Talk:Reading/Web/Desktop Improvements

Overview

Features

Repository

FAQ

Discussions: Desktop Improvements • Accessibility for reading.

---

You can comment in any language. We are especially interested in:

• Feedback on the functionalities we have already deployed. (Please see Reading/Web/Accessibility for reading/Known issues before reporting bugs.)

• Expanding the list of existing gadgets and user scripts that are related to providing a better desktop experience.

We are discussing with many volunteers and communities at the same time. It may take some time for us to reply here. We apologize for that.

Thank you!

(Translate this page)

I love the new width of the page

Please don't change it, or let others convince you to change it. The narrower width makes the reading much more comfortable, it looks more modern, and the eye can find the next line with ease. Blank space is not a problem, and who complains about it just needs to get used to it; the strangeness will go away eventually. If blank space were a problem, you should tell that to Facebook. Its feed also has lots of empty space around it. What's the problem?

The new Vector is the best skin in the history of Wikipedia. Congrats from a Portuguese Wikipedia editor. Bageense (talk) 18:38, 20 June 2021 (UTC)

Thank you @Bageense, this is very kind! All team members were really glad to hear that. While we can't promise we wouldn't listen to volunteers with various perspectives, we do hope that the limited width will be considered as an improvement. SGrabarczuk (WMF) (talk) 19:22, 1 July 2021 (UTC)

@SGrabarczuk (WMF), I'm also from the Portuguese Wikipedia. I think it would help if the white spaces got a bit grey, it looks worst in the all-white configuration. Moreover, it would be great if the desktop version looked like the mobile app version, with that awesome dark mode. Paz e concórdia (talk) 14:11, 3 August 2021 (UTC)

Hello @Paz e concórdia. Thanks for your opinion. Have you seen the task on Phabricator where we've begun discussing the background color? You may be interested in reading the comments. Regarding the dark mode - yes, that's an excellent idea, but we will not work on that. Basically, we can't deliver two separate @versions of the same page. I'll add a section to FAQ with an explanation why. SGrabarczuk (WMF) (talk) 17:15, 5 August 2021 (UTC)

I agree with Bageense. Its a huge improvement Shisma (talk) 07:20, 15 August 2021 (UTC)

I do not agree with Bageense. Usually when I read a Wikipedia article I do it from my laptop witch hasn't a wide screen. By squeezing the page together it looks like im reading the article from an iPad! On safari web browser, for example, I already have the read option to resize similarly the article (also with white spaces at the sides). An other point is that by zooming the old web page I could resize the text as I wanted (for example with cmd + or cmd -). With the improvement the size is fixed, the only thing that can happen is the white areas to enlarge or the text to ulteriorly shrink. Also, when a Wikipedia page has a propriety box (for example the chemical elements) it ulteriorly fills the text space, whereas the old version just placed it where now there is withe space. I think that it was better since the box containing technical information is less important than the article, it was logic to put it aside and not with the text. Finally, a tighter space means more scrolling witch on the older version, someone who doesn't want to always touch the mouse or keyboard, can just lean back on his chair and enjoy better the reading.

I wish the user will have the option to wide read a Wikipedia article in the future! PaDalton (talk) 15:04 21 August 2021 (UTC)

The reduced width is a single issue dealbreaker for me. The people that like the page narrow can... you know, just make their browser windows narrower? For me it is completely unusable, it gives up so much screen space, I need to scroll so much more, I have less information available in one view etc. --217.67.131.246 10:26, 10 September 2021 (UTC)

I strongly oppose this change, the current interface should be default and the one who make you scroll more, optional. Browsers already reduce vertical space, scrolling is frustrating, and increase visual fatigue because things moves. If you want easier readability, please implement a 2 columns layout like most paper encyclopaedias (this make even more sense on > 16:9 than on most book formats), scrolling should work with Page Down & Up keys & behave like 2 columns ePUB readers and physical books (bonus: incremental scrolling: bottom of 1st column overflows to the top of the 2nd one for those who wants it). See (CSS multicol) & a study on the use of multi columns and scrolling. Wasted space reduce usability. Hploter (talk) 10:42, 11 September 2021 (UTC)

I too would like to add my voice to the opposition. I always greatly admired Wikipedia as the last bastion of good UI. This abomination is merely a sign of mobile design's intrusion into the traditionally-desktop websites. I view it quite pessimistically - our voices will never be heard (because it's modern vs "outdated"), and it's only a matter of time until the last website on the Internet is turned into unreadable mess - with obnoxiously moving parts (the arrow signs), arcane and silly emojis ("hamburger sign"), wasted space (wide white bars to the sides), which requires more clicks than before (I hate the language tab already introduced which is always shrunk for logged out users). And what's the point? Mobile users already have mobile clients, afaik. Thus is mobile supremacism, pure and simple. Or do like Reddit did, and leave us alone on the old domain thingie.--Adûnâi (talk) 03:32, 19 September 2021 (UTC)

I also dislike the new default width, and is a dealbreaker for me with new Vector as a whole. When I took a look at the beta Vector, my first reaction was that I ended up on the mobile version of the site. Please do not change the default width for the desktop version of MediaWiki. I don't care about having a single design for both mobile and desktop versions, and would argue that a single design is detrimental to the user experience due to the very different nature of the platforms. The full-width website is something that has been around for over a decade, and I don't see the need to change it as the default. Mediawiki is not a tool for commercial purposes. Whitespace on the sides (which screams "banner ads") is not something needed as an informational platform. The platform exists to inform the reader and give them information, and having huge amounts of whitespace takes away what the reader can see on a single page and replaces it with nothingness. I would be perfectly fine with adding a setting that lets you switch to a mobile width for logged in users, but it should not be made the default for logged out users. This trend of creating a single design and thinning out the page width as a result is a step in the wrong direction, and is the reason why there exist situations like Reddit where a good portion of the userbase uses an older version. I am not against having a thin view option for the user, since if someone prefers that format they should be able to have it, but it should not be made to be the default experience. GalacticRuler456 (talk) 19:04, 8 October 2021 (UTC)

