Jump to content

Talk:Product and Technology Advisory Council/February 2025 draft PTAC recommendation for feedback

Add topic
From Meta, a Wikimedia project coordination wiki

Please leave any feedback on the draft PTAC recommendation (or its subpages) here. GorillaWarfare (talk) 00:11, 5 February 2025 (UTC)Reply

I think making mobile editing better is a valid goal, though tbh even desktop editing is not really what we should have in the modern day and age (yes, there is 2017 wikitext editor, no, it has too many flaws to count).
But mobile editing in particular is currently very lacking, so I am not surprised people do not contribute a lot using mobile.
1) Actual editing interface (buttons etc. which make the life easier for editors, especially on mobile, where entering links etc. is much, much harder) is only available in visual editing mode, which isn’t shown by default and isn’t many people’s preferred way of editing,
2) the UI is hard to use on a mobile device since all the buttons are located at the top and not at the bottom, out of reach for typing with fingers,
3) you still cannot edit the whole page from anywhere unless you’re logged in, even a small page,
4) it is currently hard to see the context of where in the page you are editing, leading to things like phab:T377611.
These are just some things I could list off the top of my head about this. It might also be interesting to explore some new editing workflows on mobile, for example, while I think mw:MiniEdit is useless on desktop (at least when compared to w:User:BrandonXLF/QuickEdit), something like that seems like a good way to encourage smaller-scale edits from mobile users. The interface does not need and probably shouldn’t be the same as MiniEdit, but even having an affordance to open the editor with a focus already moved to a specific paragraph in a section could go a long way to improve mobile participation.
I also don’t think having some vandalism as a result is necessarily a problem: I feel like in general Wikimedians are pretty good at accepting QOL improvements that lead to more helpful edits even if they come with some vandalism on the side. See for example phab:T54165#10439191: Wikimedians in multiple communities wanted mobile version to have the same ease of experience as desktop in regards to participating in discussions, even if it lead to an uptick in low-effort comments. stjn[ru] 01:33, 5 February 2025 (UTC)Reply
  • I'll keep it short and simple. I agree with your conclusion that improving the process of mobile editing is the most likely to bring longterm dividends to the Wikimedia projects of the future. Risker (talk) 06:05, 5 February 2025 (UTC)Reply
    I agree with Risker. I'd be happy to offer other feedback but as someone said offwiki in a comment I agree with the recommendation is both so broad and so specific that I have no useful feedback. Best, Barkeep49 (talk) 15:24, 5 February 2025 (UTC)Reply
Will Wikidata mobile editing also be in the focus of the council? Sjoerd de Bruin (talk) 07:50, 5 February 2025 (UTC)Reply
Not particularly as a focus area for this budget. This does not exclude other teams or organizations (WMDE) of working on it of course. —TheDJ (talkcontribs) 09:48, 5 February 2025 (UTC)Reply
Indeed, WMDE included "improving mobile editing experience on Wikidata" within their 2025 yearly goals, released in Q4 2024. Poetest (talk) 17:13, 25 February 2025 (UTC)Reply
My personal opinion is that focusing effort on improving editing on mobile mobile editing is a waste of resources.
Mobile devices (cellphones/tablets) are ill suited to content creation, and are by nature better suited to consuming content.
I feel like the reason most people do not contribute via a cellphone is not because the current tooling provided by Wikimedia is blocking them, but because the inherent nature of the platform is.
Wikimedia is not going to solve that problem. So rather focus efforts on where editing actually happens i.e. on a real computer like a laptop or desktop PC that has peripherals, and user interfaces actually designed to help people create content.
Focusing on the strong area will yield greater results, rather than trying to improve a lost cause. - Rooiratel (talk) 13:43, 11 February 2025 (UTC)Reply
@Rooiratel Instagram, TikTok are two platforms that thrive on mobile centric platforms and perform complex creative tasks. The vast majority of experiences with Wikipedia are indeed consumption related and even there, mobile could be improved. For example reading/finding talk pages (not even editing) are quite challenging in mobile Shushugah (talk) 21:51, 24 February 2025 (UTC)Reply
None of that changes the fact that it is way easier to perform complex creative tasks on a desktop platform. It also doesn't change the fact that given the choice people wanting to perform complex creative tasks, they would choose a desktop platform. So my point still stands.
Most "actually complex" videos on Instagram and Tiktok are edited on a desktop platform, and then just uploaded to Instagram or TikTok. Adding some filters and basic cropping of an image or video is not complex. In the same way, making small text changes or adding an image is not complex editing on Wikipedia, and is fine for doing on mobile. But for actual complex work: advanced template editing/creation, Lua modules, merging/splitting pages, etc. I don't see how a mobile platform will ever cater to those needs in a satisfactory way, no matter how many refinements are made. - Rooiratel (talk) 09:01, 26 February 2025 (UTC)Reply

