Latest comment: 2 days ago268 comments24 people in discussion
Back in March, the closers floated a suggestion in their closing statement that a third RfC may be held in six months' time if there remains a sentiment among editors to restore Vector 2010 as the default skin. Six months from the closing date is September 16 — less than two weeks away from now. Is there still interest in having such an RfC? InfiniteNexus (talk) 16:08, 5 September 2023 (UTC)Reply
By imposing Vector 2022 regardless of the community's wishes, the WMF has presented us with a fait accompli. Another RfC now would probably conclude reluctantly that changing back would be too disruptive and we are stuck with Vector 2022. Such a result might be spun as acceptance or even approval of the new skin. Certes (talk) 16:48, 5 September 2023 (UTC)Reply
No; it's not copied or pasted from anywhere. However, I will continue to consider replying where relevant, even in a debate which revisits a topic discussed before. Certes (talk) 11:10, 8 September 2023 (UTC)Reply
Another RfC now would probably conclude reluctantly that changing back would be too disruptive and we are stuck with Vector 2022. When the WMF creatively justified their refusal to force unlimited width as the default by claiming it would be too jarring a change, the irony cracked me up. Did the WMF not think that the overnight change from V10 to V22 was going to be jarring as well? I personally only found out a mere hours before launch, and I remember going to bed thinking, Oh boy, tomorrow's going to be crazy. I was right. InfiniteNexus (talk) 16:14, 8 September 2023 (UTC)Reply
Support: As far as I know, there are other Wikipedias which have been successful in the rejection of V22. Even seven months after the deployment, I continue to think that it is hideous, ill-thought-out, that the user experience is markedly worse compared to V10, that splitting the article between a left side with that bunch of messy lines called "TOC" and the a right side with that cramped wall of text, while hiding the user menus, has been an awful idea, a "change for change", a change for its own sake. Other users who were interested should be pinged. Æo (talk) 21:44, 6 September 2023 (UTC)Reply
Addendum: You should have noticed the sudden appearance of "too long" tags on a large number of articles which before V22 were perceived as having no length problems, urging their splitting into smaller articles. This is obviously a side effect of the new layout. Æo (talk) 22:00, 6 September 2023 (UTC)Reply
The mass addition of the tags was the work of a single user, Nikkimaria (talk·contribs), who added tags to articles that were substantially above the 10,000 word limit recommended by Wikipedia:Article_size. I don't Nikkimaria's tagging campaign was primarily motivated by the adoption of Vector 22 (though I will let them speak for themselves). Hemiauchenia (talk) 22:09, 6 September 2023 (UTC)Reply
Consider an alternative. While V22 kinda sucks, V10 actually isn't flawless, believe it or not. I'd think it'd be much more productive to put editor energy into a community-driven, NOT WMF-driven skin, like what Timeless was intended to be. This hypothetical skin would be able to solve the problems that do exist in V10 without creating the new problems that exist in V22, which the WMF seems reluctant to address. Doing this (i.e. presenting an alternative that actually solves more problems than it creates) would result in an RfC much, much more likely to pass, even if it'd be a few years down the line from now - and it allows the community to address fundamental questions like "is having one skin for both mobile and desktop ACTUALLY a good thing?" that have pre-baked answers fundamental to the entire concept of V22 (which argues "yes") which no RfC could possibly overturn.
I simply know nothing about what it takes to design a MediaWiki theme, but someone who does could concievably either revive Timeless or start from scratch on a new skin. I think both options would be better than reverting to V10, in the long term, and in the short term it simply is a time-sink that just might not be worth it to hold another RfC that is just as likely to attract walls of text as the first one.
If you insist on holding another RfC, I think it should be held to a stricter standard of participation than normal. (Perhaps confirmed or XC might work, but I'm thinking more in terms of something like a strict requirement that comments are constructive and moving all "V22 sux!!!!" comments to the talk page liberally would help a lot.) casualdejekyll22:32, 6 September 2023 (UTC)Reply
I might agree. Here I highlighted how one of the earlier versions of V22 might be a good compromise between classical structure and innovation, and that as such it might be welcomed even by those who don't like V22. Æo (talk) 15:13, 7 September 2023 (UTC)Reply
Certainly, require comments to be constructive. However, this question is unusual in that it mainly affects readers rather than editors (who can set a skin preference). Comments from genuine readers who rarely edit are therefore particularly welcome, though there may be practical difficulties in distinguishing them from socks. Certes (talk) 16:11, 7 September 2023 (UTC)Reply
I feel any further discussion would be unproductive and would waste the time of everyone involved (as someone who was an active participant in both RfCs). — Qwerfjkltalk17:07, 7 September 2023 (UTC)Reply
Neutral. The concern over the UI change appears to have dwindled, if the number of questions being asked at the Help Desk and Teahouse are anything to go by, so I'm not sure there's anything to gain. Seems new users aren't thrown off by the interface, and older ones have either gotten used to it or switched to a more preferred skin. I personally am glad with the changes they made to my concern about the table of contents. —Tenryuu 🐲 ( 💬 • 📝 ) 17:25, 7 September 2023 (UTC)Reply
Support, but it's pointless Community consensus was ignored twice, and then a six-month time limit was imposed to cool things down. There's no interest in this now. I support imposing prior community consensus, but it won't happen. ToaNidhiki0518:27, 7 September 2023 (UTC)Reply
@Toa Nidhiki05: I wouldn't be so pessimistic. There are some important Wikipedias, the German and Italian for example, which have successfully halted the deployment of V22 onto their projects. Æo (talk) 19:44, 7 September 2023 (UTC)Reply
Do you really not have anything better to do with your life? The majority of your edits over the last year have been dedicated to overturning Vector 22, despite the fact that as an advanced user, you can easily revert back to the old version. You are the literal definition of "vocal minority". Hemiauchenia (talk) 20:30, 7 September 2023 (UTC)Reply
And here I was thinking we based closures of discussions off the strength of the arguments, rather than number. There simply isn't consensus to do anything to change the current situation. Another RfC is hardly going to change that, much as you want to pin it on the WMF's authoritarian control. That is to say: community consensus was not ignored because there was no consensus. — Qwerfjkltalk20:51, 7 September 2023 (UTC)Reply
Yes, most of us have given up, realising that the WMF will just ignore our wishes as usual. However, the fact that those who bother to call out this abuse of privilege are a minority doesn't make them wrong. Certes (talk) 22:00, 7 September 2023 (UTC)Reply
@Hemiauchenia: Over the last year I have been very little on Wikipedia as I have had little free time to dedicate to it, and I have dedicated mosta lot of this free time to following the V22 case. Æo (talk) 20:50, 7 September 2023 (UTC) — "A lot", not "most" time, actually, as I have actively contributed to some articles and other discussions as well (I forgot about it).--Æo (talk) 20:57, 7 September 2023 (UTC)Reply
Support. (I'd already seen this mentioned at a criticism forum before the ping.) Despite the point Certes makes about a possibility of spin if the RfC is defeated (the WMF will always look to its PR image, whatever we do or don't do) and despite there having been some much overdue changes since V22 was deployed as the new default, I endorse discussing it again. The main reason being that, as Certes also reminds us, the change primarily impacted readers, who don't have the option of changing their settings to a different skin. We have a duty to consider their needs; in particular, are there any changes or reversions that are urgent but the WMF has not yet made? Beyond the question asked, I am intrigued by Casualdejekyll's suggestion that we look into community development of an alternative. I'd forgotten or not been aware that Timeless was such a project. However, quite apart from the fact that it's pretty deplorable that the WMF pays developers and represents itself as a software developer, but the most needed and useful tools continue to be the work of volunteer editors, the key issue is that V22 is the default skin so that's what readers are forced to use, not just newly registered editors unaware of alternatives; we could develop a game-changingly excellent skin, we could even promote it to the communities of other-language Wikipedias, but the WMF still decides what the readers see. And that's in addition to the problem that there are differences in both needs and preferences depending on the type of device used, user generation/expectations (I'm still using Monobook, as I think are many longer-term Wikipedians, and I have a "user hates Vector" userbox from when V10 came in as default; I detest things being hidden behind drop-down menus, but to others that's "clean" and "modern", and I like the background frame of an open book, which is probably "oldfashioned" and "fussy" in the view of many), and accessibility issues (for some readers, black text on white is necessary contrast; for others, it's physically painful). So I think it would be wasted effort, sadly, although revising Timeless might help some editors who dislike both Vector and Monobook. My number one ask of the WMF in terms of site appearance is considerably more basic: to let non-logged-in users switch to desktop view as a persistent setting. Yngvadottir (talk) 00:07, 8 September 2023 (UTC)Reply
