Wikipedia talk:Media Viewer/June 2014 RfC
What now?
[edit]The RFC has been closed, so what happens now? Does this intrusive piece of cruft get rolled back? — Preceding unsigned comment added by 69.140.0.55 (talk) 23:00, 9 July 2014 (UTC)
New features based on your feedback
[edit]Hi folks, thanks for all your helpful feedback about Media Viewer in recent days. We really appreciate your candid recommendations on this page — and survey comments also confirm many of the issues you’ve raised.
The multimedia team is taking your feedback to heart, and we are sorry for any inconvenience caused by this tool. To respond quickly to the most frequent requests, we have now pushed back other projects to focus on Media Viewer for the next few weeks.
Here are some of the new features we are now developing for you, based on community suggestions.
1. Disable Media Viewer quickly:
- Instant Opt-in: A more prominent way for registered users to disable Media Viewer, without having to go to preferences. (#703)
- Opt-out for anons: An easy way for anonymous users to disable Media Viewer, using localstorage. (#704)
2. View images in larger/different sizes:
- View original file: A prominent button to open the original image in your browser, so you can zoom in to see its details, or download it for re-use. (#630)
- View different sizes: Prominent links to view images in different sizes from the Download panel, so you can open them in your browser. (#664)
3. Discover image information:
- Make it easier to find image information: Provide clear visual cues that more information is available, with links to open the metadata panel. (#706)
- Scroll down to see more info: Use either up or down arrows to open the metadata panel below the image, to make it easier to find. (#697)
4. Edit / Learn more on Commons:
- Show Commons link to logged out users: Show a prominent link to the Commons file page to all users, so they can learn more about this image. (#429)
5. Learn to use Media Viewer:
- Add tooltips to Media Viewer: Show more tooltips in Media Viewer, so that users can tell what each button will do. (#546)
You can view more details about these features on this development planning site.
We are working hard to get these changes completed by tomorrow, so we can test them before releasing them to production. If all goes well, we expect to deploy some of them to the English Wikipedia and other Media Viewer sites by Thursday evening. The rest of them will be deployed the following week.
Please let us know what you think. Which of these features seem most useful to you? Are there other critical features that you think we should consider next? We look forward to improving Media Viewer based on your feedback. Fabrice Florin (WMF) (talk) 00:33, 11 June 2014 (UTC)
- We shouldn't have to need tool tips to know what each button does. Mystery meat navigation is by definition bad UI design, and completely breaks when you're using a device without a mouse, a tablet, or a phone. --Ahecht (TALK
PAGE) 02:41, 11 June 2014 (UTC)
- Since Media Viewer breaks the site's look and feel as well as the workflow I recommend disabling it completely. When you offer to make buttons larger or provide tooltips in order to make the product more useful you should rethink your premise as well as your process. A good desgin doesn't need tooltips, it has descriptive link texts (and not non-obvious icons) or uses an interface that is already known to everyone. Assuming that users want to learn or discover an interface is setting yourselves up for a lot of critique. --Millbart (talk) 06:45, 11 June 2014 (UTC)
- I forgot to add that there are several critical features that are missing from your list: click-and-drag to pan, scroll wheel to zoom, a method of excluding the image from display using the new image viewer in the article markup, and the inclusion of licensing data for all images despite what template or method for indicating the license is used. --Ahecht (TALK
PAGE) 13:50, 11 June 2014 (UTC)
- Thank you all for your prompt responses about these new features and other proposed Media Viewer improvements. The multimedia team just rolled out a couple new features today that may address some of your concerns:
- Thank you all for your prompt responses about these new features and other proposed Media Viewer improvements. The multimedia team just rolled out a couple new features today that may address some of your concerns:
- These features can be tested on Commons today (see sample image]) and should be released on English Wikipedia tomorrow, if they test well.
- Ahecht, I hope that #511 will address your request for 'a method of excluding the image from display using the new image viewer in the article markup'. That one should be completed by next week, if all goes well. And now that the 'View original file' button is implemented, you now have the same browser zoom features as before. Millbart, we appreciate your comments about tooltips, but have observed they are commonly used in tools like these, and have been frequently requested by other community members; we think they will help casual users who are not as tech-savvy as you. We will review your other recommendations above and keep you posted on our progress with these requests. Keep in mind that we are triaging a number of requests from other users, so may not get to them all right away; but they are on our radar now.
Thanks again for sharing your concerns, which we are taking to heart. Be well. Fabrice Florin (WMF) (talk) 22:20, 11 June 2014 (UTC)
- Please fix how Media Viewer handles long image descriptors. The map for same-sex marriage in the United States had a nicely ordered key first in English, then in other languages alphabetically (old-style file page) and now with Media Viewer, all the formatting is gone, languages and color boxes blend together into an unreadable and unsearchable mess (now, "improved" with MediaViewer). Beyond this, please accomplish the goals to disable Media Viewer quickly (and permanently) for logged-in or not logged-in users - everytime I see an image in it, I become absolutely livid at how unhelpful the new viewing style is. The option still exists to can this project, and I would urge that as the best course of action. - S201676 (talk) 08:39, 12 June 2014 (UTC)
- This is one of my core objections. When I called Media Viewer "half-baked" above, this sort of behaviour is what I was referencing. I would be much happier with Media Viewer if it simply placed the file description in its place in the interface directly, as rendered from wikitext, because we see unintended consequences like this if the descriptions are modified. My read on this is that the Media Viewer is overreaching—in particular the way it seems to treat wikitext pages as though they were safe, "semantic" data, when they're anything but. Dialling back the design in favour of more consistent function seems desirable. {{Nihiltres|talk|edits}} 15:43, 12 June 2014 (UTC)
- Fabrice: Wow, simply wow! I have pointed out that tooltips are a sign of weak interface design and you explain their existence with your (anecdotal?) observation that people demand them, especially casual users who aren't as tech savvy as I am, thus proving my point. Why do people demand tooltips in some GUIs? Could it be, because the GUI is non-intuitive, obfuscated and/or different from everything already learned? (I know, that every interface apart from the nipple is acquired knowledge, but some are easier than others) Why don't you use descriptive links? What makes you think that a logo is better than text when linking to the file description? My point is: Even the tech savvy user needs tooltips with your GUI design because it is so non-obvious.
- I realise that overlays and modal windows like lightbox are all the rage, but they are very rarely more than a nuisance. Here, it breaks the look and feel as well as the site's workflow, because it is unexpected behaviour. The site normally doesn't use modal dialogs. Interestingly you choose to ignore this criticism. As it is, Media Viewer doesn't improve anything at all. Please, make Media Viewer available as an optional gallery view link in the toolbox or wherever ("View all the article's images") if you really think that it is useful for something, just don't force it on everybody as the default, especially if it can't be disabled globally. --Millbart (talk) 09:00, 12 June 2014 (UTC)
- To Millbart, I just want to say that I think you very eloquently expressed reservations and concerns that I share and so I want to thank you for articulating this so well. I think the statements and actions of Fabrice Florin (WMF) in cynically seeking to invalidate and delegitimize the clear chorus of English wikipedia community members who have reached a consensus that MV should be opt-in by default = predictable, but utterly shambolic nonetheless. I would symbolically resign my Wikipedia editing account in protest at the WMF's cavalier, disrespectful attitudes, but I know it won't do any good. People like Florin make it very difficult to continue to assume good faith on the part of WMF, board and employees. JDanek007Talk 23:15, 11 July 2014 (UTC)
- I don't know if someone already suggested this, but if you are holding control or something like that and click on the image, then it should be able to instantly go to the Wikipedia file page for said image. Dustin (talk) 15:56, 12 June 2014 (UTC)
- For me (using Chrome) this does open the file page, but does so in a new tab. Therefore, in addition to having to having to hold control while clicking, I also need to move up to the top right corner and open the new tab. These extra steps waste my time, and would suck up even more of it if I was using a laptop with a touchpad rather than having a mouse. This again is not desired operation nor a friendly work-around to Media Viewer. - S201676 (talk) 17:51, 12 June 2014 (UTC)
- Again, proving my point. Thanks ;-) --Millbart (talk) 16:04, 13 June 2014 (UTC)
This kind of encapsulates what's going wrong here
[edit]The lead image on South American dreadnought race with Media Viewer is ... troubled.
- Why must I see a blurry image while it loads?
- There's two separate captions given in MV. The first is from the article; the second is from Commons. For the most part, they say the same thing. Why are they placed together with no divider?
- And just why does a frame cut off the caption so that I have to scroll to read the second paragraph?
- Why doesn't clicking on the image bring me to the Commons image page?
- Why does clicking on "public domain" bring me to the Commons image page? Would I be better served with the image linking to Commons, and "public domain" leading to an explanation of what that is?
- Speaking of Commons, why is the link to it so small? Why must I scroll down to see it? (oh sorry, I missed the tiny icon on the bottom right)
- The source blurb is cut off, although to be fair, this image's source is lengthy.
- "Created on 1909": minor grammar point, but why try to make this a sentence? "Created: 1909" Ed [talk] [majestic titan] 21:21, 14 June 2014 (UTC)
- As if constructing a sentence based on a date format is hard! — Scott • talk 20:29, 15 June 2014 (UTC)
Central notice?
[edit]Has this RfC been advertised anywhere? I've been half-following the Media Viewer stuff, but I only happened across this today because someone mentioned to me that it existed. It looks like a pretty small group that's been participating here so far, and since this would be a sitewide change it could probably benefit from as much input as possible. Especially because lately it seems to be just a handful of people here going over the same ground again and again (and getting more and more frustrated with each other's answers...)! Has there been any discussion of putting this on a central notice or something to attract more comment? A fluffernutter is a sandwich! (talk) 20:11, 18 June 2014 (UTC)
- @Fluffernutter:, I did see this, and to be honest, I felt that the amount of response that was coming in was pretty high, and assumed that (before or after your comment) it had been adequately advertised. I don't have a ton of experience with RfCs, so I didn't realize it was possible this level of participation could be considered too minimal to be significant. In hindsight, I should have paid more attention. My apologies. -Pete (talk) 02:01, 11 July 2014 (UTC)
Update 20 June
[edit]I did a run-through at United States and found that all of my objections/reservations/wish list were met. Thanks. TheVirginiaHistorian (talk) 07:52, 20 June 2014 (UTC)
- @TheVirginiaHistorian: Great, I'm glad it all got worked out for you and thanks for the feedback that helped make it possible. Happy editing to you. Keegan (WMF) (talk) 18:15, 21 June 2014 (UTC)
Response to the Media Viewer RfC
[edit]Thanks for sharing your comments about Media Viewer.
The Wikimedia Foundation appreciates feedback about features we develop, and we respectfully acknowledge this group’s proposal to disable Media Viewer on the English Wikipedia.
After carefully reviewing this proposal, we recommend that Media Viewer remain enabled on the English Wikipedia, for a number of reasons:
- We believe that an RfC of this type is not representative of our much wider user base.
- Readers in particular are not represented at all in this kind of discussion, and they are a key user group for this feature.
- Media Viewer was developed with extensive community guidance from a more diverse user base for over a year, through beta testing, online discussions, usability studies and other feedback channels.
Media Viewer provides important benefits to users:
- Larger images: this tool shows images more prominently, with a single click.
- Faster image load: files are shown twice as fast as the previous solution, since you don’t have to go to a separate page.
- Easier browsing: more users click on the next and previous buttons than on thumbnails, increasing overall image views.
- Better use of space: less scrolling is required to find information, due to a more compact layout.
Other factors were considered in reaching this decision:
- Media Viewer is a central part of our strategy and vision to modernize our multimedia software and user experience.
