Wikipedia talk:VisualEditor/Archive 5
This is an archive of past discussions about Wikipedia:VisualEditor. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Archive 1 | ← | Archive 3 | Archive 4 | Archive 5 | Archive 6 | Archive 7 | Archive 8 |
video on this page
Hi, I'm using Safari 4.0.5 on OS-X 10.6.2 on an Advent 4211 (Intel Atom) & I find that the video gets as far as "Hi, I'm Trevor" & just freezes. The 'Menu' button continues to work, so I downloaded the smallest version & it played fine with VLC Alanthehat (talk) 22:46, 1 August 2013 (UTC)
- I've got Safari 6 on OS 10.7.5, and I didn't even get that far. It worked on Firefox 22, though. Whatamidoing (WMF) (talk) 01:54, 2 August 2013 (UTC)
Lead section of this page
I'd like some other opinions about the lead for this page. Here are two competing versions:
That VisualEditor is in Beta means that it lacks the full functionality of the older wikitext editor. But while experienced editors will be more comfortable with wikitext editing, VE has features that are intended to be more user-friendly for new editors. For example, there are dialogs to insert a new image, or to edit a complex templates or a reference. It's hoped you'll give it a try, and the developers at the Wikimedia Foundation as well as volunteers have tried to make the wikitext editor be as accessible as possible. However, if you still wish to opt-out of using it, see Wikipedia:VisualEditor/Opt-out. |
That VisualEditor is in Beta means that it is hobbled both by bugs and by its incomplete features, lacking the full functionality of the older wikitext editor. But while experienced editors will be more effective with wikitext editing, VE is intended to be more user-friendly for new editors; for example, there are dialogs to insert a new image, or to edit a complex templates or a reference. After a huge consensus in its favor, the WMF developers have implemented a user preference to disable the VisualEditor. Go to Special:Preferences#mw-prefsection-editing, and check the box labeled "Temporarily disable VisualEditor while it is in beta", then click "Save". If you're having trouble disabling the VisualEditor, or to hide it instead of disabling it, see Wikipedia:VisualEditor/Opt-out for more detailed instructions. |
I see three main problems with the longer version:
- It uses biased language like "hobbled" and "huge consensus".
- It is redundant. It has two links to the opt-out page. It uses redundant phrasing like "incomplete features, lacking the full functionality".
- It makes a post hoc ergo propter hoc claim. AFAICT, the linked discussion had no significant effect on the decision to provide an opt-out feature. I'm not even certain that the people making the decision were aware of that discussion's existence.
Which version do you think is best? Can you suggest an alternative that you would prefer to either of these? Whatamidoing (WMF) (talk) 22:36, 31 July 2013 (UTC)
- Huge consensus is simply a statement of fact, arguably so is hobbled. As for the shorter version, it gives a completely false impression of the usability of the current version of the v/e, it doesn't even include the warning "Visual editor is an experimental tool that will run very slowly on any but the newest computers, and will sometimes do collateral damage to articles that you didn't intend to do. It is essential that anyone testing it also use the classic editor to look at the edit that they've just done and repair any unintended damage." There is also the nonsense phrase "edit a complex templates" - possibly that just requires removal of "a". As for "have tried to make the wikitext editor be as accessible as possible" that really needs a caveat such as "have tried to make the wikitext editor be as accessible as possible, providing you are using a state of the art new computer". I'd suggest that it would be more honest to replace That VisualEditor is in Beta means that it lacks the full functionality of the older wikitext editor. with We have a new Visual Editor available for editors to test. It can't yet do everything that the classic editor can do, it still has some bugs, and it runs very slowly on any but the newest machines. Click here if you agree to test it and report any bugs that you experience. ϢereSpielChequers 11:47, 1 August 2013 (UTC)