If there is to be another RfC, instead of a yes-or-no question like last time, we should present the community with a list of options for what to do. Also, Yngvadottir, could you provide the link to the criticism forum you mentioned, in the interest of transparency? InfiniteNexus (talk) 00:29, 8 September 2023 (UTC)Reply
@InfiniteNexus, I honestly believe that any other closure would be inescapable, given the nature of what we're discussing, no one will ever totally agree. As such, I think another RfC would be a waste of time. WP:STICK. — Qwerfjkltalk16:25, 8 September 2023 (UTC)Reply
I'm still undecided as to whether a new RfC is a good idea. But it's important to realize that if we choose not to this, this will be the end of it, and there's no going back. InfiniteNexus (talk) 19:08, 8 September 2023 (UTC)Reply
@Æo, I doubt that will in any way be an improvement. You mgiht get more people airing their greivances, but when it comes to consensus, this whole thing is an ILIKEIT/IDONTLIKEIT nightmare. — Qwerfjkltalk15:40, 11 September 2023 (UTC)Reply
Support. The only reason why I have this account is because reading and editing as an IP had become horrific. When I have to read/edit logged out, I cringe at how awful the user experience is. The fundamental problem remains: a mobile design for a desktop UX is by definition a bad design. The only practical solution here is to hire principal designers that have real experience designing desktop UX. I don't think an RfC is the way to get those designers hired, but it's better than nothing. TheMissingMuse (talk) 04:04, 9 September 2023 (UTC)Reply
@TheMissingMuse, you said The only practical solution here is to hire principal designers that have real experience designing desktop UX. That implies that the current ones don't have any real experience, which isn't something you know for certain (unless you have access to some information I'm unaware of). The website appearance is a controversial matter - there's not one "right" way to do it. Saying that you're a professional developer with extensive experience in mobile and desktop design doesn't mean that your opinion should be given extra weight. I can see why people might dislike Vector 2022. But there isn't any easy solution here, and claiming that either way is right or wrong doesn't help. — Qwerfjkltalk16:43, 14 September 2023 (UTC)Reply
As brought up multiple times in the RfC, the design isn't meant for mobile. While its simplicity is mobile-esque, the design is very clearly meant for desktop. I just tried to use V22 on my Iphone X and... no.
It's just a change in design trends—staring in mid-late 2010's, skeuomorphism began to be traded for flatter designs, less colors, and overall simplification. Like, look at Mac OS X Snow Leopard and compare that to MacOS Big Sur. It feels like a big change because it is the only real change to the MediaWiki interface in twelve years.
When you say: "the design isn't meant for mobile ... its simplicity is mobile-esque" you are simply restating the problem I laid out. A desktop UX should be "desktop-esque". The UX is bad precisely because it is, as you say: mobile-esque. TheMissingMuse (talk) 22:58, 11 September 2023 (UTC)Reply
I'm not sure what you are referring to here. I'm a professional developer with extensive experience in mobile and desktop design. The V22 theme is, as Cessaune says, "mobile-esque", and is thus a bad design for desktop users. TheMissingMuse (talk) 21:06, 12 September 2023 (UTC)Reply
All Cessaune said was mobile-esque is "simplicity" which admittingly is very broad but I think all of the aesthetics can fall into flat design. If you only compare the functionality the only regression is just some more dropdowns and the only dropdown that can't be un-dropdown'd is your user dropdown, which isn't something I use that much anyway and something a ton of platforms use on desktop today e.g. Google and almost all Microsoft sites. Aaron Liu (talk) 23:24, 12 September 2023 (UTC)Reply
A mobile design is not "simplicity". A mobile design is largely about conforming to a vertical viewport, linear flow, maximizing content, and minimizing controls. There are a number of other aspects as well related to controls, but when it comes to V22, they have certainly taken the mobile design route for presentation. TheMissingMuse (talk) 16:13, 14 September 2023 (UTC)Reply
How does v22 conform to a vertical viewport? How do v10, monobook, timeless and other skins have non-linear flow? Icons and drop-down are also commonplace in desktop design, just that mobile design does it doesn’t mean only mobile foes it Aaron Liu (talk) 16:29, 14 September 2023 (UTC)Reply
I'm not here to debate the finer points of mobile vs. desktop design with you. That's a discussion for the internal design team to have amongst themselves. TheMissingMuse (talk) 17:01, 15 September 2023 (UTC)Reply
Find an alternative. No one wants to talk about this anymore. The broader community doesn't need or want to waste anymore time on this non-issue. I call it a non-issue because, by now, people have either stopped using Wikipedia, started viewing it from a third-party website, created an account and changed the skin, or gotten used to V22. Time an energy should be spent in improving V22 (what the hell happened to the border thing? We were so close to getting something implemented) and not another RfC. Cessaune[talk]01:53, 10 September 2023 (UTC)Reply
Improving V22 does not seem to be feasible, however. The project as a whole, and Zebra in particular, seem to have been abandoned by the Web team (read the latest discussions here). On the other hand, as I wrote above, some Wikipedias including the German and Italian ones seem to have successfully stopped the imposition of V22 onto their respective projects. Is anyone aware of how they did it? Æo (talk) 10:37, 10 September 2023 (UTC)Reply
I don't think admins had any say in the matter. We just woke up to find Wikipedia was different and, in some people's opinions, broken. Certes (talk) 11:52, 11 September 2023 (UTC)Reply
A final possibility - people enjoying the new version and from the onset, as many who use Wikipedia for editing have stated. Regarding consensus, we don't know how many in the WMF and other internal operators of Wikipedia also felt comfortable with this change; knowing these variables is also important in understanding how global the desire for this change was, from all internal organs of Wikipedia (i.e. developers, editors, and other people heavily vested in Wikipedia (esp. English in this case) and its parent foundation also placed much energy into this, especially in preparing its implementation and what continues to follow). Being unaware of this answer does not take away from its truth. Radlrb (talk) 21:50, 11 September 2023 (UTC)Reply
Support for a six-month assessment of editor's experience compared with the creators expectations. Maybe an alternate proposal to change back to Monobook as the default, but then WMF may just mess with that too. Remember when WMF was originated to fund and maintain Wikipedia and not to dictate to it? The good ole days. Randy Kryn (talk) 15:27, 11 September 2023 (UTC)Reply
Just a dream of a better skin, not really a serious suggestion. But for those who really dislike Vector I'd suggest a week's trial of Monobook at 125% (takes a few days for the brain the adjust to change) as giving yourself an early Christmas gift. Randy Kryn (talk) 15:37, 11 September 2023 (UTC)Reply
Not sarcasm, but I do know that the default was changed from Monobook many moons ago and probably the vast majority of editors and readers are unfamiliar with its quality. If everyone would try it out for a few days Wikipedia might go back to it, but they won't try, or go back. But yes, a six-month review of the new skin is a good idea for feedback purposes and to see what the coding team has worked on. Randy Kryn (talk) 22:46, 11 September 2023 (UTC)Reply
The fact that we can't even arrive at a clear consensus on whether it's worth having another RfC is not a good sign for whether we'll be able to arrive at a clear consensus in a hypothetical RfC. InfiniteNexus (talk) 22:09, 11 September 2023 (UTC)Reply
I lean towards Support as there has not been any updates since May and even I who like the aesthetic am uncomfortable supporting an undeveloped skin.