- As recognized by the Wikipedia:Consensus policy, software development is not subject to the same policies which bind English Wikipedia editors.
- Users who do not want this feature can easily turn it off, and only a small percentage of English Wikipedia users have chosen to opt out so far. We encourage users who don't like Media Viewer to disable it.
Overall, we believe that Media Viewer’s benefits far outweigh its downsides. And while the feature still has some limitations, we have collectively identified practical ways to improve it over time.
We deeply appreciate your help in making this tool better and we hope that more users will come to value this feature as a result. Thank you so much for your feedback. Fabrice Florin (WMF) (talk) 17:45, 10 July 2014 (UTC)
- Wow. Just wow. As a reader, I want Media Viewer disabled by default. All the readers that responded to the RfC didn't like it. A majority of the people in your survey on the English Wikipedia, which I assume more readers than editors responded to, said Media Viewer IS NOT USEFUL. Friends, who also don't edit Wikipedia either, have commented on just how terrible and useless Media Viewer is. They were all surprised when I told them there is a way to turn it off because they couldn't find it. Look, I get that you like your ugly baby. It goes without saying that you recommend keeping it enabled--you rolled it out after all. But seriously, stop ignoring feedback just because it doesn't come from the imaginary readers who actually like this. You DO NOT SPEAK FOR ME when you say this is an improvement for readers. You DO NOT SPEAK FOR ME when you say people like me want this. And stop insulting everyone by ignoring the strong majority opinion that Media Viewer should be disabled. I was reading up last night on the Visual Editor debacle that a few people mentioned in this RfC. Taking a page from that incident, if you refuse to implement the consensus in this RfC, perhaps the administrators should implement the consensus in a manner similar to how they did with the Visual Editor. --98.207.91.246 (talk) 18:19, 10 July 2014 (UTC)
Representation for all users
[edit]Look, I don't see why no one sets this RFC to appear at Special:Watchlist or something, and because of this, it is misrepresentative of all editors. Only the ones who dislike Media Viewer or are on the Village pump are likely to see this. Dustin (talk) 17:54, 10 July 2014 (UTC)
- Why is everyone ignoring me? (This wasn't to the IP) Dustin (talk) 18:21, 10 July 2014 (UTC)
- I made the same point some time ago, Dustin V. S.. Then, as now, no one seemed interested in taking time out of their busy fighting schedule to bother with it. A fluffernutter is a sandwich! (talk) 21:21, 10 July 2014 (UTC)
- I have brought up similar points a month ago as well, but I received few little, if any (I cannot remember) responses. Thank you at least for responding. Dustin (talk) 21:25, 10 July 2014 (UTC)
- While I agree it should have been on the watchlists, it's worth pointing out it was put on WP:CENT. That's how I found out about it. How do you put stuff on the watchlists anyway? BethNaught (talk) 21:28, 10 July 2014 (UTC)
- A fine question. I always figured it was through one of the MediaWiki namespace pages, but I don't know. Dustin (talk) 21:30, 10 July 2014 (UTC)
- While I agree it should have been on the watchlists, it's worth pointing out it was put on WP:CENT. That's how I found out about it. How do you put stuff on the watchlists anyway? BethNaught (talk) 21:28, 10 July 2014 (UTC)
- I have brought up similar points a month ago as well, but I received few little, if any (I cannot remember) responses. Thank you at least for responding. Dustin (talk) 21:25, 10 July 2014 (UTC)
- I made the same point some time ago, Dustin V. S.. Then, as now, no one seemed interested in taking time out of their busy fighting schedule to bother with it. A fluffernutter is a sandwich! (talk) 21:21, 10 July 2014 (UTC)
- Hi Dustin and Fluffernutter: Thanks for suggesting that this RfC doesn't represent the views of the wider editor community, which is one of our key concerns. Another major concern is that this type of RfC provides insufficient representation for readers, who are typically uncomfortable participating on talk pages: another communication channel would be needed to make an informed decision about the value of this feature for its primary users. To address this concern, we are open to discussing a wider outreach to all users, if enough community members are interested in collaborating with us to collect feedback that would include a representative percentage of readers and editors. For example, we could show a one-time form to all users when they open Media Viewer, asking them if they want to keep the feature enabled or not, now that they have tried it. This would engage a lot more users than an RfC, with better representation, and without the self-selection bias of an optional survey. The form design, methodology and acceptance criteria could be jointly determined by a task force of community and foundation members, working together in good faith. But this would be a significant effort for us, and we would only consider it if enough stakeholders were willing to be bound by the results -- including some of the most vocal detractors on this RfC. Personally, I think would be a worthwhile experiment, because if this method proves successful, it could be used to arbitrate future disputes for major software features. What do you guys think? Is it worth pursuing this idea, as a possible path towards conflict resolution? Fabrice Florin (WMF) (talk) 01:39, 11 July 2014 (UTC)
- Such a study would have been nice before you rolled Media Viewer out. As it stands, the rather flawed (in your favor) survey says that only 37% of readers found Media Viewer useful. You didn't ask if the readers preferred Media Viewer to the previous file page, only if it was useful. This RfC was also dismal. Combined, this makes your study unnecessary. Media Viewer, in its current state, should not be enabled by default. How about you follow the consensus here and disable Media Viewer by default, do a whole lot more development and testing, and then run the study you propose to convince the community to change its mind? --98.207.91.246 (talk) 02:36, 11 July 2014 (UTC)
- Please do not try to suggest that the response received from the RfC can be simply invalidated because you think it was biased in terms of who gave a response. It has been pointed out numerous times that the wording of the surveys originally used to justify MediaViewer was biased, that the presentation of the tool - looking at a gallery of images - was not representative of reading Wikipedia (and though I am highly critical of MV, I do think it can be a useful tool for looking at galleries or a progression of images), and that the statisitics presented were in various ways manipulated to support the MV project. Maybe your sample size (some 15000 readers and editors from various languages of Wikipedia if I recall correctly) was larger than that of the RfC, but considering how many terrible flaws MV had on rollout day (and still possesses), I cannot say that that extra size did you any good. Point being: If the RfC was non-representative, then so have the methodologies employed by the MediaViewer development team, and for the RfC page to be disregarded so easily comes off a bit hypocritical. That said, I'll give the benefit of the doubt on assuming good faith: tell us how you think we can best gauge the community's thoughts in a representative way, and ensure us that the will of the community (and each wikipedia community separtely given the non-uniformity of opinions of MV across languages) will be respected. - S201676 (talk) 02:41, 11 July 2014 (UTC)
- LOL @ "this group’s [proposal]". Thanks for all the entertainment, Nemo 00:37, 12 July 2014 (UTC)
- Not funny. Dustin (talk) 00:43, 12 July 2014 (UTC)
Decision for local administrators
[edit]As far as I can tell, putting the following code into MediaWiki:Common.js should do the trick:
mw.config.set("wgMediaViewerOnClick", false);
This makes the decision to enable or disable Media Viewer within the purview of local site administrators. There are a variety of ways to make this code conditional, such as only applying it to users who use a particular skin (Vector, Monobook, etc.), users who are in a particular user group (autoconfirmed, sysop, etc.), users with a specified edit count or account registration date, and much more! Hope that helps. --MZMcBride (talk) 19:15, 10 July 2014 (UTC)
- Thanks @MZMcBride: - I've now done that. Testing to make sure it worked right. Please let me know if I got anything wrong. -Pete (talk) 20:04, 10 July 2014 (UTC)
- And I've reverted it. Please see Fabrice's explanation above.--Eloquence* 20:10, 10 July 2014 (UTC)
- Context for anyone following along: @Eloquence: is Erik Möller, Deputy Director and Vice President of Engineering and Product Development for the Wikimedia Foundation, the organization that developed the Media Viewer software. He stated elsewhere that this decision reflects an official Wikimedia Foundation action. -Pete (talk) 21:49, 10 July 2014 (UTC)
- And I've reverted it. Please see Fabrice's explanation above.--Eloquence* 20:10, 10 July 2014 (UTC)
- Fabrice's excuse for ignoring this RfC does not explain or justify your actions. How can you justify your defiance of community consensus? It's not like the RfC is telling the developers to do anything. It's simply disabling code that the community has concluded should be disabled by default. Please undo your reversion. --98.207.91.246 (talk) 20:25, 10 July 2014 (UTC)
- If that is ever restored, please make it so that I can enable it back without fussing with gadgets. That code disables it for everyone, regardless of their preferences. Matma Rex talk 20:19, 10 July 2014 (UTC)
- That code should never be reinserted ever again! That breaks a core feature for readers and prevented those who do want it from re-enabling it. That was a site-breaking change, and if Eloquence didn't beat me to it, I would have reverted it myself.
-- [[User:Edokter]] {{talk}}
20:31, 10 July 2014 (UTC)- Edokter: Hi. Can you please explain what you mean by "breaks a core feature for readers"? I don't follow. Can you also please explain exactly how the addition of this line of code "was a site-breaking change"? You seem to be suggesting that this line of code broke the entire site and that seems pretty extreme. Please elaborate, with or without emphasis. --MZMcBride (talk) 21:20, 10 July 2014 (UTC)
- That code disabled the onClick event for Media Viewer with no way to bypass it. Instead of disabling it "by default", it completely disabled it for everyone. The MediaViewer option in Preferences would have no effect anymore, and the only way to re-enable it would be to negate this code in your private javascript page. MediaViewer being a core feature for all readers means just that; it is a default software feature, one that can easily be disabled by users. This change completely broke it with no way of re-enabling it.
-- [[User:Edokter]] {{talk}}
23:51, 10 July 2014 (UTC)- Mate, I happen to be a reader, and just a reader as you will notice from my lack of a login. My only contributions ever have been the occasional spelling fix when it wasn't too much trouble to do so. I have however donated money in the past, and now I come to regret it. This MediaViewer thing is pretty fucking far from being a "core feature" as you call it. It's a misfeature at best. While I cannot speak for every one of those mythical "readers" that this MediaViewer is supposed to be inflicted upon, I can tell you that it personally very much offends me and totally destroys my reading experience. The incredible levels of arrogance and stupidity (even for a Frenchman) shown in these discussions have also totally shattered any trust or respect I might have had for this Wikipedia project. Just letting you know what one of those readers thinks of this piece of shit. Hope this helps.
- That code disabled the onClick event for Media Viewer with no way to bypass it. Instead of disabling it "by default", it completely disabled it for everyone. The MediaViewer option in Preferences would have no effect anymore, and the only way to re-enable it would be to negate this code in your private javascript page. MediaViewer being a core feature for all readers means just that; it is a default software feature, one that can easily be disabled by users. This change completely broke it with no way of re-enabling it.
- Edokter: Hi. Can you please explain what you mean by "breaks a core feature for readers"? I don't follow. Can you also please explain exactly how the addition of this line of code "was a site-breaking change"? You seem to be suggesting that this line of code broke the entire site and that seems pretty extreme. Please elaborate, with or without emphasis. --MZMcBride (talk) 21:20, 10 July 2014 (UTC)
- Edokter: With respect, this is a fairly warped view of breakage. Images continued to function as they have for over a decade. MediaViewer is a supplementary feature, a pseudo-re-creation of the file description page using JavaScript. Adding only this line of code to MediaWiki:Common.js of course affected all users unconditionally, but that should come as a surprise to nobody. I specifically noted that this code could be made conditional. My point in opening this thread was to point out that while the extension itself cannot be disabled, MediaViewer can trivially be disabled here.
You've inexplicably taken the tone here (with emphasis) that MediaViewer is essential to the reader and editor experience, which is flatly and demonstrably wrong. The worst case scenario you discuss here is effectively the status quo (files link to file description pages). You can completely disable MediaViewer and nothing of value is lost. In fact, the file description pages typically load faster and most certainly contain more information. And unlike MediaViewer, file description pages allow easy editing, which as a wiki we're pretty big on.