A year ago, I was dissatisfied with how narrow and overly compressed content looks with the current max-width on 1920x1080 display. Well, now I have a 2560x1600 display and it looks even worse to me, as only one third of my screen is filled with useful content. It feels like I'm using a mobile version of the website on mz laptop and it doesn't help that I don't really like a sidebar and would prefer it hidden, so that nothing distracts me from the article. I would really love to have several max-width options to choose from (e.g., leave as is, fill the entire workspace container or even fill the entire page container). Too bad it's still only possible with user scripts and styles... At the moment, it is also a single most important reason for me to stick to the old vector theme. 2-column layout, as suggested above, also might be a viable approach, though it might break a lot of stuff. Adamant.pwn (talk) 16:32, 4 December 2021 (UTC)

Is the new fixed width made for people too stupid, or too lazy, to resize their window? 2A01:CB15:338:E200:FB14:192D:A0F3:C30C 22:41, 8 December 2021 (UTC)

My thoughts exactly. (And wait till you see what they are now messing with…) Of course, if one uses Monobook instead of Vector all this silliness goes away, but having “power users” experiencing the site in a way that is more and more distant from that of an entry-level or unlogged user is not a good trend. Besides the fear that one day suddenly they will pull the plug on Monobook fills my daily Wikipedia editing experience with dread and bitterness.

I feel transported to the bad old days of the early 1990s when really bad ideas about user interface (anyone remembers those multimedia CDs?) florished and wilted, to eventually give way to clean functional designs such as early Wikipedia. To rehash all that nonsese again is even worse, as the accumulated experience is not being taken into account and every interface designer wants to do it from scratch and reinvent the wheel (like seemingly every .swf designer did in the 2000s…), compounded by the fact that, unlike back then, in the 2020s user interfaces on a screen are not exactly newfangled — yet bad ideas and attempts at bogus “paradigm shift” keep plagueing us.

One imagines that something like Wikimedia would be exempt from the constant changes that affect major commercial websites, but that what we get when we voluntarily put our necks on the stump and give away the keys of the realm to higher-ups who post more tweets and instas in a day than Wikimedia site edits in a month (not hyperbole here, and you all know it), No wonder said luminaries, accostumed to social media and mobile UIs, will be shocked to see a functional design meant for comfortable encyclopedia reading and editing: «Ew, is that “my” website?!, let’s change everything about it!» And here we are.

Tuvalkin (talk) 23:00, 21 December 2021 (UTC)

I strongly dislike the fixed width for same reasons as stated above. The new UI does seem to make using Wikipedia only harder and frustrating experience. Authors of this change seem to be bit of out of touch what most existing contributors preference. --4shadoww (talk) 17:24, 10 December 2021 (UTC)

I often use widescreen monitors. Would it be possible to select only the new width and not the other new design stuff? Jiiimbooh (talk) 21:40, 22 December 2021 (UTC)

I can understand that there is value in consistent width in terms of article presentation, however there's significant wasted screen space that's a deal breaker for me. Why is this not just an optional toggle, I'm even okay with fixed width being the default, so long as I have the ability to expand it. Supertrinko (talk) 01:05, 25 January 2022 (UTC)

I agree on the new width. The desktop page seems to provide a much better reading experience now. Thanks! -- Kaartic ^[talk] 08:35, 26 January 2022 (UTC)

Search box size and location

Hi. I have 3 initial thoughts about the sticky header.

1. I really like the easy access to the page-tools, without having to scroll up or remember the key-combinations! I've been wanting this for years.
2. I'm still hoping for easy 1-click access to my watchlist, which I access multiple times each day (and the Watchlists system is good for newcomers to become familiar with, which they won't if it's buried in a menu, and even below the "Preferences" item which most people rarely access.)

3. 

   Screenshot of three pages, showing (1) sticky header, (2) current desktop improvements, (3) classic vector, with the search-click-targets highlighted in red

   (Main point) I'm feeling very frustrated when I try to use the Search in its new location, partially because the click-target is so small, and partially because it is in an inconsistent location.
   1. The box I need to click on is very small. Previously, I was loving the increased size of the search box in New Vector. This new design is vastly smaller than the previous iteration, and quite a bit smaller than the classic design. Search is one of the most core activities readers/editors/users engage with, and deserves as much target area as feasible. [See screenshot, for emphasis/clarification.]
      • (cf. w:Fitts's law for more details about link target size - TLDR/IIUC, it's all about how many micro-adjustments we need to make, after the first initial "rough" movement in the general direction of the target.) [Edit: I believe this is also a significant part of the reason for why many users advocate for "text instead of icons" in some circumstances - we/they like the inherently bigger click-targets.]
   2. The magnifying-glass icon is in the same "area" within a browser window as the "hamburger menu" - This is confusing for muscle-memory, and should probably be consistent instead, so that I can reliably tap or click in the same area of the screen, to start a search, regardless of whether I've scrolled down.
   • I can appreciate the idea of showing the "Page name" and perhaps "Section name" at the top, but I think the Search-box needs a very large click target, so that we can throw out mouse-cursors/fingers in the general direction and be likely to end up in the exact correct spot. -- My only idea for how to resolve these 2 goals, is to show the page-name within the search box. I'm not sure if that would work.
   • I also appreciate that having it in the corner, might be beneficial for some Mobile users, because that places the click-target nearest to the thumb. But it is very frustrating for me as a mouse user.

