Wikipedia talk:Page Curation

More Autopatrollers

What about extending autopatrolled to any editor who has produced one or more GAs / FAs, regardless of whether or not they've also produced 49 almost-meritless unreferenced stubs ... ;P Pesky (talk) 21:42, 7 March 2012 (UTC)

• No, WSC, I was referring to my comment above about having a distinct and wider RfC on this precise issue. And Pesky, no ;p. Okeyes (WMF) (talk) 00:57, 8 March 2012 (UTC)

Debundling from Administrator rights makes sense on a theoretical level, but in practice I don't think there are any ongoing problems with Administrators abusing page creation. I think the threshold of 50 should be regarded as a vague guideline rather than a rigid rule. There are some people who start lots of bad stubs who may not be ready with even 100 starts; others with a couple dozen might be perfectly ready in terms of firmly understanding inclusion and sourcing standards. Carrite (talk) 17:24, 3 April 2012 (UTC)

A good point :). Okeyes (WMF) (talk) 18:11, 3 April 2012 (UTC)

One practical difficulty with making GA and FA writers Autopatrolled is that we don't know whether their other articles would be notable or not. Our FAs and GAs may be our best articles, but they aren't necessarily our most notable ones. Also the Autopatroller userright is only relevant if people actually create new articles - you could in theory be a prolific FA writer and yet have a new article count of zero. Appointing more Autopatrollers would certainly improve the Newpage patrol process, if we were to do so I'd suggest that the following routes would make most difference:

1. Currently we don't make someone an Autopatroller if they have a history of copyvio. I'm not sure what proportion of copyvio is picked up by newpage patrollers as opposed to corensearchbot and anyone else who sniffs out copyvio, but my impression is that page patrollers are not doing those sort of checks. If they aren't then perhaps we should stop worrying about copyvio when we appoint autopatrollers as it won't make a difference. If some newpage patrollers are checking for copyvio then maybe we can workout a guideline, how many copyvio free new articles does a former cut and paste merchant need to contribute before we can safely assume that they've stopped committing copyvio?
2. We used to expect 75 articles before we'd consider editors for Autopatroller; Then we cut that to 50, and that is discretionary. However we aren't very good at spotting the qualified candidates and appointing them. In particular the editors who quietly contribute a finely honed new article a fortnight often get overlooked when it comes to appointing Autopatrollers. I don't see this as something that triage can deal with, we really need more admins to plough through Wikipedia:Database reports/Editors eligible for Autopatrol privilege. There are plenty at the bottom of that report who could uncontentiously be made Autopatrollers. ϢereSpielChequers 16:54, 9 April 2012 (UTC)

Locking an article for triaging

The project page now talks about a new "article view", a dedicated interface for articles you reach through Special:NewPages, you can also "lock" an article reserving it for you (to prevent conflicts) and use a dedicated editor to edit the article without leaving the article view. If you lock an article for your triaging will others (not using NPT/NewPages) see that lock too? Will locking/unlocking be an individual contribution? --08:34, 14 March 2012 (UTC), Utar (talk)

I like the idea of reducing edit conflicts at NPP and can see a big advantage in enabling patrollers to target the articles that their fellow patrollers aren't working on. But there is also the risk of spammers tag teaming and having one account create an article and another lock it. So the mechanic of this is important. I'd prefer terminology such as reserve over lock as lock implies an overly rigid demarcation. ϢereSpielChequers 08:49, 14 March 2012 (UTC)

It's ultimately the same action, but sure. And yes, that could totally happen - if the spammers did not mind (a) still having other people able to see it and (b) others being able to review their actions as soon as it is pattrolled and the window is closed. Okeyes (WMF) (talk) 14:18, 14 March 2012 (UTC)

Yes, any technical ability to restrict edits to an article, even temporarily, needs to be controlled to prevent abuse. The biggest thing for me is preventing accidental multiple notifications (particularly speedy notifications) from simultaneous CSD nominations since these multiply the biteyness of the experience for editors. VQuakr (talk) 02:45, 17 March 2012 (UTC)