If you'd like to vent your outrage, perhaps focus elsewhere? --MZMcBride (talk) 02:55, 16 July 2014 (UTC)
- MZMcBride, we are talking about disabling a core feature of the website that had no basis whatever consensus is cited; if there was any, it was to disable it by default, not completely without even an opt-in. So you may argue technicalities, but I consider that a breaking change that had no basis in any discussion whatsoever.
-- [[User:Edokter]] {{talk}}
09:46, 16 July 2014 (UTC)- I've responded to the dubious "core feature of the website" argument here. Perhaps we simply have fundamentally different ideas of core functionality, but your view seems to fail any test in which reasonableness and perspective are employed. MediaViewer is not core functionality of the MediaWiki software or of Wikimedia wikis by any sane definition and I think the hyperbole here is inappropriate and unwarranted. --MZMcBride (talk) 13:20, 16 July 2014 (UTC)
- +1 not a core feature, certainly, useful though to those who like it. Shutting MediaViewer completely off wouldn't be better behaviour than this site's operator has shown. -- Rillke (talk) 16:26, 16 July 2014 (UTC)
- MZMcBride, we are talking about disabling a core feature of the website that had no basis whatever consensus is cited; if there was any, it was to disable it by default, not completely without even an opt-in. So you may argue technicalities, but I consider that a breaking change that had no basis in any discussion whatsoever.
- Edokter: With respect, this is a fairly warped view of breakage. Images continued to function as they have for over a decade. MediaViewer is a supplementary feature, a pseudo-re-creation of the file description page using JavaScript. Adding only this line of code to MediaWiki:Common.js of course affected all users unconditionally, but that should come as a surprise to nobody. I specifically noted that this code could be made conditional. My point in opening this thread was to point out that while the extension itself cannot be disabled, MediaViewer can trivially be disabled here.
If you want to change the configuration, you should disable it in the proper way and ask for a site configuration change on bugzilla with a description of the change that you want taken and referencing what you consider to be the consensus established on the mediawiki instance in question.. —TheDJ (talk • contribs) 20:51, 10 July 2014 (UTC)
- Indeed. File a site request, link the RfC and poke
greg-g
in IRC. -- Rillke (talk) 21:06, 10 July 2014 (UTC)- The WMF developers who might respond to such a request work for Erik (user @Eloquence: above), so I don't suggest filing a bug at this point. Nathan T 21:31, 10 July 2014 (UTC)
- Although Erik Möller is the Deputy Director; Vice President of Engineering and Product Development, I am optimistic that he will accept almost any decision by a person whose job is to deal with this kind of request on a regular basis; perhaps it needs a slightly longer explanation as usual but letting the software engineers fiddle out a good solution for this problem is better than edit warring and also shows willingness to communicate with each other. -- Rillke (talk) 22:25, 10 July 2014 (UTC)
- I'm happy to file a bug request, but may not be able to do so until tomorrow. I'd encourage anyone motivated to do so. Since the consensus reached was no surprise, and was pretty apparent as the most likely outcome weeks ago, I would hope that there are some more sophisticated thoughts out there about how to implement it without the problems associated with the Javascript implementation I attempted. -Pete (talk) 22:32, 10 July 2014 (UTC)
- I see no easy way doing this in JavaScript; as long as it is enabled by default, you would always disable it for all users, including those who like it :( -- Rillke (talk) 22:57, 10 July 2014 (UTC)
- OK. So, "enabled by default" is an option that (on a technical level) could be easily disabled? If that is the case, it sounds like the bug could be handled by anyone with the needed permissions, who is willing to implement the English Wikipedia's consensus, in spite of the Wikimedia Foundation's recommendation against it. On a technical level, is that accurate, @Rillke:? -Pete (talk) 23:41, 10 July 2014 (UTC)
- It's not more than
$wgDefaultUserOptions['multimediaviewer-enable'] = 0;
wrapped in a conditional in CommonSettings.php; rolling back to the previous state as beta feature ($wmgMediaViewerBeta
) is also an option. So far from the technical side for all interested in Wikimedia-MediaWiki configuration. This all is just hypothetical. -- Rillke (talk) 01:43, 11 July 2014 (UTC)
- It's not more than
- OK. So, "enabled by default" is an option that (on a technical level) could be easily disabled? If that is the case, it sounds like the bug could be handled by anyone with the needed permissions, who is willing to implement the English Wikipedia's consensus, in spite of the Wikimedia Foundation's recommendation against it. On a technical level, is that accurate, @Rillke:? -Pete (talk) 23:41, 10 July 2014 (UTC)
- I see no easy way doing this in JavaScript; as long as it is enabled by default, you would always disable it for all users, including those who like it :( -- Rillke (talk) 22:57, 10 July 2014 (UTC)
- I'm happy to file a bug request, but may not be able to do so until tomorrow. I'd encourage anyone motivated to do so. Since the consensus reached was no surprise, and was pretty apparent as the most likely outcome weeks ago, I would hope that there are some more sophisticated thoughts out there about how to implement it without the problems associated with the Javascript implementation I attempted. -Pete (talk) 22:32, 10 July 2014 (UTC)
- Although Erik Möller is the Deputy Director; Vice President of Engineering and Product Development, I am optimistic that he will accept almost any decision by a person whose job is to deal with this kind of request on a regular basis; perhaps it needs a slightly longer explanation as usual but letting the software engineers fiddle out a good solution for this problem is better than edit warring and also shows willingness to communicate with each other. -- Rillke (talk) 22:25, 10 July 2014 (UTC)
- The WMF developers who might respond to such a request work for Erik (user @Eloquence: above), so I don't suggest filing a bug at this point. Nathan T 21:31, 10 July 2014 (UTC)
Rillke and TheDJ: Filed as bugzilla:67826. Are we taking bets on that bug's resolution? :-) --MZMcBride (talk) 22:39, 10 July 2014 (UTC)
- Thank you. I was never a friend of gambling. -- Rillke (talk) 22:57, 10 July 2014 (UTC)
- RESOLVED as WONTFIX. It looks like we shall need to continue with local admins taking action to enforce this RfC because the Wikimedia foundation won't cooperate. --98.207.91.246 (talk) 00:33, 11 July 2014 (UTC)
- Hi MZMcBride and Pete: We understand that you are frustrated by our concerns about the reliability of this small RfC. But your rush to disable Media Viewer without further discussion seems premature and counter-productive. I would recommend that we look for a more reasonable way to resolve our differences, instead of escalating this conflict. I am sorry that Eloquence had to revert you, but your unilateral disabling of the feature would have removed it for many users who find it valuable, which is unacceptable to us. Other comments on this thread from Matma Rex, Edokter, TheDJ and Rillke seem to support this view. Would you be willing to engage in 'peace talks' to find a mutually acceptable resolution? For example, what do you think of the idea of a more representative outreach, as outlined in the Representation for all users thread above? If we put our minds to it and work in good faith, I think we could find a better way to address our differences. Fabrice Florin (WMF) (talk) — Preceding undated comment added 02:11, 11 July 2014 (UTC)
- Fabrice Florin (WMF), you owe us an immediate explanation: why do you say that this was a unilateral action? This action was taken as a method of implementing broad community consensus. You and Eloquence have given the community the finger; you have told us that we must submit to your will; we have had extensive discussion, and yet you pretend that it was done unilaterally and without sufficient discussion. We have no need of liars; do us a favor and resign your position. Nyttend (talk) 02:47, 11 July 2014 (UTC)
- @Fabrice Florin (WMF):
Before we go any further, your talk of an "edit war" is preposterous. I was reverted once, it was pointed out that my action had accomplished more than I intended, and I did not make the change again. @Eloquence: acknowledged his mistake in introducing the concept of an edit war, but apparently somehow the notion got through to you that there was an edit war, in the five minutes before it was clearly established that there was no edit war. We do not need this kind of drama; please stop using such inflammatory language here.Striking this, because I see Fabrice had reverted this part of his comment just before I posted. Never mind this part, I see it was an honest mistake. -Pete (talk) 03:43, 11 July 2014 (UTC) - Furthermore, you continue to talk about personal emotions: "we understand that you are frustrated." Frustration is not a relevant component of what is going on here. As I clearly stated in the Commons RFC, votes of the format "I Don't Like It" are utterly useless and should be ignored. My position is not one of personal frustration; my position derives from my understanding of what readers and infrequent editors need, and the negative impacts this software will have on our shared goal to help people get involved in our project of freely sharing knowledge. If you read my comments, you would understand that. Please stop misrepresenting my position.
- Finally: I had no idea this RfC would be regarded as "small." To my thinking, it seemed unwise to try to demand the attention of hundreds of users in order to make a point that seemed pretty clear based on the smaller sample. But I understand that this sample isn't big enough for you. If a bigger RfC is needed, that can probably be arranged. But -- in spite of the many concerns raised about the methodology you employed to reach conclusions about how the software is received -- are you truly confident that the result will be different? That is the part that really boggles my mind. -Pete (talk) 02:54, 11 July 2014 (UTC)
- Fabrice, I'm glad to hear you're interested in coming to a solution. If your concern is that "readers" are not well-represented by an RfC, I'm sure we can agree that editors are. As a first step, could we agree to disable Media Viewer by default for logged-in editors, with the preference toggle available for those who do want it? It's clear that the vast majority of editor participants here did not want this enabled by default. Seraphimblade Talk to me 03:25, 11 July 2014 (UTC)
- There are about 125,000 logged-in users who have made an edit on the English Wikipedia in the last 30 days. That would yield a representation ratio of 0.06%. --Tgr (talk) 07:35, 11 July 2014 (UTC)
- @Fabrice Florin (WMF): "unilateral disabling of the feature would have removed it for many users who find it valuable, which is unacceptable to us" How do you know the real numbers who find it valuable when you've added it with force? Opt-in would give you the real numbers, not a forced-in feature (I've had to make some changes to hide it, your disable doesn't work).
