- 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
Denied.
Operator: —
Wenli (
reply here)
Automatic or Manually Assisted: Automatic
Programming Language(s): PHP (CLI)
Function Summary: Watches
RfAs. Updates the vote count (0/0/0) and removes blocked/anon/duplicate users from the "Support", "Oppose", and "Neutral" sections.
Edit period(s) (e.g. Continuous, daily, one time run): Continuous
Edit rate requested: No more than 3 edits per minute
Already has a bot flag (Y/N): No.
Function Details: The bot first obtains a list of RfAs using the MW API. For every RfA, it:
- Uses
RfA Analysis Library to analyze the RfA
- Updates the vote count at the top of the RfA if it is incorrect.
- Indents votes made by anons and blocked users (using the API to determine if the user is blocked)
- Indents duplicate votes (two or more votes made by the same user)
Maybe it should move the anon votes the the "discussion header".
Ρх₥α
03:33, 4 November 2007 (UTC)
reply
This, potentially, is a very nice and useful bot; but given the fairly large disruption that would happen if an RfA page is messed up, we'll need to see the code and hammer on some
faux RfA in user space for a while.
Is the code already in a state where it can be tested, or were you looking for prior approval of process? Here is what I see to be the important stages for approval of this bot:
- Get some code working;
- Have it play with simulated RfA pages in userspace for a while; so that we can throw "hard" cases as it and see how it works;
- One we have a good behavior, and some diffs to show for it, run this by
Wikipedia talk:Requests for adminship, and possibly
Wikipedia:Bureaucrats' noticeboard to see if there is general approval for the task as perfomed;
- A trial on a couple of real RfAs; then
Profit! approval.
How does that sounds? —
Coren
(talk)
03:36, 4 November 2007 (UTC)
I can already think of some pitfalls we have to watch for:
reply
- Someone breaking the "normal" formatting - how will the bot fail?
- Users becoming banned after voting
- Stricken votes
Discuss. :-) —
Coren
(talk)
03:38, 4 November 2007 (UTC)
reply
If the bot does not detect the normal RfA format (using regexp), it will skip that page. Now, we need consensus on how the comments should be dealt with (struck, indented...) so that I can implement them. —
Wenli (
reply here)
04:13, 4 November 2007 (UTC)
reply
- There is no consensus about the removal of blocked users' comments at RfA (see BN archives from last month). Any bot which does this will be rejected, because of the lack of consensus and also it's a decision which requires discretion (and is often referred to the b'crats).
Daniel
11:02, 4 November 2007 (UTC)
reply
- At the very least it should be limited to indef blocked users only. I think the comments should be left; if other users feel the need to remove/indent/strike the comments, that's their prerogative.
Mr.
Z-man
03:43, 5 November 2007 (UTC)
reply
- Instead of removing comments, it could leave a note directly below the comment informing other users that the user voted twice or was blocked. I think that comments made by anons should be moved to the discussion section. —
Wenli (
reply here)
00:28, 6 November 2007 (UTC)
reply
We do not need a tally bot. RFA is not a vote (most people use !vote which means "not vote"), so there's no need to count it.
Majorly (
talk)
13:05, 4 November 2007 (UTC)
reply
- Although it's been often said that if there is a tally at the top of the page, it should be as accurate as possible. That does not mean it's not a good idea to remove the tally entirely. —
Coren
(talk)
14:14, 4 November 2007 (UTC)
reply
If anyone would like to see the regex that is used to update the tally, I can post it here. —
Wenli (
reply here) 00:33, 6 November 2007 (UTC)
I would argue against approving this bot; first, discounting remarks by blocked users is not something that should be done by a bot as it requires deeper investigation of the situation. Second, the edits made to update the tally are of no value and serve only to clog the history of the RfA; since the tally can at best provide a vague sense of the progress of the nomination, increasing the accuracy of the tally does not add meaningful value -- at least, not enough to justify bloating the history of every RfA by 20-50 edits.
Christopher Parham
(talk)
01:16, 6 November 2007 (UTC)
reply
- I strongly support this bot. We urgently need a bot to update the tally; many people are simply too lazy to do it themselves when they vote (as everyone should do).
Walton
One
10:18, 9 November 2007 (UTC)
reply
I, for one, do not like this bot proposal. The tally means nothing, so it makes no sense to update it. There should be no striking of comments from any automated process. In general, striking comments should be left to bureaucrats; either way, bots should certainly not be doing it. --
Deskana
(talk)
20:14, 30 November 2007 (UTC)
reply
- To me, the linked discussion doesn't seems to be in favour of such a bot. IMHO, putting a bot to update the tally does somehow recognize its importance, something I believe we shouldn't do.
- I respectfully think that no vote should be striked, but only indented. And no, not by a bot. I have never been and never will against using bots, but only if they're needed. RFA is enough complicated now, no need for a bot to step in too.
- It should be up to the 'crats to ignore votes by banned users, but it's not that simple. Just because a user have been banned/blocked doesn't mean that his vote should be ignored. Only in specific occasions this should be done.
- In conclusion, I respectfully disagree with the idea of a bot stepping in at WP:RFA, which is already seen by some as a vote, even without encouragements in this idea. Happy editing,
Snowolf
How can I help?
22:30, 1 December 2007 (UTC)
reply
- Denied. (per conserns voiced by Deskana and lack of consensus. --
uǝʌǝs
ʎʇɹnoɟ
ʇs(st47)
00:25, 2 December 2007 (UTC)
reply
- 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.