This feels very vague. Its difficult to say much about it. Its not even really a pitch. For the rejected options, its not even clear what "image upload" and "admin satisfaction" means in this context. My biggest concern is that both "Mobile experiences" and "New editor experience" are things that WMF has sunk a lot of money into in the past, much more than the amounts being talked about here. Is $300,000 - $500,000 enough to even move the needle on these objectives? How many FTEs is that even? 2? 3? Perhaps if our approach was modified, but there isn't very much here in the way of that. I feel that without a modified approach or clear plan of action on how we are going to do these things, the impact of focusing on them will be minimal. Unless we have a clear and feasible plan about how to make high impact changes in these areas, I think it would be better to focus on areas with more low hanging fruit. The theoretical impact might be lower, but the actual impact would be higher. Bawolff (talk) 18:28, 6 February 2025 (UTC)Reply

+1. I'm not surprised you want to focus on mobile and new editors, as I feel like that's been the focus for the past decade (witness the many 10-year desktop Phabs that continue to sit unloved). I have a love-hate relationship with my phone. It's a sub-$100 4G model, I don't understand why I would need 5G or should spend $1000 on some fancy device that would be a theft target. I barely manage to keep the spammers at bay in my desktop email, so I don't even try to do email on my phone. I feel like email on a phone would be an impossible disaster. Texting is a fat-fingered experience where I frequently make typos and the spell-checker changes the meaning of my message by "correcting" my typo to make it a different word than intended (curse the damn thing). Every year the percentage of texts and calls to my phone that are spam or fraud increases despite efforts of my carrier to screen them out many still get through. It's a good thing my phone is well-built and wears a protective case because it needs to survive my fits of anger when I though the damn thing to the ground. I used mobile on Wikipedia once when I was bike riding in Hawaii and didn't have access to even a laptop, but I can't imagine why anyone would want to routinely use such a device when they had other options. Where I love my mobile and it shines is in mapping, navigating and tracking my bicycle rides. But I can't imagine that the Google Maps developers and mapmakers do their work on phones. No, I picture them working on multi-monitor desktop workstations. I'd be interested in watching YouTube videos demonstrating how to effectively edit Wikipedia on mobile devices to show me how it's done, and don't do it too fast or you might lose me. Wbm1058 (talk) 16:56, 9 February 2025 (UTC)Reply
I hear what you're saying. However, you speak very much from a place of privilege. You obviously have multiple electronic devices that you're able to access. Most people outside of the Western world don't. They have a phone, and that's it. It might be accessorized with a side keyboard. But a desktop, a laptop, even a tablet is far beyond the means of people in countries where the average annual income is less than one week's pay for a typical European or North American. We *want* people from other cultures to participate and contribbute. We have a functional suite of desktop software; it can be better, but it works extraordinarily well given the number of complex bells and whistles people keep wanting added on. We need to get mobile editing (whether using the desktop interface, or the separate mobile interface) working as well as possible, because our new editors are coming from parts of the world where a desktop or laptop is not only a luxury, it's widely seen as unnecessary. Risker (talk) 17:09, 9 February 2025 (UTC)Reply
OK, I hear you too. I'll presume that most of your target audience only affords sub-$100 phones like mine and don't have expensive iPhones, though I suspect that a subset of the better-educated population in places like India can afford and do use better devices. I occurs to me where the mobile breakthrough will happen. We've all heard Alexa answer verbal questions with amazing proficiency. You don't need a desktop for voice recognition, a phone is all you should need. I recall 30 years ago when my ex was trying in vain to convince her employer to give her Dragon Dictate to help her ward off carpal tunnel. They just didn't think voice recognition was ready for prime time. But maybe it is now. Can I edit Wikipedia simply by talking to my phone and telling it what to do? That's where this device makes its breakthrough, when I no longer need to use my fat fingers. Can the Foundation make this happen with its budget, or is something closer to the budget for ChatGPT going to be needed to make it happen? Wbm1058 (talk) 18:04, 9 February 2025 (UTC)Reply
I tend to agree that if we are ever going to succeed on mobile then we need to think outside of the box on user input. I think https://www.cursorless.org/ is a very interesting project on the voice editing front. https://jenson.org/text/ is a really good article on why text editing on mobile sucks in some fundamental ways that are out of control of the website. Personally, i do agree with the vast potential opportunity mobile represents, i'm just doubtful this investment will bring us any closer to grasping it. Bawolff (talk) 18:28, 9 February 2025 (UTC)Reply
@Bawolff -- thanks for the links. I appreciated the Jenson article, because it puts words to something @RHo (WMF) have talked about a lot -- that in order for people to meaningfully contribute on mobile, we have to build workflows that make sense on mobile (as opposed to trying to squeeze the desktop editing experience onto the small device). That's why we think structured tasks are a big part of the future here. MMiller (WMF) (talk) 22:22, 24 February 2025 (UTC)Reply
Having spent 2023-4 as a carer dawn to dusk, with only my phone to fall back on, I totally agree with this priority. Such editing as I have done has been frustrating, like painting a mural through a letterbox. The basic tooling is there, but action such as wikilinking policy docs, pinging, etc, are a pain of switching subkeyboards.
While my editing (AFD etc) is atypical, editing on the phone must surely be inhibiting other users in a mobile-first(/only) world. So this is the correct priority but it bothers me that it currently lacks focus on particulars.
A couple of suggestions:
  • Improving the editing experience for occasional editors, such as the suggestion above about being able to chisel at a paragraph for better screen fit.
  • Equality of community participation tools:
  • Mobile shows alerts in a red circle at the top, which it didn't until recently, but can omit some discussions you are following? My perception anyway, based on a stack of notifications when returning to desktop.
  • Topics on a busy, sectioned discussion such as Village Pump sometimes display a grey rectangle but don't deliver the text. Probably a phone/browser memory issue, but it would be good to ensure developments are friendly to lower-spec devices.
  • And my gripe... The mobile view of Talk pages invites the mobile user to "Start a discussion" but hides the contextual project assessment templates, old PROD and AFD info. Why?
Stopping there, as getting too far into detail. AllyD (talk) 09:43, 11 February 2025 (UTC)Reply