- There are about 125,000 logged-in users who have made an edit on the English Wikipedia in the last 30 days. That would yield a representation ratio of 0.06%. --Tgr (talk) 07:35, 11 July 2014 (UTC)
- @Fabrice Florin (WMF):
- WMF isn't helping itself with retention, and it is clear that Erik and yourself are just concerned about yourselves when you state that it was "unacceptable to us" but totally ignoring those who have stated their views in this RfC, which is totally unacceptable to us as a community to be treated like sheep. So far the development of VE, Flow and MV has been very poor and one questions at how much of donors money has been spent on tools/features that have created a massive distrust in the community with the WMF. Bidgee (talk) 07:47, 11 July 2014 (UTC)
- Dear Fabrice: In your statement above you linked to the WMFs Strategy and Vision documents. Could it be the right time now, to question them? Maybe you might take a look at the Image Filter, the Feedback Tool and so on. Now it is the Media Viewer. Do I have to say Flow? --h-stt !? 17:14, 11 July 2014 (UTC)
Trust was broken
[edit]- After this decision and explanation, there is nothing more we can do except express our indignation. After all, the decision was already made when those cosmetic surveys were launched on various wikis, and quickly removed when their output didn’t show the expected approval. Maybe the WMF has the power to do whatever they consider to be best for the project. But, please, do not insult our intelligence by stating that such change has a wide support among the users. After this regrettable episode, it is no longer possible to assume good faith from the team lead by Fabrice Floren. Trust was broken between WMF and the community of editors and I very much doubt that things will ever come back to what they were before. Alvesgaspar (talk) 21:30, 10 July 2014 (UTC)
- It is possible to assume good faith, when you consider that to many people those who supported deactivation in this RfC (myself included) must look like utter tossers. The WMF is quite possibly steamrollering this through for the perceived good of the wiki, as is the case with VE and Flow. And it is true that this RfC could have been more representative. Mentioning it, was trust really ever restored after the VE fiasco? Not being around at the time I'm not aware of what the atmosphere was like. BethNaught (talk) 21:37, 10 July 2014 (UTC)
- Unfortunately, this RfC statement should probably have noted WP:CONEXCEPT, so you might not feel you look that way now. It was only ever going to be advisory. Alanscottwalker (talk) 01:14, 11 July 2014 (UTC)
- Correct me if I'm wrong (I'm really new to this), but I don't think that applies. The change that implemented this RfC earlier today was a change to a page on Wikipedia, something the WMF doesn't have authority over except by WP:OFFICE, which doesn't seem to apply here. In this case, while the WMF isn't beholden to the RfC, it has no right to keep administrators from implementing it. If WP:CONEXCEPT does apply, it looks like the claimed exception is a relatively recent addition added unilaterally by the WMF and not reflective of any consensus or foundational principle. Arguably, then, it's no real policy at all. --98.207.91.246 (talk) 01:26, 11 July 2014 (UTC)
- 98..: I think your reasoning is generally sound, but I do want to correct you on one point -- properly implementing this consensus is apparently beyond the technical ability of a mere administrator. What I attempted seems to have made it impossible for anyone to enable the feature, but the consensus only said we wanted to have the default state be disabled. I do think there are developers with this ability who are not Wikimedia Foundation staff, but I do not think it's accurate to say that Wikipedia administrators are able to implement the decision alone. -Pete (talk) 01:36, 11 July 2014 (UTC)
- Real policy? We normally read what's real policy in Policy, like WP:CON. Policy is meant to describe the way things are. This technical issue seems to have been determined that way. -- Alanscottwalker (talk) 01:45, 11 July 2014 (UTC)
- Indeed, this is beyond the ability of a mere administrator. We need a bureaucrat here, to desysop Eloquence for a blatant WP:INVOLVED violation. Nyttend (talk) 02:54, 11 July 2014 (UTC)
- Not commenting on the rest of this, but since Eloquence is a member of the staff group, no en.wikipedia adminship is needed. --Rschen7754 03:37, 11 July 2014 (UTC)
- But he also has an en.wp sysop flag, dating back from before RFA existed. That we can (and should) take from him, even if only as a symbolic indication of no confidence. MER-C 06:16, 11 July 2014 (UTC)
- Under what policy? Bureaucrats cannot remove the sysop flag except for inactivity, death, on request from ArbCom, and a resignation. --Rschen7754 08:06, 11 July 2014 (UTC)
- Only option is Wikipedia:Requests for de-adminship, even then he has staff tools (WMF would need to remove them) and he didn't really misuse his sysop tools (he only made a threat to use his staff tools [sysops can't revoke sysops]), so I see very little use. Bidgee (talk) 08:21, 11 July 2014 (UTC)
- Under what policy? Bureaucrats cannot remove the sysop flag except for inactivity, death, on request from ArbCom, and a resignation. --Rschen7754 08:06, 11 July 2014 (UTC)
- But he also has an en.wp sysop flag, dating back from before RFA existed. That we can (and should) take from him, even if only as a symbolic indication of no confidence. MER-C 06:16, 11 July 2014 (UTC)
- Not commenting on the rest of this, but since Eloquence is a member of the staff group, no en.wikipedia adminship is needed. --Rschen7754 03:37, 11 July 2014 (UTC)
- Indeed, this is beyond the ability of a mere administrator. We need a bureaucrat here, to desysop Eloquence for a blatant WP:INVOLVED violation. Nyttend (talk) 02:54, 11 July 2014 (UTC)
- Correct me if I'm wrong (I'm really new to this), but I don't think that applies. The change that implemented this RfC earlier today was a change to a page on Wikipedia, something the WMF doesn't have authority over except by WP:OFFICE, which doesn't seem to apply here. In this case, while the WMF isn't beholden to the RfC, it has no right to keep administrators from implementing it. If WP:CONEXCEPT does apply, it looks like the claimed exception is a relatively recent addition added unilaterally by the WMF and not reflective of any consensus or foundational principle. Arguably, then, it's no real policy at all. --98.207.91.246 (talk) 01:26, 11 July 2014 (UTC)
- Unfortunately, this RfC statement should probably have noted WP:CONEXCEPT, so you might not feel you look that way now. It was only ever going to be advisory. Alanscottwalker (talk) 01:14, 11 July 2014 (UTC)
- It is possible to assume good faith, when you consider that to many people those who supported deactivation in this RfC (myself included) must look like utter tossers. The WMF is quite possibly steamrollering this through for the perceived good of the wiki, as is the case with VE and Flow. And it is true that this RfC could have been more representative. Mentioning it, was trust really ever restored after the VE fiasco? Not being around at the time I'm not aware of what the atmosphere was like. BethNaught (talk) 21:37, 10 July 2014 (UTC)
Hello Fabrice. Above, you make some claims I am afraid I have to question partially. I have lived with the MediaViewer the past weeks and was happy to find in your post a way how to turn it off. With all respect and conscience for the necessity to improve Wikipedia, I make here a list of points I did not find answered:
- You say "that Media Viewer’s benefits far outweigh its downsides". Could you elaborate on the downsides from your point of view?
- You say it "provides important benefits to users". What do you mean by "users", active (authors) or passive (readers) ones? What are the benefits for authors like me? I only experienced the MediaViewer as an unpleasant disturbance of my work flow. Often I need to know the exact file name of a picture on Commons, to mark and copy-and-paste the file name for my article wiki text. The MediaViewer does not provide me this exact file name (the ending for the file type is missing). Clicking on the Commons button just takes more time. When I teach editing Wikipedia in my lessons, this turned out negativeley as well because you have to explain to the students extra things. (Btw, for picturing Wikipedia articles I never browse through the pictures of a Commons category picture by picture, I want to see the whole overview and then pick what I want.)
- You say that pictures are seen quicker. I see pictures often less quickly, because it needs time to load larger files. Until full load of the image I only see blurr.
- You say that only a few editors have disabled the MediaViewer. Possibly, because most of them did not know about how to disable it, like me?
- You say that the editors have been approached for feedback. In April I made this comment never to get a reaction. In it, I say that it is not made essentially easier for people unfamiliar with our ways to reuse an image correctly.
- You say that browsing pictures is easier. On my tablet, it isn't. When I pinch in a picture and want to see the lower parts, the part with the information shoves above, so I can never see the lower part of a picture this way.
- You say that you will not follow the outcome of this vote because passive users are not represented here. Under what circonstances would a vote on Wikipedia (any Wikipedia version) change your mind? Do you think that in these cases users' votings are irrelevant at all? On German Wikipedia they want to start a voting, but some feel that the WMF will not follow anyway. I know that we Wikipedians are often people who don't like changes, but your words sound to me like: the opinion of Wikipedians does not count in general, because there is something wrong with these people. I can imagine that this will not improve the relationship between WMF and community.
So, I got the impression, Fabrice, that the main target group of the MediaViewer is the passive reader of Wikipedia articles, not the editor who is creating the content. Is this impression correct? Then, wouldn't it make more sense to enable the MediaViewer for unregistered users, and disable it (opt-in) for registered ones?
These are some to-the-point-questions, but I love to see you again in Wikimania! Ziko (talk) 12:52, 11 July 2014 (UTC)
- No, it wouldn't make more sense to enable MediaViewer for unregistered users. Please do not advocate inflicting that upon us. As I have already stated elsewhere, I browse in permanent private mode, so all preference changes to any website that I may make are lost at the end of my browsing session. Websites should come not broken by default, as most are.
- I am by the way offended by the idea that the RfCs does not represent the wish of us users. I for one, have voted on two of them. Besides, how was it determined by those receiving a salary from the WMF that us users allegedly had a problem with the existing media pages? Where are the bug reports, feature requests, letters to the WMF, public demonstrations, etc., demanding a Flickr look-alike? Btw, I think what Flickr did is horrid too, but I don't go there often and besides, I never donated any money to them. — Preceding unsigned comment added by 109.81.211.240 (talk) 22:01, 10 August 2014 (UTC)
Hi Ziko, I hope Fabrice won't mind if I answer your question about speed instead of him, as I was more closely involved in tracking the numbers. Both normal page loading times and Media Viewer's loading times are randomly sampled and logged. Based on those logs, the median loading time for Media Viewer is around 1650ms, while the median loading time of an English Wikipedia file page is 2600 ms. The difference grows larger for slower image loads; e.g. the 99th percentile loading time for Media Viewer is 24s, while for the file page it is 48s. For a less representative but more direct comparison, we run an automated browser test which measures the two loading times with a test image; you can see the results here.
If you consider that
- Media Viewer preloads images when applicable, so for practical purposes the loading time is often zero for all but the first image on the page, and
- for the less technically apt user who is not so familiar with opening things in new tabs, the only way to look through the images in an article is to click on the first image - wait until the file page loads - click on back button - wait until the article loads - repeat for next image...