Edit: they just updated. Turns out they were developing some stuff but wasn’t communicated. I still lean support as the community can evaluate on the WMF’s promises. Aaron Liu (talk) 12:48, 12 September 2023 (UTC)Reply
In my opinion, there is still the need for a six-months community assessment. Those updates do not mark that great difference, after all; the V22 structure is the same. As I wrote here, if they put the TOC back into the article (becoming active as that "Contents" tab when the TOC is scrolled off-screen), put the user menus back in the left sidebar, and increase the width of the article's body, just as in that early prototype, the quality of V22 would increase greatly and much of the community's concerns would be addressed. Æo (talk) 16:29, 13 September 2023 (UTC)Reply
I don’t see why you want the page tools menu to be on the left sidebar and why the toggle doesn’t address your needs on article width. I agree that there should be options for the rest. Aaron Liu (talk) 16:33, 13 September 2023 (UTC)Reply
The hidden page tools menu is greatly impractical. At first it was one of the changes that I appreciated in V22, but with time I have realised that it is not an improvement at all. Æo (talk) 16:38, 13 September 2023 (UTC)Reply
In my view, it is more "logical" to have it in the left sidebar, where it would take the place of that unwieldy and useless nuisance which is the new sticky TOC. Æo (talk) 16:48, 13 September 2023 (UTC)Reply
1. To me it's more logical to have a separation between actions on the current page and "global" links. The only link that doesn't make sense is "Special pages". 2. I don't see where you have argued that the floating itself is a nuisance, not the absence of an inline one. 3. If floating, the contents plus the page tools plus the userscript buttons would be way too long to fit on a couple screens, especially if you expand all. But sure, it could be an option, just probably not the default one Aaron Liu (talk) 16:55, 13 September 2023 (UTC)Reply
2019 prototype
1. I don't fully understand what you mean but I agree on the separation between actions and global links. This is, however, a matter of internal organisation of the menu itself.
2. In my past comments I widely criticised it for having lost all the "article structuring" functions of the classical TOC; in addition, it is a nuisance, for it is never correctly aligned with the article itself. It is just an eyesore.
3. If the page tools menu is put back in the left sidebar and floats, and the classical TOC is restored, the new sticky TOC might be turned into that openable "Contents" menu in the upper right corner of the article (see the 2019 prototype). Æo (talk) 18:45, 13 September 2023 (UTC)Reply
For two, all I could read from your criticism is against how the inline one was removed, but I never understood your geometry. Could you elaborate on why inline + floating TOC is bad? Also, what advantages other than replacing panes + shadow with drop-downs does that prototype have over Minerva? Aaron Liu (talk) 19:25, 13 September 2023 (UTC)Reply
I am not against inline + floating TOC (the floating one appearing when the inline one is off-screen). I think that the inline one should remain the main one, and the floating one should be made accessory, and in any case redesigned to be made more functional. Æo (talk) 13:32, 15 September 2023 (UTC)Reply
Comment — what it seems like to me is that we have strayed from the "restore Vector 2010 as the default" RfC and have moved on to a "has WMF followed through with promises" RfC. Cessaune[talk]16:54, 13 September 2023 (UTC)Reply
Therefore, we should hold an RfC on whether or not WMF has followed through with their promises before we even entertain another "make V2010 the default again" RfC. Cessaune[talk]16:58, 13 September 2023 (UTC)Reply
If the consensus was that WMF had not followed through with their promises, us editors would create a new proposal that states enwiki's consensus-generated intention to revert back to 2010 at some specified date, with an accompanying document stating all the improvements that would have to be made to V22 in order to avoid rollback (and other, non-neccesary but wanted improvements), as decided on by the inevitable discusion that the RfC would generate. Cessaune[talk]17:13, 13 September 2023 (UTC)Reply
I know, I'm saying if there is consensus to roll back in a hypothetical third RfC, and the WMF refuses, could we force the change like we did with Visual Editor way back when? InfiniteNexus (talk) 04:26, 15 September 2023 (UTC)Reply
The idea was floated in the second RfC. I'd imagine that it would be possible. But, based off the precedent of other Wikipedias, I see no real reason why WMF wouldn't just rollback if that was the consensus. Cessaune[talk]04:28, 15 September 2023 (UTC)Reply
The idea of "unlimited width as the default" was, in my mind, a very small issue that was blown away out of proportion by people's hatred for V22. The real issue was the lack of persistence, which was eventually implemented. Cessaune[talk]04:42, 15 September 2023 (UTC)Reply
Oh, no, limited width was and still is my biggest problem with V22. This, alongside the removal of the TOC, continues to cause problems with articles that were "designed" (in terms of placement of tables and images, etc.) with unlimited width and a TOC below the lead in mind. InfiniteNexus (talk) 04:45, 15 September 2023 (UTC)Reply
I guess I'm misinterpreting the meaning of "unlimited width" here. I was talking about WMF's 'wider width' solution, I guess. I do agree that the width is the biggest problem with V22. Cessaune[talk]04:52, 15 September 2023 (UTC)Reply
The RfC had an alternate proposal to make unlimited width the default; the closers concluded that there was consensus to implement that, and yet the WMF refused because research and blah blah blah, whatever. InfiniteNexus (talk) 04:55, 15 September 2023 (UTC)Reply
They never explained that, but I think that the logic behind it is that the English Wikipedia is their leading project, and a defeat of V22 here would mean a defeat of V22 throughout the entire Wikimedia network. Æo (talk) 13:38, 15 September 2023 (UTC)Reply
Again, as I said above, no one wants to waste their time on another massive RfC of that type. I seriously doubt a second rollback RfC is the way to go. Cessaune[talk]11:55, 15 September 2023 (UTC)Reply
I think that the best idea would be to have a RfC with multiple questions: 1. Q: What are your thoughts about V22? A: Good, Bad; 2. Q: Do you prefer V10 or V22? A: V10, V22; 3. Q: Has the WMF followed its promises on Vector 2022? A: Yes, No; 4. Q: Should V10 be restored as default? A: Yes, No; et al. Æo (talk) 13:30, 15 September 2023 (UTC)Reply
1 and 2 are basically the same, 2 and 4 are basically the same, and only 3 and 4 might be worth an RfC. Remember that these are time consuming, broadening the scope too much would just make something like a trainwreck. Aaron Liu (talk) 14:31, 15 September 2023 (UTC)Reply
Sort of support a new RFC but I think it should be a quality review on how to make the skin better, not demands to change the skin back. WP:ILIKEIT and WP:IDONTLIKEIT without any further explanation as to why they do or don't like it are arguments without arguments. We can manage the content and give recommendations as to how to make the skin better, which is the whole point of these system configuration change RfCs, rather than outright saying that a skin is awful without having even played around with it. Aasim - Herrscher of Wikis ❄️ 15:53, 15 September 2023 (UTC)Reply
Sorry for the delay, I was outside the country. It seems unlikely to me that an RfC to rollback would pass at this point—too many would vote against it as tendetious relitigation or fait accompli—although I would vote support. More likely to pass, and something I would also support, is a resolution pointing out all the things we needed that they've failed to provide and denouncing them—probably in weaker language than that—for not caring about us and our concerns. —Compassionate727(T·C)08:50, 18 September 2023 (UTC)Reply
