Wikipedia talk:Survey notification

From Wikipedia, the free encyclopedia

Contents

[edit] Kudos for this Proposal

This policy is long overdue. However, this would be a policy that would be hard to convict someone of violating. joturner 22:39, 18 March 2006 (UTC)

I agree it's a great idea. I don't agree it would be difficult to "convict" someone of violating at all. There are plenty of cases where vote stacking is absolutly clear. Theresa Knott | Taste the Korn 16:54, 23 March 2006 (UTC)
I agree that this is a good policy as well, and agree with Theresa Knott regarding the easy identification of "bad" vote stacking. For instance: This edit, which shows a person soliciting someone to change their vote. SchuminWeb (Talk) 02:20, 25 March 2006 (UTC)
That seems to be a completely legitimate request. Articles often improve during the process but we all know people often don't go back to look at them again. Choalbaton 05:19, 26 March 2006 (UTC)

[edit] Magnitude

It seems to me that there's a difference between vote stacking and consensus building. I see nothing wrong with contacting a fellow editor and letting him know that a discussion is ongoing; even if I believe he agrees with me. It's only when I decide to contact 5,000 (or however many) of my fellow editors and claim they're all close personal friends that this is a problem. So while I think this proposal is a good starting point, I'd like to make sure we don't throw the baby out with the bath water. Inviting an associate to offer his opinion on a topic is not, and should not, be a violation of wikipedia policy. We need to figure out how to draw the line to separate collegial discussion and intercourse from pure politically-motivated vote-stacking. Nandesuka 16:59, 23 March 2006 (UTC)

[edit] WikiProjects

Are WikiProjects, by default, assumed to agree with the guidelines they have developed? For example, would nominating a category for renaming because it violated a project's naming conventions and informing the project about it be a violation of the third point in the proposal? Kirill Lokshin 17:31, 23 March 2006 (UTC)

Wikiprojects are expected to be read by both people who agree with the goals of the project and people who do not - I expect the most avid readers of some wikiprojects are people who think those projects are terrible. As such, contacting a project, which is an open, public space, cannot fall afoul of point 3, as long as the notifing lanugage is neutral. This is different than user talk pages, which are typically not monitored by anyone but the user in question.
Projects designed to POV push or disrupt wikipedia to prove a point are dealt with in other policies, and are not the concern here. Hipocrite - «Talk» 17:41, 23 March 2006 (UTC)

[edit] Incompatibility with AGF?

I think that certain aspects of this proposed guideline conflict with Wikipedia:Assume good faith. Specifically, it says:

Notifing only people who you believe would agree with "your" position is a violation of this guideline.

This is purely subjective - how can an outsider know what someone is believing and what his motives are behind his actions. An outsider can assume that X, who is spamming talk pages with a link to an AFD is only notifying the users who X believes will support his position. Is that outsider assuming good faith, by doing that? I know that this may seem a bit wikilawyerish, but I was wondering about it earlier. --Latinus 20:32, 25 March 2006 (UTC)

I agree, this line in the proposed policy concerns me for the reasons outlined by Latinus and Nandesuka (above in the "Magnitude" section). I would support this policy becoming official, but only if this line is altered or changed. Kit O'Connell (Todfox: user / talk / contribs) 21:31, 25 March 2006 (UTC)
An example of the potential problems this might cause -- say an article in Category:Burning Man came up for deletion on AfD. I might take it upon myself to inform the editors listed in Category:Wikipedian Burners. I do not see this as vote stacking if I post a neutral notice to them, because they have declared an interest in the event and its culture and therefore might be presumed to have an opinion on the issue. However, it is also likely they will vote Keep on many Burning Man related topics that others might not find so notable for inclusion in the 'pedia. I would not want to be accused of violating this policy as a result. Kit O'Connell (Todfox: user / talk / contribs) 01:15, 26 March 2006 (UTC)
I suggest that what you propose is already a violation of de-facto policy, and if you did it, is more than likley to get you banned. Hipocrite - «Talk» 13:47, 27 March 2006 (UTC)

