Wikinews:Water cooler/technical/archives/2016/December


Article creation wizard discussion

Hi all,

This proposal suggests WMF to fund the writing of an article creation wizard at Wikipedia, but with enough interest it may -- or may it not? -- be expanded to write an article creation wizards framework or library for use at non-Wikipedia wikis, such as here. If desired, please join the discussion before December 12. (I've sent this message to English wikis; I ask you to deliver it to non-English wikis, if you can. Even delivering it in English there may be better than nothing.)

  • What tools do we use here, now, to make article creation easier for newbies?
  • What requirements do we have for a potential implementation?
  • How would you like to inform the people of the article creation perks and difficulties on this wiki?
  • What else needs to be considered?

Thanks. --Gryllida 03:54, 6 December 2016 (UTC)[reply]

Article creation is one of the several major things on my list that I hope eventually to support using my dialog tools. The further into that development I get, the more I'm convinced it's impossible for a centralized organization such as the WMF to do properly what I'm doing. I don't think it's even possible to lay out a plan for this and then do it; every step along my path has been improvised, it's rarely been possible to see more than one or two steps ahead. What the Foundation end up doing instead is actively counter-productive, stuff that continuously curtails what ordinary users can do, restricting infrastructure development to an elite and especially to an elite employed by the WMF. The Foundation can't help operating like that, it follows from a potent synergy of their corporate culture with their corporate structure. --Pi zero (talk) 04:21, 6 December 2016 (UTC)[reply]
@Pi zero: I know we've talked here and at b:en: a little bit about development. Do you have a roadmap that you've posted anywhere to discuss? —Justin (koavf)TCM 06:16, 6 December 2016 (UTC)[reply]
@Koavf: I'll say a bit, below, about that next step or two ahead that I can see atm.
  • Regarding what has been written down — although I've largely chosen to devote my volunteer time and effort to actually developing tools, rather than talking about developing tools (or, even worse, trying to convince someone else to give me permission to develop tools). Besides the documentation for the tools, which I've tried to do well because I, myself, hate having to try to use software that's not well-documented.
  • I have in mind that a user can go to any page and have dialog-based context-sensitive help for things they might want to do there. I'm now crafting basic elements toward that goal: I'm well along in designing a mechanism to supply context-sensitive information to assistants so they can offer advice, and I'm marshaling my thoughts toward a mechanism for keeping track of situations where users disagree with advice (which is needed both so that the advice given doesn't just keep offering itself until somebody takes the advice, not realizing there was a reason it wasn't taken in a particular situation, and so that in the long run the community can take such exceptions into account when considering how to modify or expand advice). A basic principle in all of this is, of course, that the goal is to empower ordinary wiki users, and that it's impossible to do that with a rigidly structured system, or with a system that isn't thoroughly grounded in ordinary wiki markup.
--Pi zero (talk) 14:49, 6 December 2016 (UTC)[reply]
@Koavf: I didn't like my above account of where things are going, because it's really low-level. I haven't done enough to explain the big picture; that's always been a problem with explaining the dialog tools. It's a special art, to choose primitive tools that look rather innocent themselves but make it possible to do great things; it may be easy look at the low-level tools and not see the potential in them, and then how can you convince other people of their worth? My grand vision for these things is rather breathtaking, if only I could communicate it. As a small step, I'll try to explain a little more of why I want to help tools get context-sensitive information and track exceptions to rules.
  • A key principle of wikis is that when you're reading a wiki, if you see something that could be better, you can just fix it, without going through some big rigamarole with a bureaucracy and without having to have lots of special technical skills. (Yes, wiki markup is really amazingly easy — compared to every alternative. Making wiki markup better is productive, whereas anything that moves away from it is counter-productive.) My core idea is that, with just a few simple additions to the capabilities of wiki markup — mainly, text boxes and buttons for sending their content somewhere — it should become possible for a community of ordinary wiki users to grow semi-automated assistants ("wizards") in the same way that the community grows the wiki content itself. But, for that to become real, it has to be possible for an ordinary user, who is using a semi-automated assistant, to see that the assistant could be doing something better, and just fix it — without going through some big rigamarole with a bureaucracy and without having to have lots of special technical skills. And this is where things get really interesting, because those primitives I've added to wiki markup address the first half of that goal but not the second. That is, the primitives make it possible to just-fix-it without going through some big rigamarole with a bureaucracy, but I know from experience that it's logistically difficult to keep track of what you're doing when you're trying to built or maintain one of these semi-automated assistants. And the obvious solution to that problem is to provide semi-automated assistance for building and maintaining semi-automated assistants. That is, meta-assistants. How does one design a meta-assistant? By designing and building assistants, and thinking deeply about what one is doing during the process. I've chosen certain particular devices, that I believe can be built using the primitives, and that I believe are necessary for a wide class of assistants including meta-assistants, and now as I start to plan out the details of these devices I'm also thinking deeply about what I can learn from the process about how a meta-assistant has to work.
--Pi zero (talk) 20:03, 6 December 2016 (UTC)[reply]
@Pi zero: I agree that many features of the experiences between wikis here are not that user-friendly nor are they new-user-friendly. And some processes which should more-or-less be the same across wikis are different due to bureaucratic rules or because the same gadget/userscript/module/template isn't available cross-wiki (e.g. renaming a category at w:en: versus at c: or proposing an image for deletion, which is confusing at c: to me). I'm not sure how I can help you but I'd like to if I can. I think you do a lot of good work—particularly on this wiki and b:en:. —Justin (koavf)TCM 21:27, 6 December 2016 (UTC)[reply]

New way to edit wikitext

James Forrester (Product Manager, Editing department, Wikimedia Foundation) --19:31, 14 December 2016 (UTC)[reply]

Just curious.......

Who is this person, who's been modifying loads of templates here? --Bddpaux (talk) 21:46, 21 December 2016 (UTC)[reply]

I did!—The preceding unsigned comment was added by MIKEL S. SARWONO (talkcontribs)
That account is now globally locked. BRS (Talk) (Contribs) 20:33, 23 December 2016 (UTC)[reply]