Can you stop with these pointless replies? The "dead horse" thing is just an idiom and a WP:ESSAY. You already made your position very clear from the beginning; you like V22. This discussion, however, is dedicated to those who are dissatisfied with it and to a third RfC which was already planned six months ago. A majority of those who have answered here seem to be still dissatisfied with V22 and in favour of a third RfC. Æo (talk) 14:22, 22 September 2023 (UTC)Reply
is dedicated to those who are dissatisfied with it I disagree, you do not gatekeep Wikipedia discussions. I believe it’s for anyone interested in V22RFC. The “six months” thing was also just an argument and not anything set in stone. I do agree that Qwerfjkl is bordering on bludgeoning though. Aaron Liu (talk) 17:56, 22 September 2023 (UTC)Reply
That is my opinion regardless of whether or not I prefer V22. This whole thing has been discussed to death and then some. A majority of those who have answered here is meaningless; a grand total of 18 users have commented in this thread. I understand that you have issues with V22, and those issues are genuine. I just don't think further discussion on this matter will be constructive. There probably is a way forawrd, but I don't think an RfC is it. That said, I have nothing more to say about this matter so I will be withdrawing from this discussion. — Qwerfjkltalk09:52, 23 September 2023 (UTC)Reply
I continue to be ambivalent, but as I said above, even this discussion is split between those believing an RfC is pointless and a waste of time, those who still want an RfC to force the WMF to roll back V22, and those who want to have some sort of opinion poll instead. Usually, when editors are unable to come to a consensus, we do nothing, but I certainly won't try to stop those who want an RfC no matter what. I, like many in this thread, still wish to see V10 brought back as the default, or at least have unlimited width and the old TOC restored, but I'm increasingly convinced that another rollback-or-not RfC will likely end up with no consensus. And if there were consensus, would the WMF comply? If they won't comply, can we force the change on our end? Will the WMF plead CONEXCEPT again? As for an opinion poll, would that actually bring change? Will the WMF actually be willing to listen to the community's demands? InfiniteNexus (talk) 21:46, 23 September 2023 (UTC)Reply
The WMF has no qualms about having acted against consensus, because it knows what is best for the naughty children who play at editing. Its fingers are firmly in its ears. It will listen to nothing short of an all-out editors' strike, which isn't going to happen, at least not on this matter. Certes (talk) 22:15, 23 September 2023 (UTC)Reply
I would be myself in favour of a compromise between the new V22 and characteristics of V10, such as unlimited width and inline TOC. We should just decide how to formulate the RfC. Æo (talk) 06:26, 25 September 2023 (UTC)Reply
We're about to open a number of RfC's on our relationship with the Wikimedia Foundation; I would suggest waiting until at least after those discussions are opened till open this one.
No, they haven't; my plan is to open them early next week. It might be a good idea - it might drive turnout to the WMF RfC's - but I am leaning against it for five reasons:
First, the "theme" of the WMF RfC's is the relationship between the English Wikipedia and the Wikimedia Foundation; how communication between us should work, whether we are their partners or subordindates, and how much influence the community should have over the operation of the Foundation. While any Vector2022 RfC would rely on that relationship it isn't seeking to change that relationship, and I believe the WMF RfC's will go smoother, and be easier for interested editors to participate in, if the theme holds.
Second, even if there is a consensus to rollback to Vector2010 we cannot implement such a consensus ourselves; instead, we must convince the WMF to do it for us. This comes down to the relationship between the WMF and Enwiki, and I believe that the changes to that relationship that would result from a consensus in the WMF RfC's will make it easier for us to do so.
Third, I want to spread out the potential blows we issue to the Foundation. If there is a consensus in the WMF RfC's we would be telling them that they no longer have complete control over our site, and that we are requesting they make significant changes to how their organization distributes money. I am hoping we can mollify that by saying we are trying to address their concerns about hostility to their staff on this site, but I still feel it would be a good idea to give them a little time to adjust to this change before the possibility of us telling them that we are also rejecting the new skin that they poured so many resources in.
Fourth, I suspect the second Vector2022 RfC will get just as heated as the first one; my hope with the WMF RfC's is that they will proceed much like the banners RfC, where the discussion was calm and with minimal internal disagreement, and I am concerned that linking the Vector2022 RfC with the WMF RfC's will result in the heat spilling over and in doing so make it harder to form a consensus that is too strong for the WMF to ignore.
Fifth, again given how heated I expect the Vector2022 RfC to become, I think it will be a good idea to have the instructions on civility towards WMF employees in place for it; it will have the dual benefits of allowing us to test whether those instructions are fit for purpose and encouraging the WMF employees to participate.
Per WP:CONEXEMPT, if there is a decision made by WMF or developers out of our control, we can still come to consensus to formally disapprove. The last RfC did not have that, although there was rough consensus about unlimited width. I am also noting that the skin probably has evolved quite a bit in just a few months. Seriously - it went from a skin that was mostly unpolished and littered with sandwich and similar issues to one that is even more prime for readers. I did give a suggestion in WP:VPIL for an RfC about enabling responsive mode on Vector 2022 (would be opt out, probably) to unify the interface. Aasim - Herrscher of Wikis ❄️ 15:05, 30 September 2023 (UTC)Reply
It depends on what the RfC is about. 'Should we rollback to V10?' No. 'What is your opinion on V22?' Maybe. 'Has WMF followed through with promises pertaining to V22?' Sure. Cessaune[talk]16:17, 30 September 2023 (UTC)Reply
@Æo, I believe I've made my opinion on the matter clear (and I did say I had withdrawn from this discussion, so please don't ping me next time). That said, while I'm here, I'll give my opinion on this matter. First off, this is distinct from the relationship with the WMF, so it would probably better off as a separate RfC. On the matter of how the RfC would be worded, I would a) echo Cessaune's idea above, and b) suggest "How could the Vector 2022 skin be improved?" — Qwerfjkltalk20:46, 30 September 2023 (UTC)Reply
No, I'd give Nexus the credit. I was making minor corrections to his draft which was almost entirely prepared by him when HAL333 opened a very half-baked RfC at the Village Pump. Later, consensus formed to move the half-baked village pump section to merge with Nexus's draft page. Aaron Liu (talk) 17:37, 22 November 2023 (UTC)Reply
It is still the direct successor to V22RFC2, of a slightly different nature and with multiple questions, but still a sequel in the same vein. Æo (talk) 17:56, 22 November 2023 (UTC)Reply
Before we start any such RfC, I would like to clarify my opinion on the questions being asked.
