Talk:Wikipedia article creation workflow improvements


Let's be smart[edit]

Some things we could take into account when putting people into the normal article creation process vs a sandbox/namespace safespace article creation process

  • age of account
  • number of articles created
  • number of articles deleted or marked for deletion

Part of me wonders if we should just completely switch over to this model for all users, and have this namespace/draft environment be the way that users make new articles, I think the very slight efficiency hit we'd have on power users would be far outweighed by the benefits to most users, it would also save time vs creating a bunch of logic around how and when we show users the new vs old process. We could even clean up the "move article" process to be much friendlier with two separate front ends, one which is about "publishing" and article from draft, and the other existing move article flow. We could investigate better ways of user reaching out to each other for help collaborating on articles in the draft space as well as part of this work. Jaredzimmerman (WMF) (talk) 12:48, 8 November 2013 (UTC)[reply]

There's a very easy way to limit any new system to being thrown at new people only, which is using the autoconfirmed permission. For a minimum viable product, that's what we'll probably go with, rather than trying to create a metric from scratch right away. Ideally, even if someone has passed the 10 edits/3 days threshold, they might not have created an article before, but data we have right now suggests that it's non-autoconfirmed users who have the most trouble (see the Mr.Z-man analysis I linked to in the Notes on the doc).
As for simply having experienced users just use the same process... I don't think that will fly to be honest, at least not for a while. In early discussions about a Draft namespace, for instance, it's very clear that existing users want to be able to simply skip any guides and go for direct creation. Experienced users are also calling for retaining the ability to use userspace drafts. Part of that is, I think, because they assume that a draft space will involve requiring pre-review of an article in order to publish. I think this is clearly a failed idea, even for new users. If we can prevent a draft space from operating under the requirement for peer review pre-publishing, then there's a decent chance experienced editors will give it a try. But overall, creating logic to alter article creation for new users specifically is not a big burden. Steven Walling (WMF) • talk 21:32, 8 November 2013 (UTC)[reply]
This is a good idea in theory, but I agree with Steven here. Forcing so-called power users (i.e. experienced editors) to start articles in a Draft namespace would cause a great backlash against the Foundation. These power users are comfortable with the way things are and they don't like change, especially change that makes them feel like they are being treated as lower-class users (i.e. new users). Power users like the status they have earned and would react negatively to a change like this.
The other suggestions are very good and I think the enwp community would accept them. In particular, updating the "move process", especially the "publishing" part. I think this would encorage new users to improve their articles in order to "publish" them to article space. 03:37, 9 November 2013 (UTC)[reply]
I also agree with Steven. I don't see a benefit for requiring experienced article creators to use the new process. I also think auto-confirmed is fine as a transition point for now, even though it isn't perfect. Superm401 - Talk 22:15, 19 November 2013 (UTC)[reply]

Draft user stories[edit]

Here's a few...

"As a user who has written my first article I want to ask for help proofreading it so it fits with the standards of wikipedia"

"As a user who has written my first article I want to ask for help co-writing an article so I can better meet other editors, and become part of the community"

"As a user who has written my first article I want to ask for help from other users who like to contribute in other ways such as reference searching, or photo uploading/finding so that we can all play to our strengths and interests"

"As a user who has written many articles I want to make myself available to new editors who are interested in writing articles in my areas of interest so that I can help or mentor them"

Review before publication[edit]

I think #2 and #3 should be clearer with regard to the actual publication process. I define "publish" as "put it in namespace 0". You said above, "If we can prevent a draft space from operating under the requirement for peer review pre-publishing, then there's a decent chance experienced editors will give it a try." If we don't have per-publication review, is the point just to have a Draft namespace and easy tools for optional review? Would new users be able to publish pages themselves (if not, there essentially still is pre-publication review)? Superm401 - Talk 22:19, 19 November 2013 (UTC)[reply]

The short answer to your latter two questions: Yes, that's what I think we should test first. I don't even think we should call it "review" at all in fact. It should be something like "ask for help" or "invite collaborators". The focus should be on making sure that drafts aren't a ghetto we use to police newbies, but that they're more in tune with the normal collaborative cycle of Wikipedia. It might not work; we might still get the most clueless publishing in mainspace prematurely, with the smartest and most cautious lingering in draft status until they expire. But it seems clear to me that required pre-review isn't working right now, and we should give the alternative an honest try. Also, technically speaking review is not well-enforced, even if that's what we want. All a user has to do is wait to be autoconfirmed then they could bypass AfC. I don't even think there's a policy that requires them to follow AfC review steps, it just seems that way because newbies don't know how to navigate the system. Steven Walling (WMF) • talk 22:36, 19 November 2013 (UTC)[reply]

Other languages and wikis[edit]

I'm alright with focusing on Wikipedia first. Some other Wikimedia wikis, such as Wikinews, have explicit flows already that may work better than AFC. Others may not have as much of an issue with problematic articles by new users. Only a few wikis block anonymous page creation, but if our flow is useful, it should help even people who could publish pages themselves.

However, I would like to prioritize doing additional Wikipedias (in addition to English), either in the initial launch or shortly thereafter. Superm401 - Talk 22:24, 19 November 2013 (UTC)[reply]

I am definitely not opposed to doing non-Wikipedias, but I want us to try and not bite off more than we can chew at first. The other big caveat we saw with account creation was that it's hard to stage things. If you work in core like that, you have to serve all use cases (including non-Wikimedia). This made it complicated to fix Wikipedia-specific UX problems like letting users know about appropriate usernames. I can only see the issue with general page creation in MediaWiki content namespaces being an even bigger issue, though it depends on A) which part of the workflow we're talking about (i.e. landing pages vs. Draft namespace) B) how we implement things. In general, I think we'll be able to do some really interesting and valuable things if we focus on encyclopedia article creation first. Ideally I envision us doing something extension-based at first, in order to avoid mucking with things outside the stated scope. Steven Walling (WMF) • talk 22:43, 19 November 2013 (UTC)[reply]