Jump to content

User talk:Jimbo Wales/NPOV report on problems with MV

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Problems with Media Viewer (MV)

[edit]

@Jimbo Wales: Hi Jimbo, here is what I consider to be a neutral interpretation of the recent conflict:

  1. MV seems to have been implemented prematurely as a default viewer for everybody (numerous bugs, some violent protests);
  2. MV is not appropriate for editing work; many editors have considered the tool as a "nuisance" ;
  3. A vast majority of the editor’s community does not consider MV useful. That is demonstrated by the poll made by WMF in June;
  4. A vast majority of the editor’s community considers that MV should not be implemented by default. That was demonstrated in a recent RfC, whose results are in line with the WMF poll;
  5. No effective fruitful dialogue has occurred between the editor's community and the MV team immediatly before and after the RfC. The two main arguments from each side seem to be: no RfC can produce a significant consensus of the user's community, which goes well beyond the universe of regular editors (WMF); according to Wikipedia's culture the outcome of a RfC is to be respected. Thus no dialogue seems possible while MV is kept as the default viewer for everybody (editors)

I hope this start is useful. Please note that English is not my mother language and that my knowledge of Wikimedia's world is quite narrow. Hopefully other people will come here and say what I intended to say much better than me. :-) Alvesgaspar (talk) 19:16, 25 July 2014 (UTC)[reply]

I've had extensive, effective dialogue with the editing community over the past nine months while Media Viewer has been under development through on-wiki interaction across dozens of projects, by email, face-to-face discussions, video roundtables, IRC office hours and discussions, etc., in the shaping, development, and carefully documented release process for Media Viewer. Fabrice and the developers of Media Viewer have participated extensively in these conversations as well. I've even had productive conversations with you, Alvesgaspar, and conversations with the editing communities about Media Viewer continue even during this period. I'm not 100% certain this is a neutral point of view, though I do know you honestly intend it that way. I think it's fair to say I don't have a neutral point on the topic either, which is why this will be my only participation in the discussion :) Keegan (WMF) (talk) 22:16, 25 July 2014 (UTC)[reply]
You (multimedia team and WMF engineering) ignored the vast majority of editors and the majority of everyone else who have asked that Media Viewer be disabled by default. The only dialog the WMF is willing to have on Media Viewer is about how it can be improved, discounting any feedback that it isn't ready for release (and thus should be rolled back) or feedback that perhaps the concept itself is flawed. So from my perspective, there has been no productive dialog here. You've (the multimedia team and WMF engineering) decided, a priori, that Media Viewer needs to be rolled out and you are unwilling to reconsider this decision, overwhelming negative feedback be damned. Instead of imposing your "multimedia vision" from above, perhaps you could actually have some dialog about what users of Wikipedia want instead of pretending that the silent majority likes your ugly baby? --98.207.91.246 (talk) 02:52, 26 July 2014 (UTC)[reply]
@Keegan (WMF): Regarding your "extensive" dialogue: You've made just 225 edits on this wiki under this WMF account. Your first edit involving MV was as recently as April 28, and was to "clean up" (er, delete) content that I just looked at and found to be interesting. Prior to that date, your focus was on Visual Editor. On that page just prior to your cleanup, I found a link to mw:Multimedia/Media Viewer/Usability testing, and I viewed this recording of "User 1". I chuckled a few times at how she did some tasks. Did you, by chance, have this same user do the same tasks using our existing legacy file namespace system? I would be curious to see whether she found that easier than MV to navigate and answer the questions. And I would really like to know whether she preferred MV or our current system. Your second edit about MV (May 16) was to announce on Village Pump (technical) that it was launching on May 22. By then it was a fait accompli. Wbm1058 (talk) 03:55, 26 July 2014 (UTC)[reply]