[edit] This proposal would actually INCREASE opportunities for abuse

If this policy was implemented articles would be in great danger from groups that took the trouble to organise a bloc of votes ourside wikipedia (where the "stacking" could not be traced) or on some specialist page in Wikipedia that would not be linked from the deletion page, or in another language wikipedia etc etc. This is clear from two recent debates on Kurdish topics, where votes on one side came through quickly at the beginning and the opponents were slow to organise. Who knows where the Turks organised themselves, but the counter effort is quite easy to trace. This policy has an implied bias in favour of the group which acts first as it can choose when and how to strike and has time to plan in such a way that it won't get caught breaking the policy, even if it has done so. As it happens the international consensus to cover the Kurds fairly is overwhelming and it is a high profile topic, but if this proposal is implemented it will become easy for groups with an agenda to highjack and delete articles on less well known controversial issues. Choalbaton 05:34, 26 March 2006 (UTC)

I agree. We should keep discussion here on Wikipedia, out in the open. This is a bad proposal. Johntex\talk 17:59, 27 March 2006 (UTC)

[edit] Naïveté

It is pointless to ban people only from notifying people they think are likely to agree with them, as it is a matter of course that virtually all such notifications will be made to likely supporters, so any ban should apply to all notifications. But that would be bad for at least three reasons:

  • As explained above it would hand control to people who are sufficiently motivated to organise elsewhere. Often these will be the most biased people
  • Insights the people not contacted might have offered will be lost as most users in any given subject area probably do not review the deletion pages regularly.
  • It is an interference with freedom of discussion between acquaintances. Choalbaton 05:43, 26 March 2006 (UTC)
I have notified people I was certain would disagree with me as recently as three days ago. Good acts, like notifing everyone who voted in a prior poll on the same subject, should not be prohibited, while bad acts, like notifing everyone who voted yes on a prior poll, are already prohibited in fact, not yet codified in guideline. The guideline must go to motive, as what we are prohibiting is bad faith notifications, not notifications. The difference is in the motive. Hipocrite - «Talk» 13:46, 27 March 2006 (UTC)
And what about when you notified Wikipedia:WikiProject Wikipedians against censorship? The very fact that these people have a project with "against" in the name says that they have organized in favor or one viewpoint and against another. Notifying that group amounts to notifying a set of people who could reasonably be expected vote disproportionately in a certain direction. Johntex\talk 18:02, 27 March 2006 (UTC)
Hipocrite, vote-stacking is accepted no matter how egregious. Or at least, I haven't seen any cases where votes were discounted or anyone was blocked because of it. Indeed, the ArbCom has explicitly noted in at least one case that an editor was guilty of repeatedly mass-soliciting votes, but specified no action to be taken against him. (That editor does so to this day, in fact.) —Simetrical (talk • contribs) 22:28, 12 April 2006 (UTC)
Yet, this editor was banned. Derex 23:55, 23 October 2006 (UTC)

[edit] And nominators obligation to inform...

Even if there can be some practical problems in implementing this policy, I am sympathic towards the principle. I think however, that avoiding vote stacking should go hand in hand with an obligation for the one creating the nomination to inform involved parties about an AfD - the ones have that started and been involved in the article in question. Otherwise the policy is vulnerable for misuse: Imagine if someone makes a controversial proposal for deletion, only inform people that support deletion and then launches a case for vote stacking if one of their opponents spread the word about the AfD. Bertilvidet 19:33, 28 March 2006 (UTC)

[edit] A bad idea

This should be seen as a non-starter in line with Choalbaton's in section 5. I came here because this proposal is currently being advocated on the Village Pump by a user who makes determined efforts to get Kurdish articles and categories deleted even though it is obvious that there is hardly a user other than Turks and Iranians who is in favour of this. I presume his idea is that with this policy in place the Turks could all get their votes in early as usual (what a coincidence it is that that happens; surely no prior organisation is involved as he is against vote stacking!?) and he could then argue that some of the counter votes were invalid. Of course manipulation of votes is wrong in theory, but this proposal won't improve the situation. I wonder if the proposal itself was actually made in response to the Kurdish issue as the anti-Kurdish lobby seems to be prominent in pushing it.CalJW 15:30, 29 March 2006 (UTC)