then the difference becomes even more dramatic. --Tgr (WMF) (talk) 06:40, 12 July 2014 (UTC)
- MV was very slow to load (no longer use it since I've disabled it), it took almost 40s to see anything (other than the black background), where as the file page takes less then 12s (with a clean empty cache). I'm not against MV but it needs a lot more work before it should be used. Like all the other developments, they have been no where near BETA let alone RC stage and even so, with a site like Wikipedia, it should be a completed feature that is used and not one that is a work in progress. Bidgee (talk) 06:52, 12 July 2014 (UTC)
- @Tgr, did the WMF check this software on old PCs as well as new ones? I was one of the people who tested Visual Editor, one of the reasons why that had to be disabled was that it was written to take advantage of the newest PCs (like our developers probably use) and either didn't run or worse ran ridiculously slowly on old kit, even fairly new kit (like much of the world's population use). If we are seeing the same phenomenon of different experiences of speed is there a possibility that we have the same development mistake - writing software as if our market was big corporates or people who just bought new PCs rather than seeing our audience as all of humanity. ϢereSpielChequers 07:55, 12 July 2014 (UTC)
- WereSpielChequers, you make a very good point. Developers tend to use the latest and newest shiny toys. And not nearly enough time or effort is spent by a lot of developers considering older computers, older browsers and so on. Mostly because those things aren't new and shiny, they are kind of boring, old and slightly depressing. In my job as a developer, it's a constant battle to get developers to care about old browsers or indeed speed of page loading. This is why we now have newspaper websites where loading an article downloads four megabytes of data. It's madness. I'd hope that the WMF try and inculcate a ruthless byte-counting and small-c conservative approach to software development. (Also, big corporates aren't necessarily the best example there: I know people who use corporate desktops and they are some of the worst computers you could imagine. A friend recently told me that to order his new company iPhone 5S, he had to log on to a computer running Windows 98 and Internet Explorer 5.5.) —Tom Morris (talk) 22:21, 12 July 2014 (UTC)
- @Tom Morris: I have no idea how the media viewer runs on older computers (I wouldn't be surprised if it was uncompatible with some older browsers and never loaded), but as a volunteer MediaWiki developer I can assure you that all of the new shiny features being added to Wikipedia (like ULS some time ago, VE, media viewer) are only loaded on demand (so, respectively, when you open the language selector, when you click "Edit beta", when you click a thumbnail) – only a minimal loader script is loaded first (a few kilobytes, tops). There is actually quite a lot of weight put on this, both by the WMF and by volunteers. Such issues with older code are being fixed too (for example, a large part of the video player code was loaded unconditionally on every page, that was fixed a few days ago). Matma Rex talk 22:36, 12 July 2014 (UTC)
- Hi Matma, I was surprised when I was given this explanation during the VE testing. Up to that point I had expected that people who supported our mission would not design software that would negatively effect those on slower kit. Not serving the new feature to certain old browsers or operating systems is not so bad providing you simply find yourself on the "classic" software. But during the testing of VE I was told that my much less than a decade old Ubuntu box would run VE, just very slowly, and that this was a design feature that wouldn’t change. Needless to say I haven't touched V/E since. The relevance to this page is that further up we have a debate going on as to whether Media Viewer is faster or slower, if the devs have followed the same design philosophy as they did for V/E then it would be worth checking if the people experiencing performance problems are the people who don't have the shiniest newest kit. If the answer is that we have repeated one of the mistakes of VE then I would suggest that the devs might take this on board. If we have to write stuff that takes advantage of the capabilities of leading edge consumer electronics, then simply set the software to use the classic version for those PCs that it will run glacially on. That way implementation will be smoother a more positive experience for the community, and over the years the rest of us will see the new features as and when we upgrade our kit. ϢereSpielChequers 07:56, 13 July 2014 (UTC)
- @Tom Morris: I have no idea how the media viewer runs on older computers (I wouldn't be surprised if it was uncompatible with some older browsers and never loaded), but as a volunteer MediaWiki developer I can assure you that all of the new shiny features being added to Wikipedia (like ULS some time ago, VE, media viewer) are only loaded on demand (so, respectively, when you open the language selector, when you click "Edit beta", when you click a thumbnail) – only a minimal loader script is loaded first (a few kilobytes, tops). There is actually quite a lot of weight put on this, both by the WMF and by volunteers. Such issues with older code are being fixed too (for example, a large part of the video player code was loaded unconditionally on every page, that was fixed a few days ago). Matma Rex talk 22:36, 12 July 2014 (UTC)
- WereSpielChequers, you make a very good point. Developers tend to use the latest and newest shiny toys. And not nearly enough time or effort is spent by a lot of developers considering older computers, older browsers and so on. Mostly because those things aren't new and shiny, they are kind of boring, old and slightly depressing. In my job as a developer, it's a constant battle to get developers to care about old browsers or indeed speed of page loading. This is why we now have newspaper websites where loading an article downloads four megabytes of data. It's madness. I'd hope that the WMF try and inculcate a ruthless byte-counting and small-c conservative approach to software development. (Also, big corporates aren't necessarily the best example there: I know people who use corporate desktops and they are some of the worst computers you could imagine. A friend recently told me that to order his new company iPhone 5S, he had to log on to a computer running Windows 98 and Internet Explorer 5.5.) —Tom Morris (talk) 22:21, 12 July 2014 (UTC)
- @Tgr, did the WMF check this software on old PCs as well as new ones? I was one of the people who tested Visual Editor, one of the reasons why that had to be disabled was that it was written to take advantage of the newest PCs (like our developers probably use) and either didn't run or worse ran ridiculously slowly on old kit, even fairly new kit (like much of the world's population use). If we are seeing the same phenomenon of different experiences of speed is there a possibility that we have the same development mistake - writing software as if our market was big corporates or people who just bought new PCs rather than seeing our audience as all of humanity. ϢereSpielChequers 07:55, 12 July 2014 (UTC)
- Hello Ziko, thanks for your thoughtful comments about Media Viewer, which are much appreciated. Here are my responses to some of the key points you made:
- Downsides: We acknowledge that Media Viewer still has some limitations that can be improved over time, like any other software product. For example, here are some of the features we are considering for the next versions:
- Full zoom: See image details in a variety of sizes.
- Show PDF slides: Browse through PDF presentations
- Show videos: Play videos in larger size
- Mobile viewer: Better image display on mobile devices
- Users: We aim to serve all users of Wikipedia, Commons and other MediaWiki sites, including readers and editors. Readers tend to like Media Viewer more than editors now, but we expect that editors will find more value after they become more familiar with features like 'Use this file': this powerful tool lets you easily share, download or embed files on a page. It is prominently featured right below the image, and we invite you to try it: we think it addresses your request to copy and paste the file name for inclusion in articles.
- Image load: I believe Tgr has answered most of your questions. We are working with our operations team to increase load times even further, through back-end pre-rendering of thumbnails.
- Low disable rates: It is indeed possible that some editors may not have noticed the 'Disable' button in Media Viewer. To that end, we are considering providing a more prominent viewing options panel that would make it very easy to switch quickly to your preferred viewing mode. Would that feature address your concerns for this particular issue?
- Community engagement: We have extensively discussed Media Viewer with hundreds of editors over the past year, and have reviewed thousands of feedback comments, leading to substantial improvements to the product. I am sorry that your particular comment did not receive an answer, and will respond to it now: we believe the 'Use this file' tool provides practical instructions on how to re-use a file, including this new 'Attribute the author' feature. I invite you to try it out and let us know how it works for you. Though I understand that you are comfortable with the old editing tools and are not eager to upgrade, we believe this new 'Use this file' tool will make it easier for many more casual editors to add files to articles, which we view as a positive development. And perhaps you too will find this tool useful over time, once you become familiar with it :)
- Tablet issues: We are sorry about this issue, which only impacts a few users that are still using the old desktop mode on their tablets. A few weeks ago, a new upgrade was deployed on all tablets, switching them to mobile mode, and over 99% of users are now using it instead. Media Viewer is not offered for that mode, so none of the issues you report are affecting these users. We are discussing an upgrade of the mobile viewing experience with our colleagues, which may incorporate elements of Media Viewer in the future. On desktop devices, image browsing has been greatly improved, and we are logging over 10 million daily clicks on the next and previous buttons (more than thumbnail clicks), which suggests that millions of user are finding this browsing feature useful.
- Decision-making: We think it is important that all user groups be included in the decision-making process for major features that impact them, not just small groups of self-selected power users on RfCs like this one. We care deeply for the needs of all editors, but believe that discussions like these could be improved to provide better representation for other user groups like readers, donors or casual editors. In coming weeks, the foundation would like to engage community members in good-faith discussions on how to work more effectively together in future product releases, as Keegan points out below. I hope you will participate in these discussions, so we can all make more informed decisions as a movement.
- Downsides: We acknowledge that Media Viewer still has some limitations that can be improved over time, like any other software product. For example, here are some of the features we are considering for the next versions:
Thanks again for your candid observations, which we will take to heart. And I too, look forward to seeing you at Wikimania :) Fabrice Florin (WMF) (talk) 20:51, 12 July 2014 (UTC)
- Thanks Florin for taking the time, and for the links. It seems that the MV has evolved since I had a thorough look the last time.
- My main problem remains that I want to get quickly to the (pure) file title by copy-and-paste, for example when inserting a picture in an info box or more complex situations. In those cases, I don't want the double brackets etc. The MV displays, directly under the picture left, at a prominent place, the file title without file ending (e.g., .jpg) and without _s. I don't think that this is a necessary choice, because the file title is anyway something more technical and not a real "title" of the picture.
- The code proposal for embedding is in English ("File:", "thumb"; not "Datei:", "mini" as we use it in German WP). Could those code words be 'localized'?
- In the caption text of the "Embed" place the file title is used; how about using the description here instead? (For me personally, the caption text is another reason for not using the Embed code, because I usually would have to erase the caption text anyway - it is often not even in the right language.)
- Why, anyway, can I only mark the whole Embed text and not only those pieces I need (=the file title)? Is there a technical reason for this?
- As I now see the different proposed sizes for pictures in Embed: small is 300px, medium 400px, large 500px? As "thumb" is 250px, I had expected it to be e.g. small 150px, medium 300px, large 600px.
- What I was also wondering about: the caveat "You need to attribute the author / Show me how" appears at the place "Download". I would expect it at "Share". People who download a file do it usually for personal use, but sharing means to display it in public? Or, the caveat about the author should be also at "Share".
- Often I am interested in the information about the "source" and in the "description". The information about the "source" seems to be abbreviated by the MV. For this file only "unknown - Werner Lenz: Kleine Geschichte großer Lexika. Bertelsmann Lexikon-Verlag. Gütersloh 1974, S. " and three dots appear. I don't see where I can access the full information without turning off the MV. Description: I'd prefer to see the full text of the description prima vista, not in a box with scroll element.
- I like that the categories are more prominent than on the old fashion page.
- I'm afraid I do not understand what a "desktop mode on a tablet" means. I use a Chrome like browser on my iPad, no special Wikipedia app if it's that what you mean?
- To make a long story short :-): I want to get immediately "File:Leuchtkugeln München 1848-1850 Der Büchermacher.JPG" (no extra code) to mark, copy and paste it. For any reason the MV does not want me to get it. Ziko (talk) 15:49, 13 July 2014 (UTC)
- Has anyone considered what "preloading" of images mean, if you are on a slow line, or slow wireless connection. Or what happens if you pay your internet access by volume? Are the devs aware of how substantial parts of the people connect to the internet? This looks like another bloated bay area project. Guys, even in a country like Germany, there are still almost 1/3 of the users on slow lines (<1 mbit!). We do not have internet via satellite here and no one puts cables in the ground there. Think poorer parts of Europe, think Asia, think Africa. --h-stt !? 16:28, 14 July 2014 (UTC)
FabFlo said ❝It is indeed possible that some editors may not have noticed the 'Disable' button in Media Viewer.❞ "possible"??? It took me a few weeks to figure out how to disable it... If you drop the word "possible" from that statement and replace it with "certain", only then would it be accurate, there are certainly editors out there who still have not figured out how to disable it. As far as tablets go, it still sucks on tablets. — al-Shimoni (talk) 22:55, 14 July 2014 (UTC)
Wikimedia-l discussion
[edit]There is also discussion happening about this RfC and related issues on Wikimedia-l. The thread starts here. As I said in my post a few minutes ago: "Just a note that I am drafting a request to the Board about governance of WMF product launches. Similar problems have happened enough times that I think the Board needs to step in with a more active role. I am also taking a look at the policies around office actions as they relate to product launches, and will likely request that the Board examine that policy as well." --Pine✉ 08:16, 11 July 2014 (UTC)
- If WMF can over rule the community, don't you think that it is a little pointless? I don't think the WMF would care what the community thinks (I think it is already clear with VE, Flow and now MV) and would ignore any such governance regarding product launches. Bidgee (talk) 08:26, 11 July 2014 (UTC)
Where was this RFC advertised?
[edit]Honest question. I haven't seen it until the drama that started yesterday after someone concluded that the RfC is finished. Matma Rex talk 11:12, 11 July 2014 (UTC)
- +1 Thats is also a critic point of the WMF, only the "wrong people" have seen it. I do not even know that there exists something like RfC (on the En-WP). → User: Perhelion 14:25, 11 July 2014 (UTC)
- Well, it had been on the WP:Centralized discussion for over two weeks prior to the conclusion of the RfC, which means the notice was on pretty much every noticeboard and village pump in existence, not to mention countless user pages, wikiprojects, and talk pages. Plus, as an RfC, many editors subscribing to the WP:Feedback request service got notices on their user talk pages. VanIsaacWScont 22:49, 11 July 2014 (UTC)
- Speaking only for myself, CENT is about the worst place to advertise something, because the CENT box is in the "boilerplate header" section of pages where it appears, and it always looks exactly the same at a quick glance. I've seen those headers so many times that I just scroll right past them, and I'm one of the small subset of enwp users who does regularly visit places like ANI and VP. In contrast, watchlist notices or talk page notices pop up somewhere where I and most other editors are looking anyway and where my/our eyes will catch on something having changed. CENT pretty much only attracts people who watch CENT; the goal for a community-impacting RfC like this should be to catch not only those RfC-listing fans, but also a sampling of the people who aren't regularly checking in what new RfCs are running. A fluffernutter is a sandwich! (talk) 13:38, 12 July 2014 (UTC)
- Well, hopefully you learn to pay attention from now on. I know for a fact that you contribute at AN on a fairly regular basis, but if you routinely ignore the notices contained there, I don't know how you have anyone to blame but yourself, and complaining about not being notified of this discussion is disingenuous at best.