Agreed. Perhaps a time shutoff? If no actions are taken within N minutes, unlock? Okeyes (WMF) (talk) 19:01, 17 March 2012 (UTC)

As long as it was implemented in a way that prevented abuse. If this just was a specific color code on the NPT display rather than a function that actually restricted modifications to the article, then I suppose the potential for abuse is much lower. VQuakr (talk) 02:55, 18 March 2012 (UTC)

I'll find out what the plans are and get back to you :). Okeyes (WMF) (talk) 15:52, 18 March 2012 (UTC)

This is not what "locking" means. I'm not sure that's the best term. It's not about preventing other people from editing the article; it's a mechanism to prevent duplicative work (e.g., two people start triaging the same article at the same time). The system will be smart and will let you know if someone else is currently looking at an article. That's all this is.--Jorm (WMF) (talk) 21:20, 18 March 2012 (UTC)

So will it tell you "you can't edit this article now, somebody is in" or only warn you "somebody is currently triaging this article, please triage another one"? That's the basic question. --21:39, 18 March 2012 (UTC), Utar (talk)

It will tell you "someone is currently triaging this article, please select another one" if you attempt to directly access it, or it simply won't show up in your triage stream until it's freed.--Jorm (WMF) (talk) 22:43, 18 March 2012 (UTC)

Ahh, a misunderstanding at our end; thanks Jorm :). I'll amend the NPT page when I'm less sleep-deprived. Okeyes (WMF) (talk) 21:45, 18 March 2012 (UTC)

If it simply won't show then there is a risk that it can't be overridden even for obvious attack pages. This makes the system worryingly gameable. ϢereSpielChequers 15:44, 30 March 2012 (UTC)

I think Jorm just said it would be indicated :). Okeyes (WMF) (talk) 19:22, 9 April 2012 (UTC)

The "lock" feature is very useful. However, I suggest you choose another name to prevent further such missunderstandings. How about "book out"? Or "take"? Or "receive"... it makes sense to say that a user has received an article for triage.

On a more general note, the Special:PendingChanges page could probably benefit from such a system (assuming PC is adopted). The edit conflicts and associated wasting of time was probably one of the things that annoyed people most during the PC trial.

On a more philosophical note, WP:STiki manages to avoid this whole issue by only showing its users one thing at a time and never showing the same thing to two users at once.

Yaris678 (talk) 17:42, 8 May 2012 (UTC)

Liking all three notes :). We'll probably debate wording slightly later on; at the moment a priority is shipping the prototype. And pending changes new features are definitely something that should be discussed with the wider community - I'm not really in a position to make calls on that. Okeyes (WMF) (talk) 17:51, 8 May 2012 (UTC)

Oh yes. I wouldn't suggest that someone makes a big change to PC without discussing it more widely first.

In terms of the terminology... I know you say it isn't a priority now... but when will it be a priority? It's one of those things that always seems small but can make a big difference. I wouldn't want to wait until there is a whole load of documentation etc that will have to be changed... because that will make changing the terminology more difficult.

Yaris678 (talk) 18:27, 8 May 2012 (UTC)

Agreed on all fronts; terminology is Important. To be honest, I'm not sure yet; we're going to get through the current sprint and then timetable the components we still have to complete :). Okeyes (WMF) (talk) 20:50, 8 May 2012 (UTC)

OK. Cool. Timetable the components we still have to complete makes sense to me. Obviously, a certain amount of flexibility on timing may be required, but it lets people see when the different bits will be discussed. Looking forward to seeing the prototype. Yaris678 (talk) 09:25, 9 May 2012 (UTC)

Another suggestion for a name, instead of "lock": "reserve". Yaris678 (talk) 12:20, 9 May 2012 (UTC)