[edit] Not really a solution

Per some comments I made a while ago (see Wikipedia_talk:Requests_for_adminship/Archive_44#Are admin standards too loose?, I think imagining polls or surveys determine Wikipedia-wide consensus is delusional. IMO, the root problem is not vote stacking (which I agree is an issue), but self-selection of voters, and the only real solution is to implement some mechanism using statistical surveys. Direct democracy (which is what we currently use) simply doesn't scale to the population of users we have or the number of decisions we make. Some sort of representative democracy might work, but would be counter to the general egalitarian principles of Wikipedia. Rather than spend any time or effort instituting an anti-vote stacking policy, I think we'd be much better off proposing a workable method based on surveys of a random selection of users. -- Rick Block (talk) 18:34, 29 March 2006 (UTC)

We have that. ANy random person can go to Wikipedia:Articles for Deletion. No reason to adverise. --Cool CatTalk|@ 15:04, 11 April 2006 (UTC)
Um, I think "statistical selection" means something like "selected by a random number generator to be programmed into the software". The difference between that and what we have now is like the difference between getting randomly selected for jury duty and assigned to somebody's trial, versus letting anyone interested in that person's case decide to go sit on his jury ("self-selection"). Phr 00:50, 12 April 2006 (UTC)
I would like to see the software, for each issue that comes up (meaning every XFD, policy poll, etc.), pick out N of the past 24 hour's changes by registered, non-bot users and invite the author of each edit to give their opinion on the matter. This sample, being representative of Wikipedia editors and giving due weight to those who contribute more, would then be used to determine whether the motion would pass or fail. (If one of the people offered a vote declined, it would then be offered to someone else to fill out the appropriate number of people.)

What exactly N should be would have to be figured out—I would think that if no one challenges an XFD it can get by with three people or so and deleted/kept if unanimous, otherwise referred to a larger panel of perhaps a dozen. Significant policy proposals, of course, would require panels of at least a hundred, probably more.

Multiple consecutive edits would be compressed for this purpose, so editing the same page multiple times in a row wouldn't get you more chance of being selected. Possible abuse would be dealt with on a case-by-case basis.

This is, I think, what should happen if we're serious about policy being decided by polls rather than votes. Currently it seems we aren't. Of course, the community would have to be in massive agreement for such a huge feature to even be put on a "to do at some point in the distant feature" list, so this is probably moot unless someone's willing to do a lot of coding. —Simetrical (talk • contribs) 22:22, 12 April 2006 (UTC)

[edit] Madness

Any attempt to denigrate given votes or categorize efforts to bring people to a discussion as "acceptable" or "unacceptable", "in policy" or "in violation", "good" or "bad" is fraught with peril -- and that's the best possible interpretation. At worst, it's merely another loaded gun tossed onto the table.

Let's encourage those who debate controversial matters to bring as many like-minded editors to the table as possible! (I don't suggest you get all your buddies to register onwiki just for the purpose of voting; that's meatpuppetry and although I don't know we can enforce a policy against that either, it's not really relevant to so-called vote packing.) It can't be wrong to have more voices in a debate.

There is a great desire among editors to somehow avoid all politics; we wish to pretend it is possible to have a society in which there are no factions, no parties, no subgroups of the whole. But this is unrealistic. The larger the community, the more it tends to subdivide. We can't legislate human nature. All we can do in our fumbling attempts to suppress political activity is to make furtive what should be done openly, publicly, and with pride. John Reid 10:44, 31 March 2006 (UTC)

Very well said. Johntex\talk 15:44, 31 March 2006 (UTC)
We are not talking about asking 2 people to join the debate but mass advertising an afd to 16+ people of which almost all voted in a manner identical to the advertisers mostly not even commenting why, but just agreeing with the advertiser.
Wikipedia is not an experiment with anarchy. I find gathering of one sided votes highly disruptive. Infact I do not understand why there is a big fuss over this proposed policy that should be implimented at once in order to deiscourage meatpuppetary as well as senseless voting.
One can advertise a vote to lets say 3 people. I would find that to be a reasonable number. I am fine with 4 people, just not ok with 10+ or 20+
  • I would expect all of those 3 people to come up with arguments supporting their views rather than agreeing/disagreeing with the one advertising.
  • I would also hate to see the 3 people in question to vote identical to the same person on practialy any vote. This is meatpuppetary and is not welcome. If everyone is thinking alike noone is thinking.
--Cool CatTalk|@ 14:49, 11 April 2006 (UTC)
The problem is that the policy, as stated, doesn't indicate that telling 4 people is OK. Perhaps the proposal needs to be updated. I'll take a crack at it. Nandesuka 15:55, 11 April 2006 (UTC)