I hope those details help! –Quiddity (talk) 02:36, 10 November 2021 (UTC)

I don't see an issue with the search box being moved... except for why? It's exceptionally common around wikis everywhere that the search box is in the upper right. So what is the justification for changing from this norm? We shouldn't do change for change sake, but if there's a good reason, that's fine. Supertrinko (talk) 01:09, 25 January 2022 (UTC)

A question

Do you have any plan to make vector skin mobile responsive? I use this code to do this- if ( mw.config.get("skin") === "vector"){ document.head.innerHTML += '<meta name="viewport" content="width=device-width, initial-scale=0.77, maximum-scale=1.0, user-scalable=0">'; } Yahya (talk) 09:07, 11 November 2021 (UTC)

We've been working towards that goal, but it's not an official goal of the project. For third party wikis you'll be able to do this via the configuration flag wgResponsiveVector and that script is exactly what I was going to recommend for Wikimedia wikis :) Jdlrobson (talk) 00:01, 6 January 2022 (UTC)

Temporary logo customization

In legacy vector, the accepted way to temporary change the logo (for anniversaries and milestones of wikis) was through css. In the improved vector, the logo items seem to be hardcoded in the html. Any tips? Geraki (talk) 19:30, 15 November 2021 (UTC)

The instructions you're looking for are mentioned on /Features/Header, and it's /Features/Breaking changes + T245190. SGrabarczuk (WMF) (talk) 14:20, 16 November 2021 (UTC)

عدم سازگاری و تداخل در نسحه جدید و قدیم. Incompatibility and interference in new and old versions

سلام من از پوسته جدید وکتور بسیار راضی هستم. متاسفانه در نسخه 1.36.2 و 1.38 که تست کردم، مشگل تداخل نسخه قدیم و جدید را می‌دهد. در 1.36.2 با تنظیمات پیش فرض وکتور و استفاده از کامند 2،برای کسانی که وکتور وارد سیستم نشده‌اند، وکتور قدیم را نشان می‌دهد و در نسخه 1.38 برعکس است در صورتی که تنظیمات من همانند دایکرومنت پوسته وکتور است. این موضوع کلافه ام کرده است. آیا ب ای حل این مشگل راهکاری دارید؟ با تشکر

hello I am very satisfied with the new vector shell. Unfortunately, in versions 1.36.2 and 1.38 that I tested, the problem of interfering with the old and new version gives. In 1.36.2 with the default vector settings and the use of comand 2, for those who are not logged in, the old vector shows and It's the opposite in version 1.38 If my settings are the same as the docker theme. This has overwhelmed me. Do you have a solution to this problem? Sincerely Sokote zaman (talk) 22:54, 22 November 2021 (UTC)

Hello @Sokote zaman. Just to be sure, you experience this on a third-party wiki, not a Wikimedia wiki like Wikipedia in Farsi?

سلام @Sokote zaman. فقط برای اطمینان، شما این را در ویکی شخص ثالث تجربه می کنید، نه ویکی ویکی مدیا مانند ویکی پدیا به فارسی؟

SGrabarczuk (WMF) (talk) 13:39, 24 November 2021 (UTC)

سلام بر شما

@SGrabarczuk (WMF)

من میدونم شما بحث پشتیبانی از ویکی های شخص ثالث را انجام نمیدهید. قبلا این مشگل را هم در انجمن و هم در گزارش اشکال مدیاویکی مطرح کردم ولی هیچ جواب درستی نگرفتم. اگر رضایت کاربران ویکی ثالث برای شما مهم است و باعث می‌شود اشکالات اسکریپت را پیدا و ارائه دهند خدمتتان عرض میکنم این یک مشگل هست و حتی اگر یک ویکی با حالت نصب اولیه بدون کانفیگ شخصی هم اگر نصب و تست کنید متوجه این مشگل خواهید شد.

من این مشگل را در ویکی پدیا فارسی ندیدم و تست هم کردم جواب داده است. خب لااقل بگوید از چه دستوری استفاده کرده‌اید و توضیحات مربوطه را بروز کنید. من ویکی را ب یا خودم نصب میکنم و به جایی وابسته نیست.

با احترام و تشکر فراوان از همه شما را بابت این اسکریپت بسیار خوب Sokote zaman (talk) 14:52, 24 November 2021 (UTC)

Narrow width issue

This new width is a nonsense to me. If so, it means current layout of existing pages will have to be redone, more specifically pictures positioning. Large tables will take two lines instead of one !!! I guess I won't contribute anymore, if it goes this direction.--GGir (talk) 09:10, 12 December 2021 (UTC)

Of course, it's a mess, and it will proceed regardless. The entirety of the Internet is being turned into an unreadable mobile downgrade. Wikipedia used to be the site one could be sure would not feature annoying pop ups, moving parts, large fonts. But as it looks, the time will come when they implement an unskippable menu at the top that remains on the screen no matter how much one scrolls. Or a login button that only appears when one scrolls down, like in the WordPress rework. Merely brainstorming more annoying ideas, feel free to use them, my dear.--Adûnâi (talk) 11:37, 12 December 2021 (UTC)

"It is not possible for them to intuitively switch languages, search for content, or adjust reading": adjust reading, but WMF provides a limited non-adjustable width with the New Vector. They want to adapt the desktop skin after the incomplete mobile skin. The logo is also destroyed, being smaller, only because on mobile it cannot be large. --NGC 54 (talk | contribs) 12:29, 12 December 2021 (UTC)

¿Qué tal sí hacemos que ese enorme espacio en blanco sea gris?

Digo, ese es mi único pero con la nueva piel. Sin embargo, este problema de "hay demasiado blanco" ya está resuelto en la piel Timeless. En esta piel el texto negro contrastado con blanco de fondo está flanqueado por unos sutiles espacios grisáceos. Me es bastante cómodo esto, es lo mismo que ocurre en Word por defecto.

Espero que esta simple sugerencia apacigüe las inquietudes de algunos uwu. --Christian Alexis Arroyo Ortiz (talk) 09:59, 18 December 2021 (UTC)

Gracias @Christian Alexis Arroyo Ortiz. Esta es una traducción del Traductor de Google (no se español): trabajaremos en esto como parte del Refinamientos estéticos generales. Puedes leer más en T259240. SGrabarczuk (WMF) (talk) 00:07, 21 December 2021 (UTC)

So far so good

Copied from Topic:Wmjydo5ya7z8zz3g

Really liking that the table of contents now moves as you scroll down. I can see where I am at, what's next, and can easily warp to where I want to go. A2Bros (talk) 05:13, 20 December 2021 (UTC)

Contra the "sticky header" clutter and interwiki language selection

The dreaded "unskippable menu" was a joke on my part above, but it's apparently a reality (the "sticky header")... What an abomination - why would I ever want my screen space to be stolen by an ever-present menu? Are you expecting me to forget what page I am reading, so that I would need the page title at the top constantly? This is an objectively infuriating change that apparently crawled to desktop from mobile phones. I vehemently oppose such a direction - especially for the logged-out users! I actually routinely press F11 to read full-screen without the distraction of my browser UI, let alone would I wish for more clutter, unavoidable at that.