- Hi all, just my 2c here - the box at the top of the page should summarize the contents of the page itself, from what I understand. If we list there all the VE bugs we are most annoyed with, that's not a "nutshell" anymore :p Everything can be explained better in the rest of the page or in relevant pages, if it isn't yet (I recall the first part of the Known Problems page being very effective in this). But there should just be a NPOV message which recognizes "VE is not perfect at all". As for the "look at the edit that they've just done and repair any unintended damage" part, for instance, on it.wp we included a similar message in the pop-up which warns users that they're using wikitext. It is just as an example of a message that needed to be conveyed, simply, that looked a better place to do so. :) --Elitre (WMF) (talk) 12:00, 1 August 2013 (UTC)
- (Notice the message is being edited after Whatamidoing's proposal. --Elitre (WMF) (talk) 12:45, 1 August 2013 (UTC))
- WSC, I've taken your suggestion and used it to replace the previous sentence. I can't give you the "click here if you agree" bit. Also, I believe we've had the occasional complaint about slowness even from people who have fast machines (and others with older machines that claim no such troubles), so I've said "for many users". I don't want someone with a fast machine to feel unwarned if their experience is slower than they expected. Perhaps that will seem like an improvement to everyone. Whatamidoing (WMF) (talk) 18:13, 1 August 2013 (UTC)
- Hi all, just my 2c here - the box at the top of the page should summarize the contents of the page itself, from what I understand. If we list there all the VE bugs we are most annoyed with, that's not a "nutshell" anymore :p Everything can be explained better in the rest of the page or in relevant pages, if it isn't yet (I recall the first part of the Known Problems page being very effective in this). But there should just be a NPOV message which recognizes "VE is not perfect at all". As for the "look at the edit that they've just done and repair any unintended damage" part, for instance, on it.wp we included a similar message in the pop-up which warns users that they're using wikitext. It is just as an example of a message that needed to be conveyed, simply, that looked a better place to do so. :) --Elitre (WMF) (talk) 12:00, 1 August 2013 (UTC)
- Comment The statement could be extended from "VE has features that are intended to be more user-friendly for new editors. For example, there are dialogs to insert a new image, or to edit a complex templates" to add "... which are useful for a new user who is comfortable working with such terms as 'parameter' and 'transclusion'." Honestly folks, can anyone defend that as suitable for a new user? Does anyone seriously suggest that being confronted with such a dialog box would not produce 100% the wrong effect and have the user just looking for a way out of this over-technical minefield into which they have strayed and concluding that Wikipedia editing is not-for-the-likes-of-me? AllyD (talk) 06:30, 3 August 2013 (UTC)
- Comment Separating this more positive comment from my previous on templates, I do quite like the way VE presents potential images to the user. That said, it may increase the danger of encyclopaedic articles being turned into picture galleries. For example the Scotland article has a long history of this from one or more banned users. AllyD (talk) 06:30, 3 August 2013 (UTC)
Bug tracking graphs
Looking at the graphs linked from the August 2013 update, do the "Resolved" figures include "Resolved Duplicate"? If so, is it possible to have a graph that filters them out as they aren't a good indication of the number of actual issues resolved. Thryduulf (talk) 10:32, 2 August 2013 (UTC)
Temporarily disable
Cross-posted from Wikipedia:VisualEditor/Default State RFC#Was this just turned on for all users?