What are the primary features that you are looking for when selecting a skin?
This is a good question.
What problems have stopped or discouraged you from using Vector 2022?
I think this question will alienate a lot of people who actually like V22.
What features or ideas in Vector 2022 look promising?
I don't like the term "look promising", but this is a good question.
What features or ideas in Vector 2010 do you want back?
This is an mediocre question IMO. Why does it have to pertain to V10 specifically? Why can't it just ask about features that people want in general? Why does it assume that there are features in V10 that people want back?
Is there anything you'd wish to change to ensure the smooth and efficient rollout of future skins and updates to skins?
This is just off-topic IMO. I would ask this question in a separate RfC.
I think that if we ask these questions, cries of drop the stick! and bludgeoning! will quickly overwhelm any productive discussion. These questions are too specific to the V10/V22 debate. At least in my mind, the V10/V22 debate is over, so the questions being asked should reflect that.
Here are the questions I would ask.
What do you like about Vector 2022?
What do you dislike about Vector 2022?
What features, if any, would you like to be added to/removed from Vector 2022?
"...cries of drop the stick! and bludgeoning! will quickly overwhelm any productive discussion" also applies to the background section IMO, which should, in my mind, be reworded so that there is no mention of V10 or the V10/V22 debate. Cessaune[talk]17:25, 22 November 2023 (UTC)Reply
V10 is the official predecessor to V22, so that comparison with it is inevitable, in my opinion. Also, let me repeat, it should be pointed out that the deployment of V22 has not been universally successful; some Wikipedias have successfully managed to keep V10 as their default interface. Æo (talk) 18:08, 22 November 2023 (UTC)Reply
"...that comparison with it is inevitable"—maybe in a generic sense, sure, but the sentence in the Background section—While there have been countless initiatives by the WMF and the community to progress Vector 2022, there is still opposition by some readers and editors to the Vector 2022 skin—is, in my mind, not that relevant to the actual RfC.
"the deployment of V22 has not been universally successful"—yes, but for the purposes of enwiki, the V10/V22 discussion is over and has been for a while. Cessaune[talk]18:13, 22 November 2023 (UTC)Reply
I think the the first question should also be preserved and not just replaced by a normal "suggestion" question. It gives a bit more insight on how users evaluate stuff instead of just getting the results. Aaron Liu (talk) 17:55, 22 November 2023 (UTC)Reply
@Cessaune, I might tentatively suggest the third question be split into separate questions, to avoid e.g. a pile on of people who don't like the floating ToC covering potential improvements. Just a thought. — Qwerfjkltalk18:43, 22 November 2023 (UTC)Reply
In the spirit of WP:OWN, I'm not going to take credit for anything, nor is that relevant to who should start this new RfC. And it doesn't really matter anyway who starts the new RfC, as long as it is started. InfiniteNexus (talk) 01:48, 24 November 2023 (UTC)Reply
Not for me, thanks. The whole affair has been an unfortunate shambles, but I think we must just roll over and accept that the WMF have got away with abusing our trust. They are sure that they know what is best for us, and I doubt that even an all-out editors' strike would change their prejudices now. Certes (talk) 16:32, 3 December 2023 (UTC)Reply
I can and will do it by Monday.
What do you like about Vector 2022?
What do you dislike about Vector 2022?
What features, if any, would you like to be added to Vector 2022?
What features, if any, would you like to be removed from Vector 2022?
Actually, to be fair, I would wait until Jan 2024 to start it, to avoid the inevitable drop off in editing that lasts from like December 20th to January 5th. Cessaune[talk]18:22, 3 December 2023 (UTC)Reply
Good point. I thought instead that during the Christmas and New Year period contributions would increase since people have more free time. Æo (talk) 19:53, 3 December 2023 (UTC)Reply
I would still include the question What features or ideas in Vector 2010 do you want back? from @Awesome Aasim's draft. It seems reasonable to keep a reference to V10 as the foregoing official interface and therefore as the main point of comparison, possibly still used by many (including me). Æo (talk) 19:51, 3 December 2023 (UTC)Reply
You are right, it is already covered by question 3, which, however, as pointed out by InfiniteNexus below, is redundant just like question 4. Æo (talk) 18:23, 26 December 2023 (UTC)Reply
The questions aren't redundant.
For the sake of example, let's say I don't like the sticky TOC. That doesn't mean I want to remove it, though. Even if I absolutely hated it, maybe I would advocate for a button in my prefs that can switch between 'Classic' and 'Sticky' TOC modes. The fact that I don't like the sticky TOC is separate from any desire to fully remove it.
New questions:
What do you like about Vector 2022? Combats against, for example, hordes of people saying that they dislike the sticky TOC. Without such a question, tons of people saying that they dislike the sticky TOC will probably be interpreted as a general dislike of the the sticky TOC. We wouldn't really have enough data to make that point without this question.
What do you dislike about Vector 2022? Clear, simple, and easy way to gauge community imput, alongside the preceding question.
What features, if any, would you like to be added to Vector 2022? People advocated for many different features in V22RFC2. Easy way to see what people still want.
What features, if any, would you like to be removed from Vector 2022? There are things that people hate about V22, things that they want to get rid of.
What features, if any, would you like to change/augment in Vector 2022? Not everything is as simple as addition/removal.
It might be a bit long, but we'll just ask people to limit the length of their replies. We need to gather enough info to actually go to WMF with a concrete proposal, as opposed to the jumbled mess that was V22RFC2, where most ideas died at phab (assuming they got that far). Cessaune[talk]18:45, 26 December 2023 (UTC)Reply
I already objected below, but it seems my concerns were ignored. Questions 3 and 4 are redundant — they will, for the most part, be answered in Question 2. Question 1 is not helpful to the WMF, but if editors insist on its inclusion, I won't fight for its removal. InfiniteNexus (talk) 20:08, 3 December 2023 (UTC)Reply
At least for me, not everything I dislike about V22 is something that I want to remove, and literally nothing that I like about V22 would be something I would want to add/remove. So I don't see how the questions are redundant, but maybe I'm just an idiot? Regardless, redundancy isn't the worst thing in the world. Cessaune[talk]00:29, 4 December 2023 (UTC)Reply
If you're just trying to collect feedback on Vector 2022, instead of trying to achieve something actionable, probably don't even need an RFC tag. Can just be a survey page somewhere, with appropriate links dropped on busy talk pages.