Sticky header will allow users to access important functionality (logging in/out, edit, talk pages, etc.) without requiring people to scroll to the top of the page.

These are not "important functionality". And even if they were, I would not permanently trade my screen space for a thing I can access in a millisecond by pressing the Home button! Do people not employ keyboards in 2022?

And the interwiki language button is getting downgraded even more? The previous change at least showed a few important languages at the ready, one click away (based on my IP address or something). Now it will actually force me to click more times. What a joke. Every single change is counter-productive, it's impressive.--Adûnâi (talk) 23:16, 20 December 2021 (UTC)

Thank you for your comment, @Adûnâi. Please be reminded that here on MediaWiki.org, the Code of Conduct is in force. I understand you're dissatisfied with our assumptions and choices. We can talk about this - how to disable it individually, etc. Please stop using offensive and derogatory words, though. For instance, let's leave "abominations" in the Dungeons and Dragons world. Thank you. SGrabarczuk (WMF) (talk) 00:50, 21 December 2021 (UTC)

@Adûnâi Home button makes you loose your scrolling position. So that is not a solution. Having a sticky header allows you to e.g. open a talk page in a new tab. Or open a specific language version in a second window for a side-by side view.

It is quite easy to remove the header if you do not find is useful. You can do that with e.g. Stylish or by using commons.css. Just add a line like this to your css:

#vector-sticky-header{display:none!important}

Nux (talk) 22:21, 11 January 2022 (UTC)

@Nux, I fail to see how such fringe cases justify ruining the experience for 99% of the time. Even in your example, a far simpler and more efficient solution can be found - such as copying a part of the text and then CTRL+Fing it to find it immediately again (apparently, it is a downgrade aimed at the mobile users unable to do it, just like with the redundant and annoying arrow buttons on this very page when my keyboard already has the Home and End buttons). Regarding your suggestion to disable the sticky, unscrollable header - 1) I have no idea about coding that stuff; 2) does it even work when logged out? If not, it will be utterly useless to me, as I rarely log in.--Adûnâi (talk) 00:47, 12 January 2022 (UTC)

Yes, Stylus works when you are logged out. You seem to be versatile in using your keyboard so you should be able to handle CTRL+C and CTRL+V 😉. Nux (talk) 01:10, 12 January 2022 (UTC)

I hate it, but others...

Copied from Topic:Wmjryh74k5zswirt SGrabarczuk (WMF) (talk) 00:15, 21 December 2021 (UTC)

Hello, as it is with many proposals to improve the look and feel of the Wikipedia interface, some people like it, others don't care and then again others absolutely hate it. What to do? Wouldn't it be better in general to give the readers and collaborators more options for customization? Ziko (talk) 15:17, 20 December 2021 (UTC)

I suppose I've no objections as long as it's easy to opt out. However, I'm concerned that such developments divert the WMF's limited IT resources from the many stalled Phabricator tickets which represent fixes and minor enhancements that the communities actually want and have asked for. Certes (talk) 16:20, 20 December 2021 (UTC)

I love it, but...

...there is a reason that a couple of issues keep cropping up in this discussion: they represent a gut reaction of a lot of users to the changes based on massive amount of cumulative web experience. In short, the wisdom of the crowd. Please don't ignore them.

The first issue is the massive amount of wasted white space on 16:9 monitors. Readers, even those who don't log in or edit, don't necessarily want or expect an encyclopedia to look like Twitter. If logged out users can have four configuration choices for the table of contents, surely at least a wider option could be offered. Or at least compromise by using a percentage rather than a pixel number to limit the content width. The nearly rote response that pops up in most similar comments above of "some people who are much smarter than you say that you'll like this better" is not comforting.

The second is less important to many, but the downgrading and re-"organization" of the interwiki links makes linking to other language articles incredibly tedious. Even if the interwikis have to be hidden, ditch the geographic-based lists and just alphabetize them. This presents difficulties for me when I'm on an interwiki where I don't know the language well, and I scroll through a list a regions names that I don't understand. I guess the upside is that I've learned some new foreign-language terms lately.

I'm not sure where the clamor for a new interface came from, but I generally like the look of the new interface except of the blank spaces that make Wikipedia look like a social media site. I accept the gradual dumbing down of the look of most websites (e.g. greater use of emojis), but appreciate a stronger emphasis on the graphic "feel" of the new interface. AjaxSmack (talk) 03:12, 21 December 2021 (UTC)

Sidebar vs table of contents

Copied from Topic:Wmlday2rza1i9wa7

When I open the sidebar via the "<<" on the upper-left corner, the table of contents disappears. When I close the sidebar, the TOC reappears. Is this intentional? If not, then can this be fixed? George Ho (talk) 08:27, 21 December 2021 (UTC)

Humble request (toggle for width + header)

It's awesome, great work, and a much-needed refresh. I imagine there will be bellyaching about the new text width and the sticky header. I imagine most people who complain about the one will complain about the other, so I recommend a toggle (near the "mobile view" link on desktop) to simultaneously make the text full-width and remove the sticky header. It'll set a cookie. No need for a preference or anything advanced like that. I make this suggestion because there will be absolute metric tons of complaining if not, from my experience of how introducing those two design elements usually goes. Thanks again for working on this. Enterprisey (talk) 08:35, 21 December 2021 (UTC)

Hello @Enterprisey. Thank you for your support. There is a gadget disabling the limited width already. I think it will be also possible (I haven't tested that myself yet) to hide the header via CSS. Would that work as a solution? What do you think? SGrabarczuk (WMF) (talk) 13:44, 21 December 2021 (UTC)

@SGrabarczuk (WMF): No, for several reasons:

• Gadgets are per-wiki. As long as there are no global gadgets, most small wikis will almost certainly miss out anything that’s gadget-based, because simply there’s nobody installing the gadget, let alone localizing it (e.g. translating the description and any human-readable strings it produces).
• Gadgets are logged-in-only. Logged-out users cannot enable gadgets (or only using complicated ways like bookmarklets or browser extensions that explicitly load the required ResourceLoader modules, but that’s not for average users). In contrast, purely cookie-based solutions like the mobile view toggle or the table of contents collapsing work for both logged-in and logged-out users.
• Gadgets are not very visible. If one specifically looks for gadgets, they can find them in the preferences, but people won’t discover it by chance, while they may discover a link in the page footer.