I do hope that the option "Temporarily disable while in beta" doesn’t imply what I think it does. I have nothing against visual editor, but I don’t want to use it. I don’t want to temporarily disable it, I want to do so permanently. Why was this changed? RGloucester (talk) 15:05, 2 August 2013 (UTC)
- I agree totally with you. I find the statement "Temporarily disable while in beta" quite ominous. Ellin Beltz (talk) 15:08, 2 August 2013 (UTC)
- Should I also cross-post my answer there now? I understood these pages had different aims, so I am not sure why we would just replicate contents. Thanks, --Elitre (WMF) (talk) 15:20, 2 August 2013 (UTC)
Focusing on improvements
I think we should recognize that VisualEditor is the future and try to spend time and energy focusing on possible real, actionable improvements to VisualEditor. --MZMcBride (talk) 17:17, 31 July 2013 (UTC)
- Why not support the wikiproject proposal instead of making more new buried pages? There seems to be good support, so why not just start that page? Risker (talk) 17:24, 31 July 2013 (UTC)
- The two aren't incompatible (i.e., we can have both a WikiProject and an "Improvements" subpage). :-)
- But reasonable suggestions for actionable improvements to VisualEditor are being drowned out in policy questions. One way to filter signal from noise is to split off pages, as appropriate. Moving the improvements subpage to be a subpage of Wikipedia:WikiProject VisualEditor rather than Wikipedia:VisualEditor is a trivial change that we can make in the future. --MZMcBride (talk) 17:42, 31 July 2013 (UTC)
- Perhaps. But realistically, the splitting off of subpages is burying ideas and limiting discussion of them to the few people who manage to find them. Risker (talk)
- Any WikiProject will use the exact same subpage system. Only with a WikiProject the subpages will be below a new and unknown WikiProject (e.g., Wikipedia:WikiProject VisualEditor) rather than the main project page (i.e., Wikipedia:VisualEditor). I'm not sure I see your point. :-) --MZMcBride (talk) 03:05, 1 August 2013 (UTC)
- Perhaps. But realistically, the splitting off of subpages is burying ideas and limiting discussion of them to the few people who manage to find them. Risker (talk)
Au contraire! This attempt at a visual editor is a collosal waste of everyones time and can never work. At the current stage, German Wikipedia has rejected it totally, Dutch wikipedia fortunately never tried and about 88% of English Wikipedia users wont use it. Technically it cannot be fixed- time taken re-arranging the deckchairs is pointless. The WikiProject idea is an attempt to do similar job correctly- but it needs to roll back to the design stage, to establishing a programming paradigm and establishing the functional specifications after extensive analysis of users needs. -- Clem Rutter (talk) 13:28, 2 August 2013 (UTC)
- I don't think that it's at all given that it's the future. See eg. [Wikipedians say no to Jimmy's 'buggy' WYSIWYG editor; see also the preliminary results that have shown that the editor is actually driving people away and failing catastrophically at its primary intended purpose. While obviously this is something the WMF intends to push, they don't operate entirely in a vacuum, and if a united voice of rejection continues to come out of every wiki they roll this out on -- combined with statistics continuously showing that it's failing to achieve its goal -- then it will be killed or shelved eventually. Even if we can't achieve that, we can at least push it as far into the background as possible until the resources intended to improve it are redirected somewhere more useful and it is quietly forgotten. At some point they need to show that either people like the editor, or that it's improving editor retention; and while it might do better than it's doing now, there's no reason to think it will ever achieve either of those goals. --Aquillion (talk) 02:02, 4 August 2013 (UTC)
Yet another issue that needs to be dealt with...
Right, I've been hesitant to mention this, but it seems to be starting to be a significant problem. At a lot of articles that I have watchlisted (for example, USS Enterprise (CVN-80) and Deadpool (comics)), we use invisicomments (inside the <!-- --> tags) to indicate items that are correct, but frequently "corrected" as "simple mistakes" or "typos" by IPs and other new editors--in those two examples, the ship currently being designated as PCU Enterprise instead of USS Enterprise, and the Marvel Zombies version of Deadpool being known as "Headpool." These invisicomments were put in place specifically to stop inexperienced editors from changing them, with the idea being that when they go to edit it, they get these big-ass notices telling them that these are correct and not to unilaterally change them. Every time I've seen this done, it almost completely eliminated the issue of inexperienced editors constantly making these "corrections." However, since the Visual Editor has gone live, these comments have ceased to be effective. Given that the whole point of having the invisicomment tag is to provide editors with information that's important to those who are going to edit the article, but shouldn't be shown to readers, I think we need to have a change made to the VE to somehow display these comments. I don't know *how* we should do it, but my initial thought is maybe double-spacing in the editor, with the invisicomments being shown between lines with an arrow indicating their location, or something. Something that a new editor would intuitively understand isn't shown to the readers, but wouldn't be able to miss, either. rdfox 76 (talk) 23:26, 1 August 2013 (UTC)
- Yes, you're right. Something has to be improved there. In between now and then, you could use an WP:Edit notice to pop up a warning in the article. Whatamidoing (WMF) (talk) 01:04, 2 August 2013 (UTC)
- ...wow. Speaking of dense and user-unfriendly documentation, the Edit Notice instructions are *really* awkward for non-admins. x.x Thanks for the tip, I'll look into it tomorrow when I'm more... conscious. rdfox 76 (talk) 02:59, 2 August 2013 (UTC)
- Well, I'm sure there are limitations to my approach, but I've had really good success with a one-step non-admin process in the past:
- Leave a nice note at User talk:MSGJ.