- The fact is, in comparison to the notices from the WMF that this thing was coming, this RfC was shouted from the rooftops for all to hear. But if you think we should have a policy that these sorts of changes should always have a full RfC with watchlist notices, I will be the first to sign up, providing WMF abides by it. This RfC would not have happened if they hadn't foisted this monstrosity on the community without proper notice and feedback from the actual user base. But they don't do that when starting up seriously flawed programs like VE and MV for some reason. Maybe it's time for them to start treating this community with some basic courtesy. VanIsaacWScont 20:36, 12 July 2014 (UTC)
- I don't think assigning blame is really on-track for this conversation, and I certainly wasn't trying to be disingenuous. I'm sorry if I've upset you. My point here is that a factor in UX design (and thus, design of any item one wants people to read) needs to be eyeball-grab-ability. That's why we have the orange bar and the red dot for notifications, for example: something changing from one state to another draws attention to itself, especially when its colorful or displaces other parts of the page. A watchlist notice is similar in that it pushes down other content on the page, jarring the reader into noticing that something changed. The CENT box, in contrast, is static, embedded in a section most people read only once or every now and then, and always looks the same. If you're looking for it, or you're re-reading the headers every day, yes, it's absolutely there. But if you're not looking for it or focusing on that page section, you're not likely to take notice of an item appearing or disappearing from the list in the CENT box. So if you want to draw a large number of eyes from many types of editors, even those who don't lurk in or monitor the minutiae of noticeboard-type places, you're likely to have more luck with a watchlist notice than by an RfC silently appearing on CENT. A fluffernutter is a sandwich! (talk) 22:23, 12 July 2014 (UTC)
- Sorry if I'm a bit testy here, but there are several people making disingenuous remarks about this not being advertised, so they don't have to follow community consensus. Nowhere is there a policy that these kinds of discussions need to have a watchlist notice, and none of the opponents took it upon themselves to place a watchlist notice during the conduct of the RfC, but now they want to drag it out as a post hoc excuse to ignore the clear will of this community. Now, like I said, I would be more than happy to support a policy that says these kinds of decisions need to have a watchlist notice advertised RfC, but that needs to be binding on the one party that has consistently acted in bad faith in ignoring the consensus of this community: the WMF. VanIsaacWScont 23:58, 12 July 2014 (UTC)
- Can you please name the particular opponents you speak of? I sure hope you don't mean me. I just asked a honest question, I don't give a damn about this. I didn't push for the RfC to be advertised more widely because I was not aware it is happening, duh. Matma Rex talk 00:53, 13 July 2014 (UTC)
- No, this was not in reference to you, Matma. Specifically, Fabrice Florin, Dustin, and Eloquence have all argued that the overwhelming consensus of this RfC is somehow not binding because it didn't advertise in some way. VanIsaacWScont 01:09, 13 July 2014 (UTC)
- Can you please name the particular opponents you speak of? I sure hope you don't mean me. I just asked a honest question, I don't give a damn about this. I didn't push for the RfC to be advertised more widely because I was not aware it is happening, duh. Matma Rex talk 00:53, 13 July 2014 (UTC)
- Sorry if I'm a bit testy here, but there are several people making disingenuous remarks about this not being advertised, so they don't have to follow community consensus. Nowhere is there a policy that these kinds of discussions need to have a watchlist notice, and none of the opponents took it upon themselves to place a watchlist notice during the conduct of the RfC, but now they want to drag it out as a post hoc excuse to ignore the clear will of this community. Now, like I said, I would be more than happy to support a policy that says these kinds of decisions need to have a watchlist notice advertised RfC, but that needs to be binding on the one party that has consistently acted in bad faith in ignoring the consensus of this community: the WMF. VanIsaacWScont 23:58, 12 July 2014 (UTC)
- I don't think assigning blame is really on-track for this conversation, and I certainly wasn't trying to be disingenuous. I'm sorry if I've upset you. My point here is that a factor in UX design (and thus, design of any item one wants people to read) needs to be eyeball-grab-ability. That's why we have the orange bar and the red dot for notifications, for example: something changing from one state to another draws attention to itself, especially when its colorful or displaces other parts of the page. A watchlist notice is similar in that it pushes down other content on the page, jarring the reader into noticing that something changed. The CENT box, in contrast, is static, embedded in a section most people read only once or every now and then, and always looks the same. If you're looking for it, or you're re-reading the headers every day, yes, it's absolutely there. But if you're not looking for it or focusing on that page section, you're not likely to take notice of an item appearing or disappearing from the list in the CENT box. So if you want to draw a large number of eyes from many types of editors, even those who don't lurk in or monitor the minutiae of noticeboard-type places, you're likely to have more luck with a watchlist notice than by an RfC silently appearing on CENT. A fluffernutter is a sandwich! (talk) 22:23, 12 July 2014 (UTC)
How to go forward from here
[edit]The results of this RfC were fairly predictable, and the events since the closing of the RfC have been depressing but not surprising. It seems to me like we are in the following situation:
- The Foundation—and some in the community—feel that the design, user experience and tooling on Wikipedia is old-fashioned and could do with a spring clean.
- The community feel very much like the Foundation aren't listening to them. Why Wasn't I Consulted? So the Foundation either try and side-step the community, or appeal to the views of the casual user. It all ends up with drama and mess. See: AFTv5, VisualEditor and now Media Viewer. Hell, Pending Changes too.
- Foundation people start thinking the community are a bunch of Luddites who just reflexively hate the Foundation and any attempt at change. The community start believing that the Foundation are out-of-touch and they just want to push their programme of change through without caring about the contributors to the site.
That's a horrible loop: it'll just cause more resentment and distrust.
Of the people I know who work at the Foundation, they all want the best for the projects. So do the community for the most part, otherwise why would we be here? I used the MediaViewer. I don't like it—it gets in the way of my flow of work on Commons and on Wikipedia and Wikinews. I respect the people who are working on it: they aren't waking up every morning thinking "let's find a way of foisting bad software on users".
But the RfC is clear: the community aren't happy with the software in its current form. I hope the Foundation can find the wisdom to respect the community's decision at this point.
I hope the community can try and acknowledge the good faith and hard work done by the developers and not rule out a future for MediaViewer. Instead, we could commit now to holding a new RfC on MediaViewer in six months time. We name a date like, say, February 1, and on that day, we commit to starting an RfC on MediaViewer, and properly advertising it (on WP:CENT, on noticeboards, on watchlist notices). In the meantime, the Foundation can go back, seriously rethink how they engage with the community—threatening desysops is not the way to do it. If the issue is the Foundation believes that the average reader would be more keen on MediaViewer than involved editors, then the Foundation can do user testing and surveys to show that. Hell, some OTRS tickets of "oh, that new picture thing looks nice". The Foundation can also come back and show that they've improved the MediaViewer to resolve some of the issues raised in the RfCs both on en.wp and on Commons.
How can we find a productive path forward? —Tom Morris (talk) 15:49, 11 July 2014 (UTC)
- Sadly the Foundation has been on a war path with the community of experienced editors on Wikimedia projects for some years now, especially here on the English Wikipedia. The Wikimedia Foundation and its officers have long made it clear that they have no interest in the encyclopedic aims of this project or the wiki way. The wiki way of doing things is discuss, revert, improve but the most important thing is that decisions are made by consensus. This is just the last of a long string of instances where the WMF and its officer have made their disdain for the wiki process known to us. However, there really is no recourse. The board of directors, the only body with the powers to direct the Wikimedia Foundation's most senior officials or at least remove them, has long been a self-perpetuating shamble with most of its seats not held by our representative, but by chapters people and appointed randoms with no Wikimedia background. A fork is not feasible, due to the high maintenance costs of the infrastructure, the insane Google ranking of Wikipedia (we'd end in a situation like Wowpedia.org vs Wowwiki.com, where the latter 3 year old pages are regularly trumping in the ranking the fully updated pages of the former, just because of the power of Wikia's SEO and the age of the site) and the fragmentation that the community would incur into (even if we all left, we'd probably end up with multiple forks and split communities -- more sensibly any one issue would only split part of the community). Sadly, there is no real way to address this. The knife is being held by others, and we are powerless when it is used against us. The wiki idea has long been dead with those who got into power, because it is far more easy and expedient to issue diktats than defend their positions and persuade the Wikipedia community. The only, limited, mean of change is to select all elected components of the board on the basis of a campaign promise to return to the wiki way, and see if appointed randoms are just going to ignore the will of the community, not that it would happen anyway. Snowolf How can I help? 16:02, 11 July 2014 (UTC)
- Actually, what we need to do is look at the RFC process, and whether different types of RFC should be used depending on the subject matter, how much participation should be required for disabling key extensions, who should be eligible to close such discussions, how they should be advertised to the community and so on. I shudder to think that 64 people saying that MediaViewer should not be the default media viewer would override the conscious voluntary decisions of 14,681 editors who had already opted in during the Beta testing; I shudder even more that the closure of this RFC refers to completely disabling MediaViewer, an outcome that obviously does not have consensus because it was not the topic of discussion of the RFC. We're the problem here - the way that we allow a very small but noisy minority to override the already-expressed wishes of 14,000 editors is a much bigger issue. Risker (talk) 17:08, 11 July 2014 (UTC)
- If we are to accept that things were going so well that thousands of people were willing to opt-in to MediaViewer and that this overrides the results of the RfC on shear numbers, then a couple of questions need to be asked:
- 1. If opt-in was working so well, why not stick with it, rather than force MV on everyone, and only after initial outcry offer an opt-out procedure?
- 2. If those 14,000+ reviewers of MV were so enamoroed with the product and gave good feedback on it during Beta testing, how did they miss fundamental problems like poor visibility of attribution and copyright/re-use information before MV was rolled out system wide?
- 3. Why ignore the ~5000 English and ~1000 German Wikipedian responses to the Media Viewer survey that presently indicate a 28% approval rating among the communities of those language (which also happen to be the largest and most productive of the Wikipedias)? Certainly just by an order approximation, it is not possible to disregard the voices of 6000 respondents in the same way you feel justified in ignoring 64. - S201676 (talk) 18:23, 11 July 2014 (UTC)
- I don't think that 14,000 editors ever expressed their wishes in any forum. Having a feature enabled does not mean that the user liked it. It means that they had it enabled for any number of reasons:
- Using and liking the feature
- Going on, checking all of the Beta Features to help test
- Turning on Media Viewer and then never looking at any images
- Disliking Media Viewer, but having it on for debugging purposes (this is me)
- Turning on MV by accident and not remembering how to turn it off.
- Anything else you can can imagine
- The RfC was not great either -- asking if they find a feature useful is not the same as asking if it should be enabled, if it is more useful than the previous tool, etc. But it wasn't just insiders voting against "usefulness" -- I voted and hoped that would be the end of my participation. I don't have lots of edits here. I voted negative as a way of saying that I wanted the tool to be fixed before it was deployed, because it was the easy way given to express dissatisfaction.
- So you have two sets of questionable data, and it fell to the the admins to interpret it. They did so, and got a pseudo-quasi-OFFICE action in response. The software team is clearly biased on this matter as they invented new categories of administrative actions to justify their use of power.