I think the problems with MV can be simply summarized as the interface being the antithesis of what information producers need. It's pretty much the Windows 8 UI fiasco transplanted to Wikipedia. Relevant quote from [1]: "The idea of hiding the controls to give priority to content may make sense on mobile, where screen space is so limited, but it doesn’t make that much sense on a large screen, especially if users have to work harder to access hidden features." The aspects of an image which are normally useful to information producers on Wikipedia, like:

  • categories
  • which articles this image appears in
  • copyright status
  • other versions of this image or derived images

are hidden behind a mysterious interface in MV that has no self-explanatory way of giving you that info. The MV may useful to some information consumers like to a dude who wants to fap to male porn on Commons or, stated in less lurid terms, some guy who wants to use wikipedia/commons for some kind of virtual tour. But that's not what information producers do. So MV—as designed—will always be a hindrance to the editing community. Did I mention that the disable link for MV is shown to IP/not-logged-in users even though it doesn't work for them, and no explanation is provided in the MV UI that/why the disable link/button doesn't work unless you are logged in? That violates a core UI design principle; the MV architects should be able to tell you what it is called, if they took a HCI class in school. 188.27.81.64 (talk) 15:46, 28 July 2014 (UTC)[reply]

As for WMF's focus on improving MV by fixing bugs here and there, this quote from Douglas Adams (as adapted in a HCI book) is relevant: "It is very easy to be blinded to the essential uselessness of [products] by the sense of achievement you get from getting them to work at all. In other words... their fundamental design flaws are completely hidden by their superficial design flaws." 188.27.81.64 (talk) 18:02, 28 July 2014 (UTC)[reply]

Summary so far of specific problems with the MV software —

  • Useful information about an image is not easily accessible. For example: categories; which articles this image appears in; copyright status; other versions of this image or derived images. (User:188.27.81.64)

--Bob K31416 (talk) 13:11, 1 August 2014 (UTC)[reply]

  • Principle of least astonishment broken I was looking at a graph on the WWII article, and suddenly found myself looking at full screen images of Chinese civilians being buried alive by soldiers of the Imperial Japanese Army, and female SS camp guards removing bodies from lorries at Bergen-Belsen. I brought this up at MediaWiki (I think it was) and the response was effectively that since the community had rejected image filter it was tough luck. I think this is a serious issue, and the implementation of the MV means that editorial decisions about images presented as thumbnails are moot. Some similar software has thumbnails instead of right/left buttons. All the best: Rich Farmbrough00:02, 2 August 2014 (UTC).
Rich Farmbrough, I located the graph and pictures you were referring to.[2] Could you clarify here the problem with MV that you see in that article? --Bob K31416 (talk) 01:53, 2 August 2014 (UTC)[reply]
One of the things that turns me off about it is that it is all icon based. Wikipedia is a word based interface and I found myself annoyed at having to hover over each thing to see what it did. You could fit more buttons if it was text.
Like the rest of Wikipedia there is no "What links here", I cannot see a history button. It seems bolted onto the side rather than part of the wiki.
I know it is meant to be a viewer but that does not mean it cannot have the same interconnectedness that the rest of the wiki has. Frankly the way the whole site was indexed and linked together and revisioned is what drove me to edit here.
I like the aesthetics and I like the ability to link to it with an associated article. I think it can be a great tool, but I think there needs to be significant community input. Input where people make actionable improvements to make it wanted, and constructive criticism to address valid concerns.
I don't think it should be a polarized use it or don't use it discussion. Rather it should be an attempt to improve the encyclopedia. Just like we don't delete an article that can be improved to standards this tool can be made valuable. Chillum 10:58, 1 August 2014 (UTC)[reply]
The above message of Chillum (talk) was copied from User_talk:Jimbo_Wales&oldid=619532423#Mediaviewer. --Bob K31416 (talk) 12:35, 2 August 2014 (UTC)[reply]
Chillum, Re icon based and "I found myself annoyed at having to hover over each thing to see what it did." — How does that compare to the row of 11 icons above each editing box when one edits Wikipedia? --Bob K31416 (talk) 13:00, 2 August 2014 (UTC)[reply]

I don't think much of the iconic edit buttons on the edit page, I never user them. However in that case there is some justification as these are the same symbols editors have been using since Macs in the 80s and there are so many of them that text just would not work due to the requirement that it be horizontal. However in the image viewers the icons are arranged vertically thus making text fit far better.