- Perhaps that would work for you, if the other process doesn't work. Whatamidoing (WMF) (talk) 17:57, 2 August 2013 (UTC)
- Came here to make this point, and saw that it was already brought up. Just to give another example, see here. In this case, an edit notice for a single word would be overkill. Parsecboy (talk) 16:00, 3 August 2013 (UTC)
- Out of personal curiosity for this diff (I'm from it.wp), is British English "banned" here? Do we have a British English Wikipedia or need one? :D --Elitre (WMF) (talk) 16:09, 3 August 2013 (UTC)
- WP:ENGVAR: you're supposed to follow whatever flavor is already in the article and can write whatever you want on talk pages. Of course, most people don't know how to switch between versions, so people mostly write whatever they're used to and hope that someone else will clean it up if necessary. WhatamIdoing (talk) 21:11, 3 August 2013 (UTC)
- Out of personal curiosity for this diff (I'm from it.wp), is British English "banned" here? Do we have a British English Wikipedia or need one? :D --Elitre (WMF) (talk) 16:09, 3 August 2013 (UTC)
- Another example: [1]. (And we really can't subcontract editing advice to administrator-only tools.) AllyD (talk) 16:11, 3 August 2013 (UTC)
- Her advice was for a temporary solution, of course it was not intended as the bug fix, and I mentioned your concerns again in the ticket. Thanks. --Elitre (WMF) (talk) 16:33, 3 August 2013 (UTC)
- Came here to make this point, and saw that it was already brought up. Just to give another example, see here. In this case, an edit notice for a single word would be overkill. Parsecboy (talk) 16:00, 3 August 2013 (UTC)
- Well, I'm sure there are limitations to my approach, but I've had really good success with a one-step non-admin process in the past:
- ...wow. Speaking of dense and user-unfriendly documentation, the Edit Notice instructions are *really* awkward for non-admins. x.x Thanks for the tip, I'll look into it tomorrow when I'm more... conscious. rdfox 76 (talk) 02:59, 2 August 2013 (UTC)
No longer hiding
The "hide VisualEditor" section of the preferences no longer seems to work. What happened? Ten Pound Hammer • (What did I screw up now?) 07:01, 2 August 2013 (UTC)
- Yes, having clearly expressed in my preferences that I don't want to see this damaged-goods Beta, it has now reappeared and is polluting the action tab. Some people don't seem to understand the meaning of No. AllyD (talk) 07:09, 2 August 2013 (UTC)
- Is there any way to hide this "beta edit" tag? Canuck89 (converse with me) 08:35, August 2, 2013 (UTC)
- Yes, it now seems you have to disabled the "tools" again at the bottom of Special:Preferences#mw-prefsection-editing. Stalwart111 08:44, 2 August 2013 (UTC)
- So if you have previously selected "hide VisualEditor", you also need to select "Temporarily disable VisualEditor while it is in beta" from the link above. At least, I did. Getting to be a bit of a joke. Stalwart111 08:47, 2 August 2013 (UTC)
- Is there any way to hide this "beta edit" tag? Canuck89 (converse with me) 08:35, August 2, 2013 (UTC)
- (Kinda cloning from another page :) ) What you all experienced was not intended at all (see Wikipedia:VisualEditor/August_2013_update). It is my understanding that the recent changes happened to somehow "break" the unofficial, community-developed gadget that you were using. You can still use the official one which you can find in your Preferences, as noted. Sorry if this caused trouble. --Elitre (WMF) (talk) 09:18, 2 August 2013 (UTC)
- The person who wrote the gadget has publicly recommended that people use the "real" (although temporary) opt-out switch, at least for as long as it exists, because he believes that it will work better and more consistently. So if you've been using the old gadget, I suggest that you switch to the new prefs item. Whenever the prefs switch is turned off (last I heard, it will at least be around for a few months), you can always go back to the old gadget. Whatamidoing (WMF) (talk) 18:08, 2 August 2013 (UTC)