I see an RFC tag and the RFC process as being useful for asking one focused question that is trying to achieve something actionable. The "rollback V22 as default on enwiki" RFC was a good example of a focused and actionable RFC question. –Novem Linguae (talk) 22:16, 3 December 2023 (UTC)Reply
I agree with Novem Linguae. We should reduce them to one or two focused questions which ask for specific and actionable changes. Æo (talk) 18:26, 26 December 2023 (UTC)Reply
As for the proposed questions, just keep it simple. What do you like about V22, and what do you not like. What are the primary features that you are looking for when selecting a skin? feels way too broad, and What features, if any, would you like to be added to/removed from Vector 2022? is just a duplicate of what do you like/not like. Actually, now that I think about it some more, do we even need a "what do you like" question? The answers are not going to help the Web team improve V22 other than to bolster their morale. InfiniteNexus (talk) 04:52, 24 November 2023 (UTC)Reply
Someone should start concretely planning the opening of the RfC and its advertising throughout Wikipedia to as many users as possible. Unfortunately I don't have much time these days, except for occasional edits and replies. @Awesome Aasim: is the draft complete? Æo (talk) 15:55, 28 November 2023 (UTC)Reply
First we should proceed with the V22RFC3. The question about responsive Vector (I don't know what it is) could wait or it could be integrated as part of V22RFC3. Æo (talk) 17:05, 28 November 2023 (UTC)Reply
@Cessaune: in my opinion, we should still leave room for the possibility of a reversion to V10 with a specific question among the others, something like Do you support a reversion to Vector 2010?. I have noticed that in addition to the German and the Italian Wikipedias, the Russian Wikipedia too has successfully reverted to V10. I suggest that we proceed and you modify the draft according to the various opinions collected on this talk page, so as to finalise the RfC format before it is opened to the community. Regarding the opening, what do you think about doing it on January 1st? Æo (talk) 18:11, 26 December 2023 (UTC)Reply
Æo, I would wait a tad longer to make sure no one misses the RfC on holiday. I'm also somehwat sceptical about yet another "Should we revert to Vector 2010?", which could derail the rest of the RfC. — Qwerfjkltalk18:19, 26 December 2023 (UTC)Reply
I'll start it on January 8th so as to catch the stragglers who are still on break. I'm with Q on the V10 thing. I don't want people to be turned off because 'bludgeoning'. Even if we included that question I'm fairly confident that it will end in, at best, no consensus. Cessaune[talk]18:26, 26 December 2023 (UTC)Reply
In my opinion, if it is framed as a secondary and less important question there is no risk that it will invalidate the whole RfC. What do others think? Æo (talk) 16:41, 27 December 2023 (UTC)Reply
I don't know. There is that possibility indeed. How is it that other Wikipedias have successfully reverted to V10? Was it the decision of the local administrators to implement the majority choice of the local communities? Æo (talk) 17:25, 27 December 2023 (UTC)Reply
I agree with Cessaune that such a question is pretty moot. The 3rd RfC seems to be focused on evaluating the skin. Putting that question would be too much chaos already. Aaron Liu (talk) 17:02, 27 December 2023 (UTC)Reply
I'm not certain what the draft is, but before you open it I want to ask; what is your aim with this?
If the aim is to return to Vector2010, the question needs to be simple; it can't be multipart as you are discussing above. It needs to be something like The WMF has established that it allows individual projects to persist with using Vector2010, with projects like the German Wikipedia doing so. Should the English Wikipedia revert to Vector2010?BilledMammal (talk) 02:36, 7 January 2024 (UTC)Reply
Ok, so reading it through your focus seems to be to improve Vector2022, rather than revert to Vector2010. The issue with your proposed questions are they are structured as a survey, rather than an RfC - I'm almost certain that any attempt to open an RfC on these lines will quickly be closed as improperly formatted. BilledMammal (talk) 02:43, 7 January 2024 (UTC)Reply
The WMF can get away with opening an RfC like that - a consultation and survey - because they run the technical departments etc; with certain exceptions that don't apply here the community can't. BilledMammal (talk) 02:53, 7 January 2024 (UTC)Reply
"Bring back V10" has an extremely low chance of passing; we've been through two such RfCs already, not to mention the WMF vetoed unlimiting the width by default. This one seems to be about offering "demands" to the WMF about what about the skin can be improved and what things should remain, sort of a "salvage" plan. The draft is from Aasim. Aaron Liu (talk) 02:44, 7 January 2024 (UTC)Reply
The issue is you aren't going to get a list of a "demands" from such an RfC. Even if the RfC is allowed to continue you're going to get low participation and no consensus on any particular "demand". You would be better off identify points of issue ahead of time and opening an RfC specifically on those points of issue - but even then I doubt you'll get a consensus, because the community is going to be loath to overturn WP:CONEXCEPT solely for a few issues related to the skin.
I know I'm coming across as negative here, and I apologize, but I'm trying to be realistic - unfortunately, your current plan is not going to get productive results or bring about change. BilledMammal (talk) 02:53, 7 January 2024 (UTC)Reply
I don’t know what you mean by overturning CONEXCEPT. If people don’t like how it’s formatted, then just make it not an RfC and just a survey. An RfC about reverting to v10 would need to be its own instead of flooding the survey with so much participants. Aaron Liu (talk) 15:40, 7 January 2024 (UTC)Reply
I would suggest to add the question Should the English Wikipedia revert to Vector2010? (with the note The WMF has established that it allows individual projects to persist with using Vector2010, with projects like the German, Italian and Russian Wikipedias, and probably others, doing so) as the main question of the RfC, per BilledMammal's concerns, which are in line with other considerations expressed by other users (including me) in the past months. All the other questions would consequently be "secondary questions" in their separate section. Æo (talk) 15:36, 7 January 2024 (UTC)Reply
Æo, we've had two RfCs. Vector 2022 has been implemented for a year. People have moved on. If people believe strongly that we should rollback, they will make it known; I added an open-ended question for this specific purpose.
Even if there somehow was consensus that we shold roollback, WMF wouldn't rollback anyway. It's too late for that. People have gotten too used to the new skin.
And EVEN IF, somehow, WMF rolled back to V10, the rollback itself will be contentious. Again, people are used to the skin and its features. Taking away things such as the sticky TOC will piss off a lot of people. Cessaune[talk]16:01, 7 January 2024 (UTC)Reply
You are probably right, and it is likely that a reversion proposal will be addressed in answers to the open-ended question. However, perhaps adding it as a secondary question (and the other ones as the main questions) with the reference to the Wikipedias that have successfully kept V10 would make a difference. Æo (talk) 16:32, 7 January 2024 (UTC)Reply
There would be a reference to the concrete precedents represented by other Wikipedias (the German, Italian and Russian ones) and Wikimedia projects (the Greek Wiktionary?). Æo (talk) 17:02, 7 January 2024 (UTC)Reply
Those precedents would disprove your claim that Even if there somehow was consensus that we shold roollback, WMF wouldn't rollback anyway. It's too late for that.. If I'm not wrong and if my memory is not mistaken, the German and Italian Wikipedias never switched to V22, while the Russian Wikipedia originally switched to V22, but at some point in 2023 switched back to V10. This would demonstrate to the community that it is never too late to overturn a WMF-driven change. Æo (talk) 20:06, 7 January 2024 (UTC)Reply
There's a chance that WMF could accept, but the chances are much smaller as enwiki is the WMF's flagship, not to mention how we've had two prior RfCs that didn't find consensus for rollback already. Aaron Liu (talk) 20:09, 7 January 2024 (UTC)Reply
None of the wikis mentioned failed to stop implementation and failed to rollback. Our first RfC failed to stop implementation, and the second one failed to generate a consensus to rollback.
