Wikipedia:Bots/Requests for approval/NKbot
From Wikipedia, the free encyclopedia
- The following discussion is an archived debate. Please do not modify it. Subsequent comments should be made in a new section. The result of the discussion was Withdrawn by operator.
[edit] NKbot
tasks • contribs • count • sul • logs • page moves • block user • block log • flag log • flag bot
Operator: Nakon
Automatic or Manually Assisted: Supervised automatic.
Programming Language(s): PHP
Function Summary: The bot will hardblock IP address of Tor exit nodes, per m:No open proxies, for a pre-determined amount of time (currently six months).
Edit period(s) (e.g. Continuous, daily, one time run): Daily
Already has a bot flag (Y/N): N
Function Details: The bot is designed to block a list of IP addresses that are determined to be Tor exit nodes. The addresses are exit nodes, retrieved from the TorStatus page, which is linked from torproject.org's documentation page. The bot will check each of the addresses listed in the CSV file against the Tor node checker on the toolserver [1]. If an IP address is listed in the block file and is reported as an exit node by the toolserver, it is hardblocked with account creation disabled. A log of all blocks is recorded after each batch for later analysis. The bot does not block registered user accounts nor perform any other administrative functions. Blocks placed by the bot will not be indefinite, as proxies close over time. I am also an administrator on this project and on Meta. Nakon 04:34, 4 May 2008 (UTC)
[edit] Discussion
- In before zOMG adminbot. Is it going through RfA? (Or have I misread?) dihydrogen monoxide (H2O) 04:48, 4 May 2008 (UTC)
-
- Not as of yet. However, they have explained it well, and I have no problems with it going through an RfA. Soxred93 (u t) 04:49, 4 May 2008 (UTC)
- I would much rather have it run a trial, in userspace (posting who it should block), before, though. After those results are reviewed, we can decide whether to run an RfA yet. Soxred93 (u t) 04:52, 4 May 2008 (UTC)
- Yes, agreed, that'd be good. I also have no problem with it in theory...but I know some people do! dihydrogen monoxide (H2O) 05:00, 4 May 2008 (UTC)
- Not as of yet. However, they have explained it well, and I have no problems with it going through an RfA. Soxred93 (u t) 04:49, 4 May 2008 (UTC)
Redundant to a software feature. — Werdna talk 05:04, 4 May 2008 (UTC)
- Man, Werdna, you beat us to everything! Where is the software feature? Soxred93 (u t) 05:06, 4 May 2008 (UTC)
I added it years ago, but nobody's ever bothered to enable it for blocking tor. MediaWiki has inbuilt support for blocking users listed in DNS blacklist. Numerous providers offer DNS blacklists including Tor exit nodes. — Werdna talk 05:09, 4 May 2008 (UTC)
-
- Any plans on enabling? dihydrogen monoxide (H2O) 05:23, 4 May 2008 (UTC)
- Per request, I've run a small sample (25 IPs). The results are at User:NKbot/25sample. This is what the log output looks like. Nakon 05:15, 4 May 2008 (UTC)
Correct me if I'm wrong, but wasn't User:TawkerbotTorA supposed to do the same thing (but the RfA failed)? —paranomia (formerly tim.bounceback)a door? 14:44, 4 May 2008 (UTC)
Of your 17 trial blocks, [2][3][4][5][6] are already stale. Many Tor nodes use dynamic IPs. How do you intend to avoid this? --uǝʌǝsʎʇɹoɟʇs(st47) 14:47, 4 May 2008 (UTC)
- I discussed this with someone on IRC last night. I could add a function after the blocking run that would check all blocks placed by the bot against the toolserver and unblock any closed nodes. Nakon 17:21, 4 May 2008 (UTC)
- This task seems rather duplicative of the job my bot KrimpBot does - KrimpBot already contains logic for detecting nodes as they open and close and tags the IP's talk pages appropriately, only stopping short of blocking as it would require an RfA - would you perhaps like to collaborate somehow? (As proposed, I'm also concerned that the methods you plan to use - grabbing a list from a third party and making several thousand HTTP requests to a toolserver tool, instead of retrieving and parsing the directory entries yourself, seems prone to inefficiency and outdated data.) krimpet✽ 20:28, 4 May 2008 (UTC)
- As I recall, the half-life of TOR exit nodes is somewhere around twelve hours. Given that, blocking for six months seems like massive overkill. --Carnildo (talk) 22:39, 4 May 2008 (UTC)
There does seem to be a community consensus against adminbots. Can you explain what special measures you will take to run an adminbot? AKAF (talk) 09:01, 5 May 2008 (UTC)
This kind of admin bot, in particular, isn't going to get consensus support from the community. It would only get my support if there is a way for TOR node operators to get an exemption by showing that they have restricted editing access to Wikipedia through their node. TOR is not inherently evil, it's just that the way it interacts with editing Wikipedia is problematic. rspeer / ɹəədsɹ 06:46, 6 May 2008 (UTC)
- The authoritative directory does allow one to see if access to Wikipedia has been restricted from that node by parsing the node's exit policy - that is how KrimpBot works, for example. krimpet✽ 23:37, 15 May 2008 (UTC)
I maintain that this is best implemented in MediaWiki. If the existing measures are inadequate, I will write an extension to specifically target Tor nodes. — Werdna talk 09:50, 6 May 2008 (UTC)
- Agreed; this sounds like an excellent idea. It would be nice to get some proper consensus first on enabling this functionality. —paranomia (formerly tim.bounceback)a door? 18:49, 8 May 2008 (UTC)
- Agree per last two comments. archanamiya · talk 19:17, 13 May 2008 (UTC)
Extension is half-written. Currently liasing with tor as to maintaining a comprehensive, accurate list, and tor nodes will be treated as if they were blocked with account creation disabled. Should be live within a few weeks. I don't see any scope for this bot. I will not decline, as I have a conflict of interest. — Werdna talk 13:28, 19 May 2008 (UTC)
- I agree with declining if this extension will take care of the bot's actions. Nakon 14:05, 19 May 2008 (UTC)
- The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section.