Adding a NOINDEX tag to unpatrolled articles

Hey guys

After suggestions here, we've opened a Request for Comment on adding the NOINDEX tag to unpatrolled articles - basically ensuring they can't be syndicated by google. If you've got an opinion or any comments, head on over there and post your two cents :). Okeyes (WMF) (talk) 13:10, 20 March 2012 (UTC)

A (massive, sort of) update

Hey guys! Big update on what the developers have been working on, and what is coming up:

coding

• Fixes for the "moved pages do not show up in Special:NewPages" and "pages created from redirects do not show up in Special:NewPages" bugs have been completed and signed off on. Unfortunately we won't be able to integrate them into the existing version, but they will be worked into the Page Triage interface.
• Coding has been completed on three elements; the API for displaying metadata about the article in the "list view", the ability to keep the "patrol" button visible if you edit an article before patrolling it, and the automatic removal of deleted pages from the queue. All three are awaiting testing but otherwise complete.

All other elements are either undergoing research, or about to have development started. I appreciate this sounds like we've not got through much work, and truthfully we're a bit disappointed with it as well; we thought we'd be going at a faster pace :(. Unfortunately there seems to be some 24-72 hour bug sweeping the San Francisco office at the moment, and at one time or another we've had several devs out of it. It's kind of messed with workflow.

Stuff to look at

We've got a pair of new mockups to comment on that deal with the filtering mechanism; this is a slightly updated mockup of the list view, and this is what the filtering tab is going to look like. All thoughts, comments and suggestions welcome :). I'd also like to thank the people who came to our last two office hours sessions; the logs will be shortly available here.

I've also just heard that the first functional prototype for enwiki will be deployed mid-April! Really, really stoked to see this happening :). We're finding out if we can stick something up a bit sooner on prototype.wiki or something.

I appreciate there may be questions or suggestions where I've said "I'll find out and get back to you" and then, uh. not ;p. I sincerely apologise for that: things have been a bit hectic at this end over the last few weeks. But if you've got anything I've missed, post it here and I'll deal with it! Okeyes (WMF) (talk) 16:12, 30 March 2012 (UTC)

A request

I know that this is a minor aspect, but if we could get the Page Logs included (so as to help quickly detect "I didn't hear that" type events and cases where the article is deleted for a reason and has come back as a zombie. I know it's a truly minimal portion of the NPP workflow, but it's one of the things I check to see if there's a CSD that applies. Hasteur (talk) 18:58, 3 April 2012 (UTC)

That would be awesome :). So we're talking: user creates article at URL X. I delete it. User recreates article at URL X. [non-admin], or [admin who doesn't have OCD] comes along, and can't immediately see if anything was previously at that URL, treats it as if it were a first creation? Okeyes (WMF) (talk) 19:16, 3 April 2012 (UTC)

Exactly. In 90% of the cases the peeking at the "Page Log" is just to help firm up a "Is this suitable for WP?" feeling. It's a extra bit of information, but can help the triagers decide to treat it gently, or to take out the cluebat, or to bring out the rod of smiting. Hasteur (talk) 20:17, 3 April 2012 (UTC)

That is a really great idea. I'll email the development team now, hopefully have a response for you this sort of time tomorrow ;). Okeyes (WMF) (talk) 22:39, 3 April 2012 (UTC)

Just want to second this as A Great Idea. As an admin, whenever I check an article I look to see if there are relevant deleted revisions, it can be a tremendous help in determining whether or not to delete something (obviously for G4s, but also to see what other admins have had to say). The Blade of the Northern Lights (話して下さい) 03:16, 8 April 2012 (UTC)

Nudge Any update? Hasteur (talk) 15:19, 9 April 2012 (UTC)

Not yet, but I've got a meeting with the team this evening, so I'll bring it up if they haven't by then :). Okeyes (WMF) (talk) 15:24, 9 April 2012 (UTC)

• Looks like the answer is "sure, that'd be relatively simple, I'll include it" :). Okeyes (WMF) (talk) 21:29, 9 April 2012 (UTC)

Thanks! --joe decker^{talk to me} 20:44, 18 April 2012 (UTC)

April Debut?

