharoth 14:45, 8 January 2007 (UTC)
    WP:BLB hehehe you figure this one out crz crztalk 00:27, 9 January 2007 (UTC)
    Wikipedia:Biographies of living bots? :-) Carcharoth 01:59, 9 January 2007 (UTC)
    I'm glad to trust this bot once I know I can see the code for myself. Why is that so hard to understand? --badlydrawnjeff talk 13:00, 8 January 2007 (UTC)
  4. Neutral I fully trust Dragons flight and his ability to program a good bot, but is a bot even considered a user? As of now, only users are allowed to be nominated for adminship. The policy on Administrators needs to be reworked before this nomination can pass. Cbrown1023 16:10, 7 January 2007 (UTC)
    Remove the words "user with an" from the sentence: "Any user in good standing may nominate any user with an account." That should do it – Gurch 16:13, 7 January 2007 (UTC)
    Everyone understands that a "Request for adminship" is not really the best name for the process that is being followed here, but the bot needs to operate with sysop rights since page protection is an administrator-only function. In yesterday's discussion on ANI there was agreement that consensus in favoring of operating the bot under a sysop flag needed to be gathered, and this is the best and most visible vehicle for doing so. Newyorkbrad 19:50, 7 January 2007 (UTC)
    Neutral, until question 1 has been answered. Staying neutral for now. I strongly support the bot in its current function, but I also strongly oppose the idea that further functions of this bot could theoretically be approved at Wikipedia:Bots/Requests for approval rather than here. Until it is explicitly stated that this will not happen, I can't support. Sorry. --Conti|✉ 19:26, 7 January 2007 (UTC) Changing to Support. --Conti|✉ 20:32, 9 January 2007 (UTC)
    Neutral - I'm too worried the bot has read I-Robot. :) Actually, I'd be more supportive if the code was public. Do any of the BAG crew agree that we are better off with the source private? ---J.S (T/C/WRE) 18:06, 7 January 2007 (UTC)
    Actually, if the bot had read I Robot then it would be the perfect robot. 1. Do not harm humans. 2. Follow orders. 3. Protect yourself. In that order. Methinks you are getting some weird impression of what I Robot is actually about from some hackneyed movie of the same name. --Cyde Weys 22:24, 7 January 2007 (UTC)
    That was a book before it was a movie you know, not the the two have much more in common than the name. HighInBC (Need help? Ask me) 23:29, 7 January 2007 (UTC)
    ... which is exactly why I posted my comment (did you respond to the wrong person?) --Cyde Weys 00:42, 8 January 2007 (UTC)
    Bah, ruining my joke. ---J.S (T/C/WRE) 05:42, 8 January 2007 (UTC)
  5. WTF?? Neutral - Why the heck is a bot running for adminship? Maybe I'm not aquainted with this stuff, but isn't this for actual people & isn't there a separate page for electing bots? Also, isn't a fully automated bot kinda like what happened in Terminator I, Terminator II etc? And look what happened there, the computors started killing everybody & messing up the space-time thingy by sending Arnie back & forth. ;) Anyway, at the risk of sounding extremely dumb, what is this thing in layman's terms & what does it do? Who will be running it? (I assume nobody, which I think is kinda bad...) And how will this make Wikipedia a better place? And most importantly, what do we do if it causes armageddon & starts saying stuff like "I'll be back..."? Thanks, Spawn Man 07:24, 8 January 2007 (UTC). P.S. Armageddon tired of all this confusing stuff... ;)
    I refer you to my neutral vote above. I have made the relevant text larger so that you (and hopefully others) will notice it. Thanks – Gurch 12:06, 8 January 2007 (UTC)
    The Terminator is Governor of California so while we're on about it...ProtectionBot for ArbCom! Why not? :) --WikiSlasher 03:17, 9 January 2007 (UTC)
  6. Neutral - I can't support unless i see the bot at Category:Administrators open to recall. Prevention is better than cure. -- Szvest - Wiki me up ® 11:46, 8 January 2007 (UTC)
    I'm inclined to agree, on that count -- with the caveat that I'd prefer people don't start a recall just because of "zomg adminbot" and are bringing up actual problems, there should be strong accountability, and this is especially true for a bot of this nature. Whether through this mechanism or something else, it seems a good general idea. Luna Santin 12:17, 8 January 2007 (UTC)
    Why would there be any need for Category:Administrators open to recall, or indeed any other process? If the bot does something it isn't supposed to do, it gets desysopped. If we think Dragons flight was acting maliciously, as opposed to making a genuine mistake, he gets desysopped too. To achieve these things we go straight to a steward in the former case, and first to a steward and then to ArbCom in the latter. No extra bureaucracy required, surely? – Gurch 12:38, 8 January 2007 (UTC)
    I think it's a good idea - don't think of it as extra bureaucracy, think of it as an additional gentle soothing layer of reassurance for those who feel a little uneasy about a sysop-o-bot. That can only be a good thing. Proto::► 15:03, 8 January 2007 (UTC)
    Note though we should all remember that the bot is not an administrator - it's just a humdrum bot with the sysop flag ticked so it can protect pages. But I'm all for adding it to the category mentioned if it reassures people. Proto::► 15:05, 8 January 2007 (UTC)
    It can be added to that category if you want, but it's not really necessary. The administrator recall process is designed such that desysopping is hard. In comparison, it is incredibly easy for malfunctioning bots to be desysopped, blocked, have their approval withdrawn, etc., without having to go through such a process. The Bot Approvals Group knows what they're doing, and if any issue did show up with ProtectionBot they'd resolve it much, much more quickly than the recall process would ever proceed. --Cyde Weys 17:45, 8 January 2007 (UTC)