I just don't get why the outcome would be any different than what it was previously, given that people have had time to adjust to the skin. Cessaune[talk]23:07, 7 January 2024 (UTC)Reply
@Sarri.greek: Thank you. So, you successfully rejected V22 on the Greek Wiktionary? I see it has not been implemented on other Wiktionaries either, including the English Wiktionary. We would like to hear your opinion about the third RfC which we are planning to open soon. Æo (talk) 15:45, 7 January 2024 (UTC)Reply
@Æo:, thank you for your response. Wiktionaries have different characteristics, and V22 was only applied experimentally at https://fr.wiktionary.org/ I think WMF team understands the peculiarities of wiktionaries. And, we, a small wikiproject, realise too, that skins are the property of WMF and will be changing every X years. So, we need to focus on securing the presentation of our bodytext, our _TOC__ styles, plus we would give to all our guests their desktop.style choice with a standard Sitenotice at all our pages like this one as courtesy. We have not found statistics for how many viewers use this or that skin when they have the switch.style.choice available (but most, use mobile view anyway). Thank you. Sarri.greek (talk) 17:05, 7 January 2024 (UTC)Reply
Thank you, M @Æo: to consider me, an outsider from el.wikt. Too many questions though. It seems that the subject is "How can we improve V22? State what is missing." But I cannot see any statistics like 'percentage of editors using V.Classic, V22, monobook' etc. neither any hint for a rollback. Sarri.greek (talk) 19:20, 10 January 2024 (UTC)Reply
Thanks for the ping. I'm not qualified to suggest improvements to the current default skin as I continue to use Vector 2010. By imposing Vector 2022, the WMF took responsibility for the user interface, making its shortcomings their problem and not ours. Certes (talk) 19:24, 10 January 2024 (UTC)Reply
The fact that a lot of readers managed to find this Wikipedia space page and vote that they didn't like it just shows that there was not consensus to implement Vector 2022. ~WikiOriginal-9~ (talk) 20:45, 14 June 2024 (UTC)Reply
Not exactly how it works. In fact, I'm relatively certain that you know that that's not how it works. We don't define consensus (or a lack of one) like this. Cessaune[talk]20:50, 14 June 2024 (UTC)Reply
Yet ~60% of the community was against it. This might not be "consensus" according to the criteria and terminology of Wikipedia, but it was a clear absolute majority. Æo (talk) 15:07, 22 June 2024 (UTC)Reply
Well, I'd argue since this isn't a content issue, it doesn't make sense in this case to apply the same policies and guidelines. This seems like the rare case where WP:IJUSTDONTLIKEITshould be considered a valid rationale — the appearance of the website is for the benefit of readers and editors, so whether they like it does matter and is relevant. InfiniteNexus (talk) 09:29, 26 June 2024 (UTC)Reply
The issue with allowing "I don't like it" votes is that user interface changes are often initially met with disproportionate backlash due to the fact that people are accustomed to the old version of the interface. I hated V22 and now I'm fine with it, despite it looking essentially the same as the day it was made the opt-out version. "I don't like it" can be taken to mean "I'm not accustomed to it" or "I'm not used to it" or many, many other things. And if one doesn't provide any rationale, how are we meant to know what you mean?
Also, the goal of an RfC is to figure out what the best course of action is, not to settle on a binary solution. RfC questions often pose dichotomies in the interest of simplicity, but I've taken part in multiple RfCs where the closer's conclusion differs from a yes or a no to the central question. "I don't like it" 1) doesn't impart any useful info that can then be turned into improvement, 2) turns the user's !vote into a vote, and 3) doesn't contribute to the problem-solving atmosphere that is engendered by lively RfC discussion. "I don't like it because XYZ" is much more useful. Cessaune[talk]10:26, 26 June 2024 (UTC)Reply
Certainly, I agree every !vote should be evaluated carefully, and less weight should be given to "I just don't like it" !votes with no additional rationale, but I don't think they should have been discarded entirely when assessing consensus. The fact that a significant number of readers and editors opposed V22 should have been taken into account, but bureaucracy of course once again got in the way of common sense, and V22 was retained because there was "no consensus" to revert it (even though there was no consensus to implement it in the first place). InfiniteNexus (talk) 13:39, 26 June 2024 (UTC)Reply
WP:V22RFC2: While most experienced editors are already familiar with what makes for a strong argument, this RfC saw a large influx of new editors, most of whom are readers who wanted to make their voices heard. Many of these readers are not familiar with Wikipedia's policy of consensus. The strongest arguments are those based on our policies and guidelines, while the weakest are those based on subjective opinion, and here is where we should start our discussion. This request for comment came soon after the change of the default skin, and many !votes were based on personal opinions about how Vector 2022 was better or worse than Vector 2010. While these are usually considered weaker arguments, they were not entirely discarded but were not given as much weight as other points. The closers did exactly what you're suggesting.
WP:V22RFC: Overall, there is a positive reception to the changes. Supporters were enthusiastic in their appreciation of the UX improvements, and most editors currently opposed expressed specific and narrowly-scoped concerns, rather than wholesale objections to Vector 2022.
The most substantial concern, and the only clear blocker, was the issue of fixed-width. The idea of using a community-maintained gadget is deemed insufficient. It should be possible to achieve a full-width experience using a WMF-maintained toggle, which is clearly visible and available to both logged-out and logged-in users. There were also notable concerns about non-intuitive icons in the sticky header and the behaviour of the language selector, which we believe need to be addressed to achieve a firm consensus.
If all the concerns outlined above are satisfactorily addressed then we see community support to roll out the change. I think that all of these concerns were satisfactorily addressed, besides the non-intuitive icons thing, which, again, I'm gonna attribute to 'I'm not accustomed yet'—the icons are pretty bog-standard—and people didn't largely complain about the intuitiveness of the icons in V22RFC2, rather the necessity. Based off V22RFC's closing statement, I believe there was a consensus to implement that was not precluded by the no consensus outcome of V22RFC2. Cessaune[talk]15:48, 26 June 2024 (UTC)Reply
These icons hadn't been invented for half of Wikipedia's history, I think it's a bit recentist to say they're bog-standard. DaßWölf23:40, 26 June 2024 (UTC)Reply
I barely edit Wikipedia anymore but I still read it in everyday life, including at work, so I unfortunately often end up using Vector 2022. I haven't changed my mind about the default fixed width and its wasted potential, nor the unprofessional modernist design. I do appreciate that readers are being given a little more of a choice, but again it's all lost behind a new backward design change that can be countered only through a completely mystifying icon -- since when does a private browsing mask stand for text size and page width customisation? "AA" would at least fit in with other websites.