The first two issues can be resolved by using a browser extension (although it introduces another drawback, browser-dependency, and thus it may not even work for many browsers, especially for smartphone and tablet ones), but the third one, visibility, can be resolved only within the skin.

A solution that consists of pure CSS except for the small amount of JS toggling the class is a good direction to avoid HTTP caching issues—but this CSS and JS should live within the skin, not in some gadget. —Tacsipacsi (talk) 16:24, 21 December 2021 (UTC)

@SGrabarczuk (WMF), thank you for the response! I anticipate the majority of the complaining coming from readers without accounts, and it would be surprising to require an account for a minor display change. I think this should be treated exactly like the table-of-contents toggle (thanks Tacsipacsi for that example). I feel that this would serve a large number of readers for a relatively small amount of work. (This might finally get me into MediaWiki dev :) that is if WMF product is okay with this change.) Enterprisey (talk) 23:03, 21 December 2021 (UTC)

Larghezza finestra

Veramente pessima la larghezza dello spazio utilizzato, si perde da un terzo (pagine in ns0) a metà (pagine di discussione) dello spazio attualmente disponibile ValterVB (talk) 18:06, 21 December 2021 (UTC)

Too narrow (again)

Why waste all that space? I have a nice wide screen so I want to take advantage of it. Those who want to have it narrow can just reduce their window sizes? Overall, not a great look. Not exciting. Quite boring.--Xania (talk) 11:23, 22 December 2021 (UTC)

What is our objective? > Currently the interface doesn't highlight the community side & list item 4

List item 4 under https://www.mediawiki.org/wiki/Reading/Web/Desktop_Improvements#Currently,_the_interface%E2%80%A6, (connects with "...doesn't highlight the community side"):

The large difference in experiences among our desktop interface, apps, and the mobile web, makes it difficult for readers to connect our products. There is a lack of unity in the concept of Wikimedia sites.

I think this is saying that if user experiences across devices and Wikimedia sites are similar, it'll be easier to help connect the products together. But does that help the user? Maybe it'll help discovery of other, less popular Wikimedia sites, but this new design hides many community links behind a menu...

(Sidenote: Hmm, maybe increasing the size of many user controls on the top right could help, but wait what's the difference between alerts and notices again? I guess alerts are just more important notices, since they both show the "All notifications" and "Preferences" buttons at the bottom.)

I'm probably confused about this particular objective. Like how it relates to the table of contents + menu sidebar change.

AltoStev (talk) 18:08, 22 December 2021 (UTC)

Concern with using logged in vs. logged out to determine sticky header design

Based on the most recent update on the sticky header feature, it seems that you plan to use whether a user is logged in or not as a proxy for whether they are a reader or an editor and which interface they should therefore be given. I understand the temptation to do this, but I think it's a fundamentally bad idea.

There is a large group of users who are logged in but do not identify as editors. This is good, since it allows them to customize their preferences, and it means that if they ever do decide to fix a typo or otherwise wade into editing, their edits will be associated with their identity, making it much easier to communicate with them and improving their newcomer experience (courtesy ping MMiller (WMF) for the Growth team). However, these users presumably do not want a ton of buttons for editing features, and if their first experience after creating an account is that they now suddenly have a ton of new buttons that clutter their interface and look like junk, that'll push them to just log back out.

I've long felt that we need to be much more deliberate about how we invite readers to become editors. In an ideal world, there would be a master user preference that could be toggled between "display no editing features", "display basic editing features" (e.g. the edit button; default for IPs), and "display advanced editing features" (e.g. the en-WP left sidebar tools section). As a more practical matter here, it's important that at minimum there be a preference option for users who don't want their user links to persist in the upper right as they scroll and that this be the default for non-autoconfirmed users.

There are so many ways in which the ideal experience for readers vs. editors differs, and the longer we keep using logged in vs. logged out as a cheap proxy rather than building a better system, the more debt will accrue and the more non-editing logged in users will be harmed. {{u|Sdkb}} ^talk 20:52, 22 December 2021 (UTC)

Hi @Sdkb -- thanks for the ping. I definitely see why newcomers came to mind as you were looking at the sticky header work. I'm following along closely with the progress, and I know that the Web team has newcomers in mind as they work on the desktop improvements project -- part of the objective is around helping readers intuitively understand how Wikipedia comes together so that some of them are driven be curious and learn more, and maybe participate themselves. I think that's part of a difficult balance you're touching on -- we know that the vast majority of readers will never edit, but we do want to make opportunities visible enough for those who might edit to try them out.