[edit] Salvage some good from bad

Even if this proposal gets shot down, I'd like to see part of it survive. We should at LEAST prohibit people from mass mailing people at random (or all administrators, or all members of a certain Wikigroup, or all X, or all Y) in an indiscriminate manner. --Dante Alighieri | Talk 16:55, 11 April 2006 (UTC)

[edit] Oppose article 3

The rest seems very good to me (as guideline, not as policy), but the article 3 is not a good idea. Loom91 08:14, 12 April 2006 (UTC)

Article 3 is in fact the core of the idea. Could you be more specific about what you don't like? I have addressed my main problems with it by rewriting it for clarity and adding article 4. Nandesuka 11:45, 12 April 2006 (UTC)

[edit] Meatpuppets

This proposal intersects meatpuppet policy and might reference that. The main difference is that meatpuppets are non-Wikipedians while VS means going after Wikipedians who otherwise wouldn't visit a particular issue. I'm uncomfortable with the formulation that it's ok to notify up to 2 3 4 of your bestest closest friends since it begins to be instruction creep. I think the issue of self-selection mentioned earlier is significant and if this VS proposal doesn't help then statistical sampling is the next thing to try. Phr 16:11, 12 April 2006 (UTC)

[edit] AFD etiquette

The AFD page mentions two kinds of notification that are acceptable or even desired:

  • "Notify the creator and/or main contributor(s) of the article when nominating, as they may be able to address concerns raised. Templates {{AFDWarningNew}} {{AFDWarning}} and {{Adw}} are available to this purpose."
  • "If the article links to other pages, place a notification on those pages (if they are relevant enough). This will (hopefully) get those with knowledge on the subject the ability to participate on the debate."

Perhaps a VS guideline should ask that people comply with those two forms of notification before they start notifying people individually? Perhaps there should be a template for the 2nd type mentioned above? Additionally, it might be desirable to incorporate "Be open. Don't make cross-posts that initially appear to be individual messages" from Wikipedia:Spam#Internal_spamming. Шизомби 01:21, 13 April 2006 (UTC)

There is an interesting case of internal spamming presently occuring on AFD. In addition to be open about cross-posting, it would also be ethical to indicate why that particular group of people is solicited. I don't think we can really do anything about mass solicitations, except ask that it be quite open in all regards. Unfortunately, this sort of thing is one reason user box's had such potential to be divisive. Brillig20 08:25, 13 April 2006 (UTC)
I'd intended to post over here before, but that case is what prompted me to get around to it. About 3/5 of the people notified had the Category:Members of the Vast Right Wing Conspiracy user box, which could be coincidence, or could really be (gasp!) a right wing conspiracy - which is sort of funny I guess.
Another possible thing that could be done would be to have a template for user talk pages that people could place at the top of their user page. Something on the order of a "do not call registry" or a "no solicitors" sign by one's door, it could state something like "This user would prefer not to be contacted about AFDs." It could also contain, or people could add, exceptions - like the two mentioned in AFD etiquette. Шизомби 15:59, 13 April 2006 (UTC)

[edit] Suggested wording change

I suggest changing

In summary: telling a few good friends about an ongoing discussion, even if they agree with you, is fine.

to

In summary: telling a few good friends about an ongoing discussion is fine, even if you're sure they will agree with you.

but then I'm a language lawyer. Cheers, Chris Chittleborough 03:28, 16 April 2006 (UTC)

Well, I've WP:BOLDly made this change. (BTW, I'm too new around here to have an opinion about this proposal. I just wanted the language to be clearer. Maybe I've read a few too many ISO specs ...) Cheers, CWC(talk) 01:00, 22 April 2006 (UTC)

[edit] RV of edits by Morton devonshire

I reverted his last two edits, which I think go against what the proposed policy is trying to aim at. Morton devonshire is (I don't think he would disagree with this characterization) someone who has in the past notified about fifty people about a single AfD who could be expected to recommend the same thing as he, and hoped that those people in turn would contact other people to do the same. A "vote stacker" such as he might have some interesting and possibly valuable ideas about what a proposed policy on vote stacking should be, but I think they should be brought up on the talk page here first.