    Bots don't care what is their block log, as they don't have feelings to be hurt by mistaken blocks. If there is a problem, block first and ask questions later, don't wait around for long drawn out process. (but be ready to do what the bot does if necessary... that's a concern I actually DO have, that we will come to rely on this bot and forget to double check things) There is no need for this bot to be in any recallable category. For that matter, there is no NEED for the bot's MASTER to be in any recallable category either!!!! Certainly Dragons flight would be welcome in the category, if so desired, but I am just not at all keen on anyone using that category membership as a requirement for anything. Please. ++Lar: t/c 20:39, 8 January 2007 (UTC)
    If the bot stops working, the following safety valve operates (see User:ProtectionBot): "If you want to monitor ProtectionBot yourself. You may include {{User:ProtectionBot/Stopped}} on any page. This template will normally be empty as long as ProtectionBot is operating normally." - I suggest that {{User:ProtectionBot/Stopped}} is permanently transcluded into WP:AN or WP:ANI or both. Also, please see the talk page of this RFA for more discussion on human oversight of ProtectionBot - many more eyes are needed on that discussion. Carcharoth 23:44, 8 January 2007 (UTC)
    Comment Just remember this: It is in the nature of man to create monsters and it is in the nature of monsters to destroy their creators. Tread carefully. 65.102.35.249 00:48, 9 January 2007 (UTC)
    What ridiculous and inaccurate hyperbole. If any of us actually could make a computer capable of rebelling against its masters we'd be billionaires. Strong AI is an incredibly hard problem to crack and the potential monetary gain from such a creation is basically unlimited. --Cyde Weys 15:21, 9 January 2007 (UTC)
    Maybe he was referring to Wikipedia processes such as RfA. They do have a life of their own, you know... Abu-Fool Danyal ibn Amir al-Makhiri 15:58, 9 January 2007 (UTC)
  7. Neutral If it's a bot, it wouldn't need it since it can revert faster than a human can edit. This would be oppose, but supposedly as an admin it would never point of view push, so that makes it a neutral. Anomo 02:33, 9 January 2007 (UTC)
    I'm sorry? Did you understand that this bot will not be performing reversions or content edits? It will only be protecting and unprotecting high-risk templates and images, which requires the sysop flag. That's the only reason the bot is undergoing an RfA. —bbatsell ¿? 03:33, 9 January 2007 (UTC)
    Support. I admire the opposers' general "open-source-at-any-price" attitude (and have far less sympathies for "OMG a robot admin what next?" ones), but this is a tool designed to do a tedious job. The recursive tracking of templates used on the main page is difficult for a human to get right, (just because it's recursive), and it's important to get it right. Practicality certainly has to win over philosophy. Duja► 16:03, 9 January 2007 (UTC)
    This is the neutral section, but your vote looks like a 'support'. Are you sure it's in the right place? --ais523 16:09, 9 January 2007 (UTC)
    Bah, after 3 edit conflicts and 42 timeouts, it's no wonder... Seems I already managed to vote above, or someone else did it for me :-) Duja► 16:26, 9 January 2007 (UTC)
  8. Abstain only because I'm a techno-dummie. I fully trust Dragons flight, and strongly believe something must be done about the main page vandalism, but I simply do not understand the technical issues involved with bot programming well enough to be able to give a support vote. SandyGeorgia (Talk) 20:12, 9 January 2007 (UTC)
  9. Too few main namespace edits. Oh, also source code should be made public. Not that I do not believe in Dragons flight's good faith, but personally it contradicts Wikipedia' spirit. Necessary probably, but contradiction nevertheless. -- ReyBrujo 21:16, 9 January 2007 (UTC)
  10. The bot is good, bot operator good also but the only problem about having a bot admin is that a single mistake can cause huge problems. Cheers to 2007! User:Sp3000 01:09, 10 January 2007 (UTC)
    It seems to me that if the bot made a single mistake, say, leaving one page unprotected, this would still be a smaller proportion of unprotected material than currently seems to make its way onto the main page. Browsing through Talk:Main Page and its archives will show periodic complaints of vandalism to the main page. With the bot, though, the coding can be fixed to prevent future errors, and furthermore, any of +1000K administrators can protect the page in the mean time. Dar-Ape 01:34, 10 January 2007 (UTC)
  11. Abstain I would really like the source code to be release, so we can inspect and discuss the source. →AzaToth 03:46, 10 January 2007 (UTC)
  12. Neutral - I don't object to using it, but let's not call it an administrator. It's just a utility. Deb 12:48, 10 January 2007 (UTC)
  13. Neutral - Uncomfortable with the idea of a 'bot running unsupervised with admin powers; persuaded by oodles of knowledgable and well-respected supporters arguing passionately in favour, but only to the extent of squeezing a neutral out of me, rather than a support pile-on. I earnestly pray we don't live to regret this and hope to live to reflect ruefully on my stupid cautiousness here when it works like a dream until discarded for some jolly good reason in the future.Users are invited to post "told you so"s in January 2008. --Dweller 16:44, 10 January 2007 (UTC)
The above adminship discussion is preserved as an archive of the discussion. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the talk page of either this nomination or the nominated user). No further edits should be made to this page.