With respect to the actual sticky header work and their thinking about the different needs of readers and editors, @OVasileva (WMF) will be able to give you more details on the thinking and plans likely at the beginning of January once work ramps up again. MMiller (WMF) (talk) 21:04, 22 December 2021 (UTC)

Hi @Sdkb - thanks for your comment and questions. I too agree that we should move towards a more modular approach for individual features and needs, preferably by showing the user the interface options they are most likely to need. I see the sticky header as the beginning of this type of process - it collects many editing features in a way that is easier to understand and more accessible to newcomers. Another approach, as you pointed out, would be to create individual configurations for the sticky header and it's functionality. This is difficult technically as it requires the team to maintain many different versions of the page that we have to then test against when building out any new functionality. It also places a lot of pressure on the user to be able to identify where each setting lives and what its purpose is. However, we are currently discussing other potential options for configuration - for example, allowing individual users or entire wikis to be able to configure smaller portions of the sticky header via gadgets or a user preference (such as adding or removing links to functionality and features that are relevant to them). I'd love to hear your thoughts on this. Our hope is that it would be a more sustainable way to approach the question of configuration. OVasileva (WMF) (talk) 12:32, 4 January 2022 (UTC)

@OVasileva (WMF), thanks for the comment! I think that when left to the community, there's nearly always a significant bias for editor needs over reader needs. Logged in but non-editing users are a pretty much invisible group. That's part of why I think it's important to have something like the no/basic/advanced editing features toggle in the settings, since then when the community makes interface changes to benefit editors, we'd be doing so for ourselves, rather than for ourselves and also this other invisible group that might want something different. {{u|Sdkb}} ^talk 22:50, 4 January 2022 (UTC)

Add extra button to the WVUI search

On the Hungarian Wikipedia, there’s a gadget that adds an extra “open in new tab” button to the search widget. (You can try it by turning on the gadget Search in new tab / Keresés új fülön in your gadget preferences, or by loading the ResourceLoader module directly with mw.loader.load('ext.gadget.keresesUjFulon').) It works fine on legacy Vector, but on new Vector it disappears when the full (WVUI) search widget loads. How can I inject my button in the WVUI widget? —Tacsipacsi (talk) 13:57, 9 October 2021 (UTC)

Sticky Header

How can I use this new feature in a third party wiki?

Reading/Web/Desktop Improvements/Features/Sticky Header --89.35.180.240 22:58, 24 December 2021 (UTC)

We'll have better third party support soon for the new Vector. We will be creating a new skin. Please track phab:T291098 to follow progress. Jdlrobson (talk) 23:58, 5 January 2022 (UTC)

Unecessary "deadspace" on the sides of the site

Why is there empty space on the sides of the articles? It just gives off, to me, an unprofessional vibe, as well as squishes the actual contents of the article. What is gained from the unused space? --Lamaredia (Kasper D.L) (talk) 14:31, 26 December 2021 (UTC)

Please see Talk:Reading/Web/Desktop Improvements/Reading/Web/Desktop_Improvements/Features/Limiting_content_width. Jdlrobson (talk) 23:57, 5 January 2022 (UTC)

Example of "deadspace" on sides of article

Move other projects to the sticky header

Hi. I don't know if this is planned, however I leave the following suggestion which seems to me important. With the new sticky header which will soon be implemented —which I encourage— it would be very useful to move “other projects” (distinctly and Wikidata included) next to languages, for several reasons:

• over all, easier and more comfortable access;
• consistency with the languages tab, since it was moved;
• better recommend other projects to readers;
• “frame” templates, such as Template:Sister project links (Q5830969), create duplicate links, are heavy-looking, and often shift the layout: such a new tab would allow to improve the global rendering by making these templates useless.

Thank you in advance for looking into this issue. Best regards — Baidax (talk) 17:35, 4 January 2022 (UTC) (edited)

Sticky Header won't switch between language versions

I've just noticed that you've activated the new sticky header in my account on German Wikipedia. Thanks! I think I like it. However, I'm afraid the language selector won't work. I cannot switch between languages in the sticky header. The standard language selector at the top of the page, however, works nicely. I'm running Firefox 95.0.2 on macOS 10.13.6. Thanks and Best regards, Aschmidt (talk) 20:40, 5 January 2022 (UTC)

I think this is T297579 (which will hide the language button in the sticky header). If you want to use languages in the sticky header you will need to go to Special:Preferences and select "Use a compact language list, with languages relevant to you." in Appearance/Languages. Jdlrobson (talk) 23:56, 5 January 2022 (UTC)

Thanks, @Jdlrobson, this is probably true. I'm afraid, I'd rather not use the compact language list because, as an author, I prefer to have all languages displayed in the list. I'd like to see at a glance how many language versions of Wikipedia find a lemma notable. So, I'd prefer if you please could try and solve this issue. Apart from this, I like the sticky header. :) Kind regards, Aschmidt (talk) 00:40, 6 January 2022 (UTC)

@Aschmidt: If you want to see just the amount, I think the compact language links feature is actually good for you: if you enable it, the button on voy:de:Berlin says 21 Sprachen (using old Vector, it lists the 9 most relevant ones—it’s always 9, as long as there are at least 9 languages—, and the button says 12 weitere). So you don’t even need to count the links, the software does it for you. —Tacsipacsi (talk) 12:42, 10 January 2022 (UTC)

Thanks, @Tacsipacsi, for your hints. I'm afraid, I would prefer to be given the direct links to all other language versions, too. I keep the so-called compact language list switched off. Best regards, Aschmidt (talk) 16:42, 10 January 2022 (UTC)

How to disable sticky headers?

Sticky headers really annoy me. There is absolutely no benefit, but just wasting useful space. How can I disable them without switching back to legacy Vector? --Bombenleger (talk) 18:16, 6 January 2022 (UTC)

I also immediately noticed it and came to comment my disdain for sticky headers. My browser already permanently reserves like 3% of my entire screen, now Wikipedia will double that. I don’t need a instant access to any of the buttons in the the header. I search Wikipedia through google, not Wikipedia and that search button is already part of the 3% screen space my browser reserves. I don’t need to be permanently reminded what article I am reading. Most people only speak one language, they do not need to have access to the language switcher 100% of the time. I only check my watchlist once, I do not need access to it 100% of the time, that logic applies to the other buttons in the top right.