I strongly dislike vote stacking, and would like to see some sort of policy put into place regarding it, but there are still problems with the one as currently proposed that need to be improved upon. Шизомби 18:52, 6 May 2006 (UTC)

There is no voting on Wikipedia -- it's about determining consensus. When viewing an Afd, the Closing Administrator is obliged to consider first the opinions of people who regularly edit the article, but that shouldn't be the only opinions she weighs. If a person properly considers Wikipedia policy regarding NPOV, notability, and the other policies which are relevant to Afd's, and applies these precepts to the article in question, then their opinion ought to be considered by the Closing Admin as well, even if they were invited to look at the Afd. We are a community of consensus, not votes. Comments from users whose histories do not show experience with or contributions to Wikipedia, and particularly, to the article in question, are traditionally given less weight and may be discounted by the Closing Administrator. Morton devonshire 19:30, 6 May 2006 (UTC)
One of the problems I see with vote stacking (or consensus stacking, if you like) is that it works against broad consensus. It serves to promote the consensus that exists among a narrower group of people who share an ideology or a support or opposition to a subject. One could argue that when stacking is being done by one "side" that the other "side" should simply do the same, however I think it would be better for all "sides" to refrain from such behavior.
Vote stacking tends to make the conscientious closing admins' jobs harder. The lazy admin will count the recommendations, and when there has been stacking, the stacked consensus will simply be followed. VS also seems to tend to result in "ditto" posts and baseless recommendations, rather than valuable observations about what ought to be done with an article. The conscientious admins have to sift through far more people's posts and histories to determine the best course of action; vote stacking increases the size of a discussion while decreasing its value and wastes their time. Шизомби 20:09, 6 May 2006 (UTC)
The problem really is that for most Afds, there isn't a broad enough consensus to make a determination -- it's just the same-old POV-pushers, again and again. The community wants a NPOV encyclopedia, not one dominated by hardcore ideologues. The broader the consensus, the better. I'm not here to debate you, because based on our earlier interactions, I think you are non-appeasable and will just keep responding endlessly. Can we have anyone else's thoughts, please? Morton devonshire 04:49, 7 May 2006 (UTC)

I think the name change from Vote stacking to Survey Notification is probably a sensible one, although notification should probably be lower-case and possibly should be plural. I don't like Morton devonshire's other edits, but won't revert for now - what do other people think? Шизомби 06:04, 7 May 2006 (UTC)

[edit] policy already exists

This policy already exists at [1], doesnt it?--Urthogie 12:52, 7 May 2006 (UTC)

Sort of. Straw polls, however, don't have "opening and closing times" or "binding" results quite like AfDs and the like. I think the idea here is to develop a policy or guideline that extends to things like those as well. Шизомби 17:26, 7 May 2006 (UTC)
Ok, that makes sense. But maybe you should remove the overlap.--Urthogie 20:04, 7 May 2006 (UTC)

[edit] Accept that the system is flawed anyway

There is absolutely no way of knowing whether any vote represents the views of the majority of Wikipedians as they only ever get a tiny number of votes. What is needed if anything is a higher participation rate, and this policy works directly against that. And as I said before, it is more full of holes than a Swiss cheese, handing the advantage to people who willing to take the trouble to stack a vote outside of Wikipedia, while lowering everyone's guard as it will be assumed that stacking is not happening. Choalbaton 21:11, 7 May 2006 (UTC)

Call me naïve, but I think that many AfD votes do better than merely gauge consensus: they are discussions, invoking (relatively) objective guidelines, that help determine whether an article should "actually" be deleted, transcending popular opinion. If that makes sense. Ardric47 08:21, 18 May 2006 (UTC)

[edit] Mostly good

Seems good to me. But the policy should cleary draw a line between what is acceptable and what is not. For example I have been recently accused of vote rigging by posting an info about a vote to a regional noticeboard. It would be nice if the policy would address such a case.--Piotr Konieczny aka Prokonsul Piotrus Talk 21:15, 24 May 2006 (UTC)

[edit] Request of known votes

One addemdum that I would like to propose. I believe that it should be a violation, in the case of a CFD, AFD, or another concensus-building areas if: 1) The vote is a repeat of a similar vote which was finalized as "no concensus" or the equivalent and 2) that someone notifies only the commenters on the previous similar poll which had a specific vote. (It is more likely than not that they would "vote" the same way again, although it is possible they would not.) Instead, a notification to prior parties should be done uniformly, if at all. This could result in spam to uninterested parties, if they voted on a non-concensus poll before. Thoughts? JRP 07:58, 30 July 2006 (UTC)