- My suggestion is to take a month or two, turn off Media Viewer by default on enwikipedia, fix some bugs, come back to us with a better, more finished project, and ask for a better RfC. Two months is not a big deal, and everyone will be in a much better temper to deal with this issue. In the meantime, the original decision should stand, so that the community trust in the WMF isn't completely eroded.
- For the record, I don't think that Media Viewer is bad software, I just think it's incomplete, and needs more time to mature. I've filed 2 bugs so far (1 indirectly), and encourage others to do the same. BrentLaabs (talk) 03:14, 12 July 2014 (UTC)
- @Tom Morris: I don't see a productive path right now. Not before the Arbitration Committe decides ([1]), not before WMF acknowledges that the consensus of community will be respected and not with the model suggested above. If a new RfC is to be started (or continued) that should be done as soon as possible but after the MV was presented an opt-in. It is the WMF who has to demonstrate that it has the full support of the community, not the contrary. -- Alvesgaspar (talk) 17:25, 11 July 2014 (UTC)
- Alvesgaspar, the closure of the RFC did not reflect consensus: read it again. The close says that there's consensus to disable MediaViewer. There is no such consensus, because that is not what was discussed in the RFC. The RFC was about whether or not MediaViewer should be the default. This is what I mean when I say that WE are the problem: we can't even get a major RFC closed with a proper interpretation of the discussion. You can't complain about the WMF ignoring consensus when *we* are ignoring it too. Risker (talk) 17:34, 11 July 2014 (UTC)
- Risker -- I'm well aware what the Rfc was about, please don't put words on my mouth. When I say that the change should be reversed I mean that the Media Viewer should be disabled by default for everybody, just like the RfC indicates. The choice to use it or not will remain of course possible. Alvesgaspar (talk) 17:43, 11 July 2014 (UTC)
- Please be clearer then :) As it was mentioned above, the original implementation of this RFC disabled the media viewer for everyone, with no option to reenable it. That one should definitely not be reversed. Matma Rex talk 18:00, 11 July 2014 (UTC)
- Risker -- I'm well aware what the Rfc was about, please don't put words on my mouth. When I say that the change should be reversed I mean that the Media Viewer should be disabled by default for everybody, just like the RfC indicates. The choice to use it or not will remain of course possible. Alvesgaspar (talk) 17:43, 11 July 2014 (UTC)
- You don't own Wikipedia --In actu (Guerillero) | My Talk 18:50, 11 July 2014 (UTC)
- Alvesgaspar, the closure of the RFC did not reflect consensus: read it again. The close says that there's consensus to disable MediaViewer. There is no such consensus, because that is not what was discussed in the RFC. The RFC was about whether or not MediaViewer should be the default. This is what I mean when I say that WE are the problem: we can't even get a major RFC closed with a proper interpretation of the discussion. You can't complain about the WMF ignoring consensus when *we* are ignoring it too. Risker (talk) 17:34, 11 July 2014 (UTC)
- @Tom Morris: unfortunately it was too late for the Media Viewer launch, but Rachel was hired in late May by the Wikimedia Foundation as the first Director of Community Engagement, Product, to develop a system with us (community liaisons) hand-in-hand with us (community members) for future product releases. She should be posting to something to this effect shortly on Wikimedia-l as well has here. Thank you very much for the insights, as always :) Keegan (WMF) (talk) 21:05, 11 July 2014 (UTC)
- «The design, user experience and tooling on Wikipedia is old-fashioned and could do with a spring clean» — thank you for including the word tooling on the list. This is something I believe the WMF have missed, slightly, with focus on extensions developed off-wiki. --Gryllida (talk) 09:30, 13 July 2014 (UTC)
- In my opinion as a volunteer developer, as much as it pains me to say, I have sorta given up trying to solve this. nothing has worked, nothing will work, and I fear we are heading towards a fork of both community and software. —TheDJ (talk • contribs) 16:33, 13 July 2014 (UTC)
- Hello all - as Keegan mentioned, I recently started with the Foundation to focus on product development collaboration within the various communities. We are in early stages of considering more effective ways of collaborating with the various communities in product development, and we will be developing this strategy with community involvement. It's a little soon for me to be reaching out for feedback on a wide scale, though I have met several users already and I do welcome contact (please just understand it might take a little time to get back to you!). Please understand that we're a small team; we need a little time to set this up properly with you, but we are dedicated to increasing collaboration across the foundation-supported communities. -- Rdicerb (WMF) (talk) 15:52, 14 July 2014 (UTC)~
- Hi @Rdicerb (WMF):, as I said above trust was broken and I do not intend to cooperate with WMF on the improvement of MV until the feature is disabled as a default for all users. I'm afraid that is the only way for me to believe that you are acting in goof faith. Until now, all of your actions suggested the opposite. -- Alvesgaspar (talk) 16:40, 14 July 2014 (UTC)
- goof faith [sic] indeed. Legoktm (talk) 18:34, 14 July 2014 (UTC)
- Hi @Rdicerb (WMF):, as I said above trust was broken and I do not intend to cooperate with WMF on the improvement of MV until the feature is disabled as a default for all users. I'm afraid that is the only way for me to believe that you are acting in goof faith. Until now, all of your actions suggested the opposite. -- Alvesgaspar (talk) 16:40, 14 July 2014 (UTC)
- @Legoktm: -- Please enlighten me if your apparently humorus comment [sic] is directed towards my poor use of English or towards my opinion on the issue. And also if your opinion is to be meant as personal or (WMF) official. Under the present circumstances neither seems correct to me. Alvesgaspar (talk) 18:58, 14 July 2014 (UTC)
- Your opinion on the issue. You're hypocritically requiring "good faith" from the WMF, yet refusing to do it yourself. You claim "trust was broken", but I don't really see how. Did the WMF ever promise they would honor the outcome of the RfC? My comment was obviously my own, and I'm not sure why you'd think it would be a WMF one. Legoktm (talk) 20:32, 14 July 2014 (UTC)
- I don’t believe that this kind of ad hominem attacks help. No, I am not a hypocrite when I say that WMF has not acted in good faith in which the relationship with English Wikipedia’s editors is concerned. If WMF did not intend to honor the outcome of the RfC, why wasn’t that stated clearly before it started and why did their members participate? If WMF does not intend to disable MV despite the overwhelming negative reactions, why these pathetic attempts to get the community involved? If WMF did not intend to follow the indications of their own statistics, why were they made in the first place? In terms of hypocrisy, WMF has much more to explain than me. -- Alvesgaspar (talk) 21:43, 14 July 2014 (UTC)
- Alvesgaspar, while I admit I do not know all of the specifics of your concerns yet, I recognize that you feel wary. I know there is history there. I am not going to make a call on MediaViewer, as it is not my product to decide upon. My post was actually in the context of the title of this section "How to go forward from here?" as posed by Tom Morris. Quite honestly, I do not have an easy answer for that. It's a huge task and we need time to work through it. There is a lot to do as far as better communication and involving all areas of the community in the product process. We all have the common goal of building the world's repository of knowledge - I hope we can start with that. Beyond that, my team and the Foundation will be welcoming respectful discourse on how to create a Community Engagement strategy that truly includes the community (I do not just mean the users participating or even reading this comment). This is a partnership - right now it's clearly an uneasy one, but please trust that we want to increase the effectiveness of collaboration. In coming weeks we will be considering some strategic initiatives, and I welcome communications from community members who wish to collaborate towards our mission. We'd like to open up a wider conversation to a greater cross-section of the community post-Wikimania. Rdicerb (WMF) (talk) 23:53, 14 July 2014 (UTC)
- @Rdicerb (WMF): Your grandiloquent words about a conversation with a greater cross-section of the community seems pretty shallow to me especially when there is such an urgent and serious problem to solve with the local editors, those who keep this project rolling. A problem about which you seem to know nothing. To have a better perspective of what is really going on maybe you should first take a look into the Arbcom discussion going on now. I say again: there is a narrow path for future communication with this community and such path involves that WMF, as a sign of good faith, disables MV as a default viewer. -- Alvesgaspar (talk) 03:27, 15 July 2014 (UTC)
- We've waited five weeks for the WMF to listen to us and remove this misguided and deeply flawed feature. Telling us to wait longer while you "consider some strategic initiatives" is insulting. The WMF must heed this RfC much sooner than later if any trust is to rebuilt here. --98.207.91.246 (talk) 05:25, 15 July 2014 (UTC)
Focus on the features
[edit]One suggestion for "where do we go from here" is to focus on the features. In my experience that is often the best approach when dealing with programmers. Nobody like to hear people saying "I don't like your work", much better to say "can we change this". For example, my use of Wikimedia Commons is quite specialised, I do a fair bit more categorisation than the average commonista. So no surprise that I don't want Media Viewer, whether or not it works for the great majority, it isn't written with me in mind. In commons user preferences there are two options:
- [ ] Place categories above all other content.
- [ ] Place categories above content, but below image on file description pages.
Currently Media Viewer is incompatible with these choices, but there is no warning of that in the user preferences. I would suggest that one should either assume that any of us who have opted in to one of those options should by default be opted out of Media Viewer, or better still change Media Viewer so that if people have opted in to making the categories more prominent it respects that user choice. A couple of days ago on my third attempt I found and clicked the option to opt out of media Viewer. If the option to opt out of Media Viewer had been repeated on the gadgets page next to the options that it overrides I would have found it and opted out a month ago. If we can focus on the features that cause problems with new software then the developers have the option of either fixing that and making everyone happy, or accepting that this doesn't work for some people or projects and only trying to promote it to others, or at least acknowledging that those who oppose a new feature may be people for whom it is negative. Focussing on the features can work in both directions. Everyone here uses computers in some form, by definition none off us are luddites, describing those who are annoyed by a new feature as Luddite is not only offensive, it can be taken as a sign that developers either don't know or don't care about the flaws in their new software. So I would suggest that the devs and other WMF staffers in these sort of implementation arguments try to focus on the bugs and unpopular design features of their new software and remember that if you call your opponents Luddite, what some of them hear is "I don't know or care what's wrong about this software, I've got a deadline to meet or a sponsor to satisfy". ϢereSpielChequers 10:37, 13 July 2014 (UTC)
- How is it not compatible ? The categories show up just fine for me on the file description pages. If you want the gadgets to support MMV, you should ask the volunteer author/maintainer of the gadgets. Gadgets are volunteer driven utilities that are not supported by the developers and WMF. That's why they are easy to change and equally easy to break. —TheDJ (talk • contribs) 15:35, 13 July 2014 (UTC)
- There is indeed a work around of using the file description pages, the annoyance and inconvenience comes when I revert to my normal way of working and click on the picture. Thanks for the explanation as to why gadgets are easily broken, in terms of the user experience I could understand if there was some process by which very rarely used gadgets weren't tested for compatibility. But systematically ignoring collateral damage to gadgets is not a strategy conducive to good relations between those whose developments do such damage and those who suffer it. ϢereSpielChequers 06:14, 14 July 2014 (UTC)
- That is exactly why some developers think that we should abolish Gadgets and users scripts all together. Because they are not supported, yet people still expect them to be supported. The confusion and blame game that is the result of that is not healthy. —TheDJ (talk • contribs) 08:40, 14 July 2014 (UTC)
- There is indeed a work around of using the file description pages, the annoyance and inconvenience comes when I revert to my normal way of working and click on the picture. Thanks for the explanation as to why gadgets are easily broken, in terms of the user experience I could understand if there was some process by which very rarely used gadgets weren't tested for compatibility. But systematically ignoring collateral damage to gadgets is not a strategy conducive to good relations between those whose developments do such damage and those who suffer it. ϢereSpielChequers 06:14, 14 July 2014 (UTC)
Proposal to reach consistency/agreement first, before actioning this RfC
[edit]I feel that:
- We've only reached a small fraction of the userbase;
- An RfC is a messy way of doing this: we get a "yes/no" feedback on the entire product;
- The Multimedia team are collecting statistics on multiple values (usability, overall experience, etc);
- While I have no close familiarity with the details of the exact way they're doing things, theirs show positive responses from users (and iirc they targeted the ones who aren't logged in, as well.)