Sticky headers are a diet version of toolbar hell from the Internet Explorer days. I really really don’t understand why people keep doing them. Akeosnhaoe (talk) 11:12, 10 January 2022 (UTC)

@Bombenleger, @Akeosnhaoe, you need to add

.vector-sticky-header {display:none;}

to your global.css. SGrabarczuk (WMF) (talk) 13:24, 11 January 2022 (UTC)

thank you @SGrabarczuk (WMF), I love it! Bombenleger (talk) 17:27, 11 January 2022 (UTC)

Sticky header zero language switcher

Hi. Is this really a good idea to create a button "0 languages" that opens a huge empty popup on click on pages with no interwikis? IKhitron (talk) 16:15, 7 January 2022 (UTC)

For that matter, I don't see why the language switcher should be prioritized for the sticky header at all. If a user has navigated to a page and read through it enough to scroll, chances are they're at the language they want to be at. They don't need a prominent button to switch languages from that point. {{u|Sdkb}} ^talk 19:55, 7 January 2022 (UTC)

Courtesy ping SGrabarczuk (WMF) and OVasileva (WMF) {{u|Sdkb}} ^talk 19:58, 7 January 2022 (UTC)

• I think the feature should be scrapped. Apparently it leads to drop of uses of Wikipedias in secondary languages (which ones they are vary by country/region, probably not important for the US). --Jura1 (talk) 20:55, 7 January 2022 (UTC)

The 0 languages is a bug we identified early in testing. That's not meant to show and will be fixed in this week's deploy.

Tracked in Phabricator
Task T298705

Jdlrobson (talk) 03:02, 11 January 2022 (UTC)

Great, thanks. IKhitron (talk) 00:20, 12 January 2022 (UTC)

A problem of cache or someone else?

(from fr.wiki 1 and 2)

Hi, reporting my personal experience, in differents pilot wikis (fr.wikipedia, fr.wikiquote, vec.wikipedia) I was not able to see the sticky header for a while, I need to change accounts, log in and log out 3 times and then refresh the page a lot of times before to see it appear and work. 5th and 6th January I was not still able to see it. Now I can. Actually, there are users on Wikipedia in French (and Wikiquote in French for an individual tester) that can't see the sticky header, Malik2Mars, Paul.schrepfer and Daehan (feedback of 9th January). One of them (Daehan) was able to use the sticky header active when deployed (the 5th January), now he can't, this is strange. There is a reason for that? Thank you for answering.--Patafisik (WMF) (talk) 10:25, 10 January 2022 (UTC)

There's an A/B test running, so 50% of user accounts will not see the sticky header. When that test finishes run you should be able to use it. There was a period while we set up the A/B test where you may have seen it if you do not see it now. Jdlrobson (talk) 02:58, 11 January 2022 (UTC)

Sticky header: issue with fixed-width space for page title

The space dedicated for page title is always 500px wide. This number is fixed and therefore long titles will be cut even on wide screens which may look awkward. On the other hand, when the screen is really narrow, the title stays 500px wide and pushes buttons out of the screen. Here is an example. Msz2001 (talk) 15:32, 10 January 2022 (UTC)

Tracked in https://phabricator.wikimedia.org/T298885 Jdlrobson (talk) 02:58, 11 January 2022 (UTC)

Sticky header: language switch button wraps on narrow screens

When using a narrow screen, the language switcher wraps and renders as two lines of text. This looks ugly and IMO should be changed so that the content is always displayed as a single line. Example is here (second image). Msz2001 (talk) 15:35, 10 January 2022 (UTC)

Tracked in Phabricator
Task T298887

Jdlrobson (talk) 02:59, 11 January 2022 (UTC)

Talk page and page history

Tracked in Phabricator
Task T299115

Any reason why sticky is only available on the main page and not on a talk and history pages?

Seems a bit weird to me. Both talk and history have links on the sticky header, but when you navigate to them then the sticky is gone. Especially when I visit someone's talk page I would like to be able to navigate to his/her user page. And talk pages can be quite long. Also if I get an answer here I will get a link to the bottom of this page. And I should see a sticky to be able to open the main page in a new tab or view other notifications. Nux (talk) 22:30, 11 January 2022 (UTC)

Hey @Nux!

1. Talk pages - that's not entirely up to us. The Editing team is responsible for talk pages - see their current project Talk pages project/Usability. They will make the decision whether and how the sticky header will be implemented on talk pages. We're working closely with them on that issue.
2. History pages (also, special pages) - at first, we decided not to enable the sticky header there because it doesn't offer that much functionality (there's no option to edit that page, for example). We'll consider doing it, though. There will be a dedicated Phabricator task.

SGrabarczuk (WMF) (talk) 20:49, 12 January 2022 (UTC)

Keyboard shortcut does not focus search field in sticky header

When pressing the keyboard shortcut for search (alt + shift + f) the search field at the top of the page is focused, even if the sticky header is showing. It would be better if the search field in the sticky header was focused. As it is now the page will scroll to the top which is unnecessary. Sebastian Berlin (WMSE) (talk) 16:10, 12 January 2022 (UTC)

Scrolling by pages + sticky header scrolls too far

Steps to reproduce:

1. Use Firefox. (Chromium might behave the same, but I haven't tested.)
2. Scroll to the top of the page, so that the sticky header isn't visible.
3. Note where the text is cut off at the bottom, by the edge of the viewport.
4. Hit the 'page down' key.
5. Wait a moment for the the sticky header to appear.
6. Note where the text is cut off at the top, by the sticky header.

Expected results:

There would be some overlap between the text visible at the bottom before scrolling and the text visible after scrolling, so that you don't need to manually scroll up with arrow keys to re-find your place.

Actual results:

The sticky header obscures the overlapping text and effectively scrolls too far, going past some text that was not visible before to scrolling.