Hey ChatGPT, Is it possible to edit Wikipedia using only voice commands on a mobile device such as a smartphone?

Yes, it is possible to edit Wikipedia using only voice commands on a smartphone, but there are some limitations and steps involved:
  1. Voice-to-Text for Editing: Most smartphones have voice-to-text functionality built into the keyboard (such as Google’s Gboard or Apple’s iOS keyboard). You can use this feature to dictate text for the article you're editing on Wikipedia.
  2. Wikipedia Mobile App or Browser: You would need to use either the Wikipedia mobile app or a mobile browser to access the Wikipedia page you want to edit. Both support text editing, but you’ll have to tap into the editing box to start typing or dictating your changes.
  3. Voice Commands for Navigation: If you use accessibility features like Voice Access on Android or Voice Control on iOS, you can also navigate through the app or browser, select edit buttons, scroll, and even select specific sections of the page using voice commands.
  4. Editing Process: Once you're in the edit mode, you can dictate your text, make changes, or even copy-paste text using your voice commands, depending on the apps and settings.
While you can get quite far with voice commands for navigation and text input, you might still need to do some manual adjustments (like formatting or adding citations) depending on the complexity of the edit.

Sounds complicated. Maybe this comes naturally to kids growing up with this technology, but it doesn't to me. I grew up in an era when products still came with manuals, and you weren't required to figure everything out by guesswork trial & error. I'm still exhausted from my efforts to become a MediaWiki hacker, so don't feel motivated to try this without some serious guidance. Create some high-quality YouTube videos showing me how, and make them really engaging so that they go viral and get pushed into lot of people's YouTube feeds. Wbm1058 (talk) 12:25, 11 February 2025 (UTC)Reply

First thing I would do is to stop purposely butchering the mobile experience. One example: links to "sister projects" are totally absent for mobile users. That's a nasty way of treating your "family". Unfortunately, WMF has move toward homogenizing the experience, by hiding the sister project links in the new Vector skin. (phab:T334792) --Ignacio Rodríguez (talk) 14:02, 11 February 2025 (UTC)Reply

Sorry, this page is a sort of stream of consciousness (which is hopefully helpful to the Council in some way) to which I can't help myself from adding another thought. My second-generation Samsung smartwatch told me it was end-of-life yesterday (its face came unglued) so I just upgraded to the current 7th-generation model. I just noted one of what I'm sure are many pointless changes. The second-generation's power button is at 4 o'oclock and its back button is at 2 o'clock; in the 7th generation, these buttons were reversed. Why, I have no idea. I recall when I got my first cellphone investing a lot of time into learning and understanding all its features and how they worked. Then I got a mandated upgrade and everything was different so I have to go through the learning session all over. By the time I got my third and fourth phones, I gave up and decided to not bother investing a lot of effort into learning stuff that would only be of transient value to know. Imagine the next generation of automobiles reversed the position of the brake and gas pedals and replaced the steering wheel with a joystick. Wbm1058 (talk) 15:25, 11 February 2025 (UTC)Reply