Perhaps it's crossed information, but wasn't there supposed to be a first level demonstration of the new interface this month? I ask because there's only a few days left in the month. Hasteur (talk) 13:08, 29 April 2012 (UTC)

Argh, I'm an idiot and misread the page title the first time around. Yep; we've had very very minor delays to accomodate some community requests (and because I've spent hours on end playing around with the software to discover any and all rare bugs. My favourite is one that appears precisely 50 percent of the time when you take a particular action). May 2nd is looking to be the date of deployment - we've scheduled and booked it - and after a couple of days of us kicking the feature around and checking it doesn't make anything else on Wikipedia explode, I'll be running around asking everyone to try it out and give feedback. Okeyes (WMF) (talk) 15:26, 29 April 2012 (UTC)

How is the deployment test going. Edinburgh Wanderer 16:35, 3 May 2012 (UTC)

We ran into some issues when we tried to deploy last night and couldn't :(. Luckily, forewarned is forearmed, and now that we know what the problem is we've politely booted the other project I'm working on out of its deployment window today in order to get things released. It should be done by tomorrow morning - if there are further delays, I'll let people know :). Okeyes (WMF) (talk) 16:56, 3 May 2012 (UTC)

Okay, tried again, more issues! Ugh. I'm really sorry about this :(. The new plan is to try again on Monday, then conduct some internal testing, poking and prodding to make sure nothing explodes, then opening it up to the community probably the following Monday or Tuesday. Okeyes (WMF) (talk) 23:00, 5 May 2012 (UTC)

How did the Monday (May 7th) deployment go? Interested minds would love to know. Hasteur (talk) 20:53, 8 May 2012 (UTC)

It is being tested, and bugs are being worked out. Deploying to the English Wikipedia helps us see a huge array of possible problems that we simply don't see in a test environment. Works great, though, IMO.--Jorm (WMF) (talk) 22:40, 8 May 2012 (UTC)

bug in the current version of New Pages

Hi guys, just wanted to make you aware of a bug in the current version of New Pages, to ensure it doesn't happen in the new version: http://en.wikipedia.org/wiki/Wikipedia_talk:New_pages_patrol#possible_BUG_.28error.29_in_the_wikipedia_software Azylber (talk) 11:20, 3 May 2012 (UTC)

Thanks for letting us know :). I'll check on the prototype right now. Okeyes (WMF) (talk) 13:13, 3 May 2012 (UTC)

Yep, it's in the new version as well; filing a Bugzilla report. Okeyes (WMF) (talk) 13:16, 3 May 2012 (UTC)

Cool! Happy to help. Azylber (talk) 14:47, 3 May 2012 (UTC)

Thanks! Let me know if you see anything else you might like changing :). Would you like me to let you know when we have the functional prototype deployed on Wikipedia so you can give it a spin? Okeyes (WMF) (talk) 15:16, 3 May 2012 (UTC)

ok, but I'm not really into patrolling. The only reason I've found this bug is because I've written 200 articles, and I always check the NewPages page to see if they've already been patrolled. By the way, it looks like you have a "nice" backlog :p Azylber (talk) 16:16, 3 May 2012 (UTC)

Heh; okay, neat :). Okeyes (WMF) (talk) 16:24, 3 May 2012 (UTC)

Tags and Twinkle

As a developer of Twinkle, I was thinking about the possibility of somehow intertwining Twinkle's "Tag" module with the tag list in New Pages Feed's "curator" view. It seems a bit silly to have two different JavaScript tagging tools in use. Perhaps Twinkle could share the same list of tags and categories as NPF, or perhaps Twinkle could even call upon NPF's UI to present the same tagging interface within a Twinkle-style window. Just some thoughts; Twinkle is fairly popular and it would be confusing to have two competing tagging interfaces/lists of tags floating around. What do you think about the matter? — This, that, and the other (talk) 10:48, 9 May 2012 (UTC)