My only issue with this is that it could lead to a block under WP:SPAM, and it seems that this proposal does not rectify that. SchuminWeb (Talk) 11:43, 30 July 2006 (UTC)
That's true, perhaps. There is a real problem with one-sided "pinging" of re-votes. And escalating it just seems like such a terribly bad idea. JRP 13:26, 30 July 2006 (UTC)
Still, before we can go forward on this issue, we need to rectify it with WP:SPAM. That's a biggie. SchuminWeb (Talk) 03:02, 31 July 2006 (UTC)
If we accept all of WP:SPAM, which I suspect that we should, then I would argue that re-pinging of people that have voted in a specific way in a specific poll is a special case of Votestacking. To quote from WP:SPAM:
Votestacking is sending mass talk messages out to editors who all share a common userbox and informing them of an upcoming vote. A general consensus has formed that this behavior is unacceptable.
In this case, we have mass-talk messages which are aimed at a particular group with an "official" opinion. In the same sense that userboxes define how you nmight vote in a controversy, this type of vote canvassing provides even a higher degree of assurance that you might vote in favor. So, by WP:SPAM, I would suggest that both should be disallowed. Of course, this is only a guideline... JRP 13:01, 31 July 2006 (UTC)
In practice, though, it seems that WP:SPAM is treated as a de facto official policy, and therefore "this is only a guideline" isn't a good reason for dismissing it, as violations of WP:SPAM are considered "blockable offenses" (I have other issues with use of blocking, but this is neither the time nor place). And with WP:SPAM covering vote stacking already, I'm now not quite seeing the reason for this proposal's existence. SchuminWeb (Talk) 22:33, 31 July 2006 (UTC)
Possibly this proposal should be merged with WP:SPAM, but notice that it's considerably stricter in certain senses than that. People are blocked for talk-page spam, but the spammees' votes aren't necessarily discounted, and neither does anyone tend to do much of anything about e-mail spam. This proposal's suggestion is to standardize the ways to do both, and it proposes a different solution (everyone says where they came from) from that currently practiced (block people you know are mass-soliciting votes). —Simetrical (talk • contribs) 19:50, 1 August 2006 (UTC)

[edit] Internal spamming is not a blockable offense

It's been used recently by the School Project people to fight deletionists on the afd lists. Admins knew about it and no one was blocked. Mattisse 20:15, 1 August 2006 (UTC)

[edit] It is helpful to be notified when an article is nominated

It has happened that articles or categories I would have been interested in voting on were deleted when I did not even know that they were nominated for deletion. Sometimes I even fail to notice nominations of articles on my watchlist. Although people usually put "afd" or "cfd" in the edit summary of the edit on which the nomination is added, if I do not check my watchlist between the time which the afd/cfd nomination is made and the time of the next edit, I will not notice the "afd" in the edit summary on my watchlist, and will not know the article is nominated for deletion. This proposal might be more fair if it was possible for the system to somehow notify people when an article is being notified for deletion. In addition, it once happened that Category:Entertainers_by_age_upon_death was nominated for deletion almost immediately after a previous cfd had closed with no consensus. Someone notified me that it had been renominated, and I appreciated the heads up. I don't know who else was notified. The most fair thing to do in that circumstance would be to notify everyone who previously voted, and would be unfair to only notify people who voted a certain way. However, it seems very appropriate to me to notify people that an article or category has been renominated quickly after a previous afd/cfd closed. If you toss a group of 10-15 coins together enough times, you are bound to have an occurance with less than 25% of the coins fall on heads. Similarly, if an article is nominated for deletion enough times, even if 50% of wikipedias think it should be kept, and 50% think it should be deleted, a particular nomination will eventually result in a consensus to delete. I think there should be limit to how often an article or category can nominated for deletion, and a way for the system to notify people of an AfD before this proposal becomes policy. Q0 09:18, 14 August 2006 (UTC)

[edit] Problems

Copied from the main page... SchuminWeb (Talk) 05:03, 16 September 2006 (UTC)

There are two main problems with survey notification:
Retrieved from "http://en.wikipedia.org../../../s/u/r/Wikipedia_talk%7ESurvey_notification_a647.html"