I don't recall which of this I mentioned in the original RfC; a lot of these complaints were covered well by others. We shouldn't ascribe positive or neutral views simply because someone didn't complain about a feature. If they saw only a single thing wrong with the design and deployment process, they might've well been in the support !camp. I know I would have. DaßWölf23:40, 26 June 2024 (UTC)Reply
I don't think Vector 2022 is unprofessional (or at least less professional than V10), and I am interested as to why you think that. I know people who consider phpMyAdmin to be top-of-the-line when it comes to professional design, and I strongly disagree. If professionalism is built on skeuomorphism, drop shadows, and decade-old practices, as some of my colleagues implicitly suggest, then sure. As it stands, I don't think that either V10 or V22 is more professional than the other.
I like Aa a lot more than the current icon, and I think it should be changed to Aa. I think I said this somewhere, but I can't recall where. Cessaune[talk]01:17, 27 June 2024 (UTC)Reply
I found V22 sloppy and poorly polished when it was proposed, and while many of the issues have been addressed, it still gives page layout bugs and slips every once in a while. These problems are minor, and usually of the sort that some element gets hidden or forced to overflow etc, but these are all polish issues that were absent in Old Vector.
I also think that good interaction design is independent of skeuomorphism or other aesthetic top coats. The fact that Ol' Vec' gave visual cues—in the form of gradients, borders and colours—to the start and end of page content was good design. Maybe to some a gradient that implies dimensionality is skeuomorphic, but it’s still a fundamental tool used in modern designs that few would call skeuomorphic (iOS was the big one for popularising that term, and the present design still uses drop shadows). An all-white page without any visual cues to different zones is bad design. Menus that can be hidden is great; menu buttons that are ambiguous, and show in a different order when you scroll below the fold are bad design. There is a lot to V10 that was visually dated, but it could hardly be called unprofessional in the same way that V22 can. — HTGS (talk)09:02, 1 July 2024 (UTC)Reply
V22 with borders is as professional as V10. I've gotta check Phab but something related to borders as a toggle under the Appearance tab might've been filed. Cessaune[talk]17:50, 1 July 2024 (UTC)Reply
I don’t know anything about upcoming changes to the design, but borders are just one possibility. There are many right answers here (borders, colours, drop shadows, gradients, dividers, etc) but I find it absolutely bizarre that none of these were considered at any point until designers got complaints and suggestions from the community relying on them.
I have no issue with design decisions coming from on high. It is – quite clearly – hard to create good communication design or interaction design with a crowdsourcing model, but I am still disappointed with the outcome we have, and so while I don’t feel that calling V22 “unprofessional” is necessarily polite or productive, but when comparing the two, it isn’t wrong. — HTGS (talk)02:05, 2 July 2024 (UTC)Reply
In all the original images for V22, the prototype layouts had borders. Somewhere along the way they removed them, and I'm not sure why. I hesitate to dig around in the RfC, but there was a link floating around that brought one to multiple different border designs. V22 prototype, September 22, 2022
V22 as implemented on January 18, 2023
I don't like the lack of borders in the design, but it doesn't make sense to call it bad. I guess that what I would consider the most professional (and not necessarily the most preferred) design would be the one that holds up the best under statistically sound and transparent testing, which was something that I begged for in the RfC. And borders didn't make the cut, very unfortunately. It would be unreasonable of me to suggest that what I think is good design is better than what testing finds to be good design (or at least better design). That's why I like the idea of having a border toggle in the Appearance tab: it creates an opt-in scenario where everyone wins. Cessaune[talk]07:51, 2 July 2024 (UTC)Reply
I think the prototypes are actually basically showcasing the same interface. It's just that in the first screenshot, you have the main menu. Until last December, the main menu, and only the main menu, did indeed have a bsckgroundfor some reason. IMO that looked bad because neither Tools nor ToC nor the top bar had a background.
They arrived to this conclusion because in the fifth prototype testing in 2022, they presented nine prototypes for the borders and backgrounds. People favored design #1 and #9 the most; yes, the first and the last. The designs were changed by a radio menu in a corner that laid out the selections in that order. The order wasn't randomized, and so they can introduce "both primacy bias towards the first option and recency bias towards the last option", as this paper says. #1 had more support, and they blindly went for it. #9 (now known as Zebra) had a lot of support in RBV22 but they did the very suspicious research below. Aaron Liu (talk) 13:56, 2 July 2024 (UTC)Reply
A 3% increase in pageviews per session in the treatment group attributed to Zebra
A 3.4% decrease in edits per session in the treatment group attributed to Zebra
A 17% decrease in the click rate of the table of contents
An 87% increase in the page tool pins per session
(emphasis mine), which is... surprising, to say the least. They then decided that these stats indicated that it would make the experience worse. Aaron Liu (talk) 21:51, 1 July 2024 (UTC)Reply
And lo! Separation may not have even been the issue!
We noticed that a significant amount of the decrease in edits and pageviews came from screen sizes narrower than 1200 px. We combined these results with the results of our user tests. These did not indicate significant differences in readability between the test and control designs.
we plan to improve readability and focus on the content through the following:
Introducing improvements to Zebra, optimizing for narrower screens, and repeating the tests.
They did not repeat the test. They took the results for granted:
As we reported previously, the A/B test didn't prove that the color-based separation was an improvement.
Wow. Decisions about the interface are made based on calculations and a 3.4% difference! I am speechless. Æo (talk) 00:21, 10 July 2024 (UTC)Reply
3.4% is quite a big difference. And what's wrong with using data to justify a UI change? In fact, I'd argue that using data is extremely valid. Cessaune[talk]05:59, 10 July 2024 (UTC)Reply
Well put, HTGS. Good design isn't built on abandoning practices solely because they're currently unpopular, particularly not when the offered alternatives aren't as good. I don't think any kind of annoying nostalgia is worse than groping with a touchpad for the invisible barrier between TOC and the main text to get at the extra thin TOC scrollbar which only shows up on mouseover. Mind you, this is four years after the first open beta. Did Vector take this long to catch up with Monobook?
There's also the interaction between developers and open beta testers, including enwiki users since the rollout. Why insist on calling it "Vector 2022" when it's clearly more different from Vector than Vector was from its previous skin? More importantly, why all the painful RfCs just to get the fixed width toggle, which happens to be the last item to load and is turned off by default? There are pages and pages of pre-enwiki-rollout archives on mw: of fr: users asking for full width or a toggle and told in various ways it'll never be done. Surely the developers knew it would be an easy job as it turned out to be after one of the enwiki RfCs. This toggle is also still unavailable to readers on most other Vector 2022 wikis.
And now we have another button to click and another sidebar to get rid of to get back to the original experience. Admittedly, this particular change helps the problem of empty "ad space" a little bit, but at the cost of letting readers see even less of the page by default on each screenful -- fewer lines of the so-called "standard" line length which depends on the text size you chose.
It kind of feels like the developers are less focused on building a functional skin than driving a point against user customisation and skeuomorphic cues (the latter amusingly seem to be making a silent comeback, see e.g. Windows 10 vs 11). Is that professional? DaßWölf01:48, 2 July 2024 (UTC)Reply
One note: you can toggle the width on every wiki since the rollout of the "appearance" sidebar. I agree with many of your other points in the first two paragraphs. Aaron Liu (talk) 02:23, 2 July 2024 (UTC)Reply
mw:Talk:Reading/Web/Desktop Improvements