  • I think mobile editing is most suited to quick typo corrections, linking, patrolling and discussions on talk pages, and the latter two need improved tools. You can't really expect people to use it for content creation. My two pet peeves with mobile editing are the inability to display heavily indented text on talk pages (phab:T116686), and the difficulty in finding the undo button. Femke (talk) 18:50, 13 February 2025 (UTC)Reply
    So y'all have decided to focus on mobile, in particular, the mobile front end, in order to make it easier for mobile users to contribute. Fine (not really my priority, but OK). The next step is to descend from the clouds and survey the surface looking for good landing sites. Phabricator's mobile homepage shows seven mobile front-end tasks, most having been submitted five or more years ago. This task isn't one of them. No, it's one of the 80 tasks still needing triage. I'd assume these tasks should be triaged in order to determine where to put them in the list of priorities. phab:T116686 has been waiting for triage for over nine years. Wbm1058 (talk) 14:11, 17 February 2025 (UTC)Reply
    @Femke: As an anecdote, I would like to note that I edit exclusively on mobile and have created several articles. QuicoleJR (talk) 15:07, 25 February 2025 (UTC)Reply

I was reminded of this comment from nearly four years ago:

I do 99% of my editing on Android smartphones, using the desktop site, which is 100% fully functional on modern mobile devices. The only time I sit down at a desktop computer is to work on large image files. We would all be better off if the WMF shut down all these poor quality smartphone/mobile apps, which are an impediment to collaborative editing. I cannot imagine the amount of money that has been wasted on these crappy apps over the years, but "small fortune" comes to mind. User:Cullen328 20:12, 15 March 2021 (UTC)

This gets to my point about steering a car with a joystick. If the user interface works basically the same on a phone as it does on a laptop, then I don't need to suffer the learning curve of figuring out phone editing. Do you have any statistics on what percentage of mobile editors do it Cullen's way versus using a mobile-specific app? Wbm1058 (talk) 12:22, 21 February 2025 (UTC)Reply

Maybe we're due for an update on the years-old Advanced mobile contributions feature set that adds more contributor features/capabilities to the mobile web experience. Why hasn't that led to an increase in mobile contributions beyond around 12% of successful edits? Wbm1058 (talk) 15:49, 21 February 2025 (UTC)Reply

As a long-term contributor who edits exclusively on the mobile website, it would be nice if the mobile view actually had all of the features work properly. As just a few major examples, mobile view users cannot mark edits as minor outside of reverts, and for some odd reason cannot even save page creations without switching to desktop view! (the latter is probably a glitch) Fixing issues like these would go a long way to improve the mobile editor experience. QuicoleJR (talk) 15:12, 25 February 2025 (UTC)Reply

This editor has posted several messages to my talk. They're using VisualEditor on mobile so seemingly in your target market for development. However they have yet to make significant positive contributions in English mainspace. If you assume they're trying in good faith, maybe someone here can try to help them. My time is oversubscribed; I'm working on another bot, and this editor keeps interrupting my work flow. Wbm1058 (talk) 17:19, 28 February 2025 (UTC)Reply

This mobile editor has managed to change their user name from Amogelang22 to Lekgau tshiamo but has yet to figure out how to make productive article-space edits. Go figure. Maybe they're a Russian agent trying to disrupt us, as they've disrupted the Federal Government. Wbm1058 (talk) 12:38, 8 March 2025 (UTC)Reply

For topic 4 it would be great to see some integrated tool with functionality of VicuñaUploader as a result. And another consideration (let's call it topic 5). While Wikipedia editors are divided in their opinion on the implementation of AI in Wikipedia, it seems impossible to keep it AI-free in the long term. Maybe it's time to think about some beginner-friendly general framework to integrate AI tools to use Wikipedia for all purposes, including input and output of data, different inner tasks. Not ready-made instruments that will become obsolete in 12 months, but only an infrastructure for these instruments that will last longer, some kind of "AI Toolforge" with API for both web-based chatbots and local LLMs. AI models are managed by relatively simple text prompts which are much more understandable to the general audience (i.e. non-programmers), so I think that number of users of such tool will be much bigger than the number of bot owners and active users of special instruments. Homoatrox (talk) 11:48, 19 March 2025 (UTC)Reply