- Before actioning this RfC, we should make sure that the statistics and the RfC outcome are made consistent;
and this consistency should be made by one or both of these means:
- adjusting the statistics process, or
- re-opening this RfC and reaching more people with it by a banner at pages top.
--Gryllida (talk) 09:35, 13 July 2014 (UTC)
- I see no problem running this for longer and advertizing it more widely. However, the statements made so far by WMF employees indicates that were participation to double, or (highly unlikely) quadruple, it would still be dismissed. In this context running it again might only cause more frustration. Perhaps WMF employees could suggest what procedure should be followed for a community vote, such that they would be prepared to take action on the outcome, and what they would accept as a meaningful number of minimum votes? --Fæ (talk) 10:10, 13 July 2014 (UTC)
- I see issues with the approach of re-opening the RFC and advertising wider.
- In a nutshell: Let's do it for all and with less noise.
- The whole approach of being against media viewer is sick. You need to be catalytic; think what should be done next, with least drama and community bashing.
- This RFC is a pretty static thing. The stats WMF Multimedia are collecting are :
- for all — literally a thing logged-out users see and respond to even (you can compare to that only if you show a global notice at page top for everyone);
- more fine-faceted, asking what people would like to remove or improve;
- enables feedback from many languages and projects, that helps to improve the media viewer in long term — not just one (but this one included, so that you can improve the situation here).
- Looking into these stats and improving their surveys should be a nice thing. On the plus side, the team itself would observe the results and attempt to address them. --Gryllida (talk) 11:27, 13 July 2014 (UTC)
- I see issues with the approach of re-opening the RFC and advertising wider.
Scale of the requests for comment process
[edit]Currently the community relies on "requests for comment" which are messy and reach inadequately small portion of audience due to people being lazy to edit the sidebar to promote these requests.
Dear community, you should know better. The sidebar is under your control. You're free to enact a policy by which you put any requests for comment and requests for permission into the sidebar, to have the user-base respond.
I'm disappointed in your desire to make that much noise instead of properly using the tools that you have.
--Gryllida (talk) 23:23, 13 July 2014 (UTC)
- So, your opinion appears to be that "the community" is lazy, should know better and wants to make noise rather than using tools. Thanks for sharing a piece of your mind, however please remember that the time of the unpaid volunteers that compose the community is Wikimedia's most valuable resource, and randomly telling off anyone that reads this, is unlikely to be productive. --Fæ (talk) 03:39, 15 July 2014 (UTC)
- @Gryllida:: I'd like to echo Fæ's sentiments. I feel your comment was not helpful. RomanSpa (talk) 21:11, 17 July 2014 (UTC)
- +1 (Fæ). These problems are complex, and require a careful balance among competing interests. Yes, there may be better ways of going about things, but "you should know better" is not a helpful comment. -Pete (talk) 22:14, 17 July 2014 (UTC)
- Mostly, this comment belongs somewhere else, not on the talk page of an RFC that, even when it was active, only attracted the attention of 111 people. A broad discussion on the manner in which RFCs are conducted (which is what this seems to be pointing to) should be, for example on Wikipedia talk:Requests for comment, or possibly WP:VPP. Risker (talk) 01:35, 18 July 2014 (UTC)
- Yes, Fæ and all, sorry. However, in long-term, I would ideally like to see more flexibility in using the tools and control that we have over this software on-wiki. (Risker's suggestion to talk about changing the process is relevant, but may be out of my depth given that people had ignored about a half of my suggestions at village pumps before.) --Gryllida (talk) 17:00, 19 July 2014 (UTC)
Structure of the requests for comment process
[edit]- See also: Survey tool for features
Fabrice Florin, would you please consider working on, or finding and enabling, a surveys extension for the wiki software, and using it for the Multimedia team so that it's also available for everyone else to use? The off-wiki survey sounds in-transparent at best, and people are too shy (understandingly) to use the survey-monkey tool for their RFCs.
It would
- make RFCs easier for people, I believe. They would be able to gather statistics, not the mess we have here where each responding user has to read through large chunks of text and edit sections manually. Proper surveys leave a matrix of responses, correlations, and statistics.
- either be a private vote, or entirely transparent with options to view responses of everyone else, like this RFC is. Preferably the latter?
- probably leave in its own namespace, such as Request for comment:*.
It would be useful for easing requests for comments, and requests for permissions. —Gryllida (talk) 23:23, 13 July 2014 (UTC)
Update: Quick access to enable/disable, opt-in/opt-out metrics
[edit]![](http://upload.wikimedia.org/wikipedia/commons/thumb/4/40/Media_Viewer_-_Viewing_Options_Panel_-_Mockup_1.png/220px-Media_Viewer_-_Viewing_Options_Panel_-_Mockup_1.png)
![](http://upload.wikimedia.org/wikipedia/commons/thumb/7/78/Media_Viewer_-_Viewing_Options_Panel_-_Mockup_3.png/220px-Media_Viewer_-_Viewing_Options_Panel_-_Mockup_3.png)
Hi everyone. We appreciate the ongoing conversation about Media Viewer, and we’ve been discussing practical ways to address concerns and improve metrics used to establish the default configuration for this tool.
With that in mind, we would like to propose a new feature which we think can address many of the issues you reported here: the Viewing Options Panel would make it very easy to disable (or enable) Media Viewer. This feature would prominently display a ‘cog’ icon at the top right corner of the screen, as shown in the thumbnails to the right — and in this rough prototype. Clicking on that icon would display a settings panel with two viewing options:
- 'View quickly’ — enable Media Viewer (or keep it enabled)
- 'View all the details’ — disable Media Viewer and show the file page instead
The Viewing Options Panel would let users quickly select the mode that works best for them, switching back and forth to compare them. (Note that Media Viewer already includes a “Disable/Enable’ link, but it is located below the fold and can be hard to find; this new feature would bring it above the fold, where everyone can see it.)
This would make it much easier for users to decide for themselves if they want Media Viewer enabled or not. It would also enable us to collect better user data on whether or not this tool is useful -- which can inform our discussions about default state for different user groups. To that end, we plan to track the number of enable and disable events by user group on this existing opt-in/out dashboard (this dashboard now tracks clicks on the “Disable/Enable’ links, which would be replaced by the Viewing Options Panel).
What do you think of this new feature? Does it seem worth developing at this time? Any suggestions for improvement? We would greatly appreciate your feedback on this Media Viewer discussion page.
We are also considering a controlled experiment on the English Wikipedia. We invite you to comment on that separate proposal on its research page.
Thanks again for taking the time to share your thoughts and feedback on Media Viewer. Your comments are really helpful to us, and we look forward to working with you in coming days to find a resolution to your concerns. To be continued, Fabrice Florin (WMF) (talk) 16:42, 18 July 2014 (UTC)
- This is a practical way to address the concerns expressed in this RfC if and only if you disable Media Viewer by default for all users and this is your method of opting in. I wish you'd quit ignoring that the main concern expressed in the RfC is that Media Viewer SHOULD NOT BE ENABLED BY DEFAULT. Also, don't call it "View quickly." If I disable Media Viewer, I can be looking at a full sized image faster (with two clicks and a page load!) than it takes for the downscaled image to show up in Media Viewer. Also, ditch the mystery meat navigation. Use text labels so I don't have to guess at what the various buttons do. Finally, your opt out doesn't quite work. It'll stick for awhile and then I'll click an image and BOOM MEDIA VIEWER, something I never want to see again. --98.207.91.246 (talk) 17:14, 18 July 2014 (UTC)
- With this Arbcom process going on, it is not certainly the right time to address your proposal. -- Alvesgaspar (talk) 17:24, 18 July 2014 (UTC)
- I see little reason to pay attention to small tweaks, while the software remains enabled. For the last decade, we had a system that supported the creation of the most widely read publication in human history. There's broad agreement that it can be improved upon, but also broad agreement that the Media Viewer software is not an adequate improvement.
- If and when the Media Viewer is disabled by default, I will be willing to discuss details. Until then, I see messages like this as gasoline on the fire. I'm not interested in participating. -Pete (talk) 17:27, 18 July 2014 (UTC)
- I think it is surely a step into a better direction. However, I still very much care about the default (as others above). There are still many people around here with low bandwidth connections and, so far, Wikipedia was friendly also to these users. By default you got small pictures and by going to the preferences you could move to larger pictures. We should keep this conservative approach. A fully blown picture is simply a road block for many users and they get only the chance to disable it once they have been able to load it. My other concern that there is still no edit button within the MediaViewer. The Wikipedia is the project anyone can edit — this functionality should be accessible also for non-regulars who are not familiar with the interface. --AFBorchert (talk) 21:51, 18 July 2014 (UTC)
- When VisualEditor was introduced, people had a right to be upset as it disrupted their workfrlow and was too buggy to be usefull. It took a while before the WMF realized the same and disabled it. In this case however, there is no such disruption. MediaViewer is about presenting images and other media in a more up-to-date manner. And it can be disabled very easily. So, I predict that disabling the MediaViewer by default is... how to put it mildly... not gonna happen.
-- [[User:Edokter]] {{talk}}
22:50, 18 July 2014 (UTC) - AFBorchert, more prominent file edit interface is on the must-have todo: [2]. --Gryllida (talk) 17:07, 19 July 2014 (UTC)
- Fabrice Florin (WMF), please consider removing the "show this interface to only logged in users" note from that card. I fundamentally feel that anyone should know how to edit, including logged-out contributors (and iirc those are not allowed to upload files but are allowed to edit descriptions; whether they're allowed to upload new versions of the same file, I dunno). --Gryllida (talk) 17:12, 19 July 2014 (UTC)
- When VisualEditor was introduced, people had a right to be upset as it disrupted their workfrlow and was too buggy to be usefull. It took a while before the WMF realized the same and disabled it. In this case however, there is no such disruption. MediaViewer is about presenting images and other media in a more up-to-date manner. And it can be disabled very easily. So, I predict that disabling the MediaViewer by default is... how to put it mildly... not gonna happen.
- The right thing to do is not to endlessly tweak MV, but to roll back to a known good option (i.e. the old way), then tweak MV to your heart's content, fix bugs, get navigation down, use WMF and opted-in users as guinea pigs, and get MV to the point where it's feature complete, rock solid, and fast. Then and only then hold an RFC to see if MV should be adopted, and if so, how and when it should be deployed.
- I think MV is bad--but the constant stream of minor tweaks and twiddles you guys are applying to try to fix it makes the user experience absolutely terrible. Rolling out fundamentally flawed beta software to the world without consultation is wrong. Spending weeks trying to fix it in place is even worse. — Preceding unsigned comment added by 69.140.0.55 (talk) 18:51, 19 July 2014 (UTC)
New MediaViewer RfC
[edit]Wikipedia:Village pump (proposals)#Media Viewer RfC Question 1 & 2 (shortcut: WP:VPR#rfc 5D8BCA2) --Trofobi (talk) 07:38, 12 October 2014 (UTC)
- For anybody who finds a dead end link above, that's because it was moved here: Wikipedia:Village pump (proposals)/Media viewer 2014 -Pete (talk) 01:37, 18 May 2015 (UTC)