Request to deploy the sticky header in all the namespaces in Vietnamese Wikibooks

Hi, the Vietnamese Wikibooks community is requesting the Web team to deploy the sticky header in all 3 content namespaces. I'm not really active at the Wikibooks so I don't understand why there are 3 main namespaces in it, but they are asking for it. Could it be possible for the team to do this? I'll tag the requesting person for more follow-up info if needed: b:vi:user:Đức Anh. Bluetpp (WMF) (talk) 11:10, 20 January 2022 (UTC)

Page-title and Search

In a nutshell, I believe

• the Search would benefit from being (a) more consistently placed, and (b) more accessible from the sticky header by having a larger click-target.
  • See image at File:Vector-sticky-header-version1-search-size.png (and more details above)
• the Page-title is already visible in multiple locations, and might not need this additional instance? AFAIK only users with "Fullscreen - F11" enabled would hide the existing locations where it is already displayed, whilst scrolled.
  • See image at File:Page_title_repetition_and_cutoff.png

Hope that helps! –Quiddity (talk) 20:53, 21 January 2022 (UTC)

Drop down menus

If I want to protect or delete a page, or something else, I have to make an extra click to the "Page" actions menu in order to collapse it. I usually delete thousands of pages each year - this means I have to click thousands more times. I am finding this extremely time-consuming. If there is any way to change this, I would really appreciate it. —user:Hasley 23:33, 22 January 2022 (UTC)

regarding Reading/Web/Desktop Improvements/Features/Limiting content width and transclusion

If recent changes are transcluded this change the page width. Although I appreciate the width-limit and I like the special treatment of special pages, this seems to be not really consistent. I don't really know what could be done about this. Could there be a switch within the transclusion to specify which behaviour is wanted? Though the programming-effort for this will be extremely high, I suppose, because transclusion is effected and not the mere interface. I don't know. Probably it's best to keep it as it is, because it's not a bug, but a (visual) inconvinience? Regards HirnSpuk (talk) 19:11, 23 January 2022 (UTC)

Addition: It looked like it when I hit "show preview" but it's shown correctly after publishing the page. At least for now. Maybe the effect is connected to: phab:T270802? If I notice anything else, I'll post more info. Regards HirnSpuk (talk) 23:07, 23 January 2022 (UTC)

+1: Same happens, when transcluding {{special:prefixindex}}. Slowly it seems to me being a bug? Way to reproduce: edit a page, transclude a special page, show preview. Tested with prefixindex and recent changes. Firefox 96 (64-Bit), no gadgets and stuff. At least it seems not critical. Regards HirnSpuk (talk) 23:16, 23 January 2022 (UTC)

Nice to have: personalized background

Hi, there is my "nice to have" wish. We are discussing about background color of new modern Vector skin. In addition of this, you could leave users personalize their background. Possibly with all images from Commons. Or, if there are technical constraints, you could suggest some background colors and some patterns like those ones or this one (without logo and text) from which to choose. Benefit: no more problems with grumpy comments for the limited width and too much white space.--37.103.19.52 09:16, 24 January 2022 (UTC)

Claims about the current interface

I fully agree that the desktop interface doesn't match the expectations of modern web platforms. Wikimedia has always looked quite outdated. But to be honest that's less to do with the structure of the web page around articles, and more to do with the design of articles themselves in my opinion. We can only create plain boring tables, not beautiful tables and articles.

"It is challenging for readers to focus on the content." According to whom? I've never heard this complaint about the desktop experience before.

Completely agree that many people don't know how wikis function. I'd welcome changes to the interface that welcome new editors and make it easier for them.

"isn't consistent with the mobile version", this is worded as though the mobile version is somehow the default. The tone changes if this was written as "The mobile version is not consistent with the desktop version" or "The two versions are not consistent". I just want to be cautious that we shouldn't change the desktop one to match mobile just because it's easier than the other way around when that may not be the best thing for a desktop experience.

Consistency does not meant they have to look exactly the same and function the same. We should not lose the advantages that a certain platform provides (i.e. a wide screen on desktop).

Bit concerned about the claims made in this section. Supertrinko (talk) 01:37, 25 January 2022 (UTC)

This is not an improvement

At least not at this stage. I just had the misfortune of encountering the "improvements" on the Persian Wikipedia and I have to say it is utterly horrendous to try and use. Aesthetically and functionally displeasing, the new skin does not represent an improvement over the status quo at all. If this change is to be enforced I sincerely hope it is not done in its current state. This is possibly the single most unnecessary set of changes I've seen in my two and bit years of contributing to and decade plus time using Wikimedia projects. I could complain about a lot of the proposals I've seen on this page, but in essence, why are desktop using being given a mobile-esque interface? 5225C (talk) 03:52, 25 January 2022 (UTC)

i agree. after the "improvements" users would have to click a few more times to get to the same links = it's a waste of time and effort. RZuo (talk) 08:19, 25 January 2022 (UTC)

Side panel on Wikidata item pages is forced to bottom of page

@SGrabarczuk (WMF) and OVasileva (WMF): Hello, not sure exactly what the state of deployment is for this (i.e. if all latest changes are deployed everywhere), but was trying it out on Wikidata (where I primarily edit) and the div classed as wikibase-entityview-side on item pages (which houses all the sitelinks) is pushed down to the bottom of the page (see wikidata:Q1 for example).

In legacy vector this element is at the top of the page on the right hand side, which seems more appropriate given the importance of the information (as a means to access the subject on other wikis) and the clearer separation from the data statements. I don't think it necessarily has to live in the same place, just that the current placement seems incorrect, especially considering how much scrolling is needed to reach it on some item pages. --SilentSpike (talk) 12:25, 25 January 2022 (UTC)

@SilentSpike: Oh yeah, the infamous limited content width. On my 1366×768 screen (minus a few pixels of vertical taskbar), sitelinks at the bottom with legacy Vector as well. The limited content width of new Vector forces this layout even on gigantic iMacs. (By the way, you can see the very latest version on beta cluster, every change that’s been merged immediately appears there.) —Tacsipacsi (talk) 13:19, 25 January 2022 (UTC)

Feedback regarding the new language switcher

Hello, I am a French person using Wikipedia in both French and English. I usually decide the language of the article depending on the topic. If it's related to France, I know the French version is going to be better. If it's a general topic, I usually select English.

However, it's been a few months that the language picker has been moved to the top of the screen when an article is display in French. And I really can't get used to it. Sure, it's easy to discover - but it's more complicated to use. Having to click twice is a regression for anyone switching language frequently. An action that used to require half a second now takes 2 seconds. If you do this about ten times every time you browse Wikipedia, it does become irritating...

I hope this feature is not shipped to other languages, and France gets the sidebar back. If you want to make this feature more discoverable, maybe you could add a shortcut but keep the sidebar. Or display a tooltip once in a while to educate newcomers. But please, keep "heavy users" of language switching in consideration. Thanks