I think it should have the 4 basic nav links that each page has: What links here, Related changes, Permanent link and Page information. These are the tools that make every page interconnected. There is no other part of Wikipedia where you cannot find these tools on a page.

If text based buttons were used, even in a stylish font, there would be room for all existing buttons and the standard 4 nav links. There should also be a link to the file's wiki page and its talk page both indicating in red if they are empty. Chillum 15:58, 2 August 2014 (UTC)[reply]

@Bob K31416:: You think those icons are useful? I never use them (either). And icons used by MV are not anywhere near a de-facto standard convention; the editor icons, I can sorta figure out what they'd do if I clicked on them. And I don't see 11 of them. only 5 are icon-only (bold, italic, some squiggle that's probably a signature insertor, a link of some kind, and probably an image insertion) the rest "Advanced", "Special Characters", "Help", "Cite" are rendered in text at least in Google Chrome. JMP EAX (talk) 20:03, 11 August 2014 (UTC)[reply]

Questions to help us think through what an NPOV statement will be like

[edit]

1. "MV seems to have been implemented prematurely as a default viewer for everybody (numerous bugs, some violent protests)". This is clearly at least in part an overstatement as there appears to have been no actual "violent protests" (bricks through windows? snipers on rooftops?). Is it possible to restate this as something like "In a poll, 83 people, which is 92% of the participants, felt that the default viewer had been released prematurely and that numerous bugs should have been fixed first." With a link to that poll, if it exists?

This is relevant because "released too early due to bugs" means that there is an actionable thing to do which is fix those bugs pronto. A prioritized listing would be ideal.

2. "MV is not appropriate for editing work; many editors have considered the tool as a "nuisance" - stated this way there is at least the suggestion that no possible version of MV could ever be appropriate for editing work, leading people to a path which involves having a separate user experience for readers and editors which I think we can all agree is not ideal. This also doesn't raise any actionable fixes for the developers. Can we explain the key 3 or 4 reasons (if there are that many) why MV is not appropriate for editing work and suggest how to fix it while maintaining the increased usability for readers?

3. Further down on the page, there is an extended and persuasive argument that a big problem with MV in its current form is that key information such as categories, which articles use the image, copyright status, and other versions are hidden behind a "mysterious" interface. Could we have a concrete proposal to the Foundation about how to fix this? Would it be sufficient to have a big obvious link to "Full info" which *instantly* (through the magic of javascript) unhides that information? Would it be better to have the information directly on the page when you scroll down? What are the plausible objections to or problems with these two solutions?--Jimbo Wales (talk) 10:55, 1 August 2014 (UTC)[reply]

Bugs have been logged. Attempts by the community to get bugs prioritised in the past have failed miserably. Again this speaks to the fundamental work that we need to do on the relationship between the community and the foundation. All the best: Rich Farmbrough00:08, 2 August 2014 (UTC).
  • @Jimbo Wales: -- After reading carefully your objections I came to the conclusions that our interpretations of the conflict are so different that we will hardly agree on a NPOV text. For example, you appear to take as a fait accompli that WMF decision concerning MV is irrevocable and that we should just live with it, trying to minimize the damage. I don't. For me (as well for many other editors) no fruitful discussion is possible while MV remains enabled as a default. That is clear from the lack of response to the recent attempts of the MV team to discuss the subject. I believe that such little step would immediatly clear the heavy clouds still in the sky. To be benign, it is a mistery for me why they haven't take it before. Alvesgaspar (talk) 13:40, 8 August 2014 (UTC)[reply]