- This, and the reaction to the present switch, strongly suggests that it is important to make sure this particular preference is preserved when implementation changes. e.g. that (1) there is sufficient warning when the present off-switch stops working (2) that the preference is transferred when it moves back to a gadget. The WMF did not make the VE mandatory, but it most definitely worked way too hard to try to steer people into using it, including by accident. Functionally and behaviourally, this reminds people of Facebook. And I appreciate it's unpleasant to have people not assume good faith in you, but what they're actually doing is assuming that past behaviour is the best predictor of future behaviour; and if that gives a negative answer, then it's actually up to you to behave in a manner that appears better than you have been. The WMF is behaving better now, but you cannot reasonably expect people to instantly forget how you were behaving during the past couple of months; trust has to be earned back - David Gerard (talk) 20:23, 2 August 2013 (UTC)
- The person who wrote the gadget has publicly recommended that people use the "real" (although temporary) opt-out switch, at least for as long as it exists, because he believes that it will work better and more consistently. So if you've been using the old gadget, I suggest that you switch to the new prefs item. Whenever the prefs switch is turned off (last I heard, it will at least be around for a few months), you can always go back to the old gadget. Whatamidoing (WMF) (talk) 18:08, 2 August 2013 (UTC)
- (Kinda cloning from another page :) ) What you all experienced was not intended at all (see Wikipedia:VisualEditor/August_2013_update). It is my understanding that the recent changes happened to somehow "break" the unofficial, community-developed gadget that you were using. You can still use the official one which you can find in your Preferences, as noted. Sorry if this caused trouble. --Elitre (WMF) (talk) 09:18, 2 August 2013 (UTC)
- I don't think anyone is assuming bad faith. It would mean that someone is assuming that WMF is trying to cause trouble. For example, that they know that "VisualEditor" is terrible and are deliberately pushing it to annoy the users. I don't think anyone thinks WMF is doing that.
- On the other hand, "Assume good faith" in many cases means "Assume incompetence". That's why (given that we have Wikipedia:Competence is required) we can demand total compliance to this principle - to the limit of "naivete", until sanity gets threatened. And I am sure that most of us are assuming just that - that WMF is highly incompetent. That is, they sincerely think that "VisualEditor" is useful, because they really know little about usability. They insult, annoy and anger us by paying no attention to our legitimate concerns, because they do not have sufficient social competence to see that they are doing this. They leave many major bugs, because they do not have sufficiently good programmers, testers and managers. And because they have too much enthusiasm - that is a great vice in each bureaucracy.
- So, I would ask you not to talk as if someone is not assuming good faith. Please, remember Wikipedia:Assume the assumption of good faith. --Martynas Patasius (talk) 17:45, 3 August 2013 (UTC)
Article: Mention Scripts Enabled Required Prominently
I am a very casual user. I usually access and edit on Wikipedia with JavaScripts blocked. I don't recall ever having a problem with functionality in editing because of scripts being blocked. So, when I noticed the "Request for Comment on implementation" I was looking at this article trying to figure out why I didn't see VisualEditor even though I was logged in and how to enable it. After a while, I realized it was because I had scripts blocked. I wonder if it might be a good idea to prominent note early in this article that this is script enabled. Or if there is little chance of other users as dumb as me with script blocking on their end (lol). jrun (talk) 05:57, 4 August 2013 (UTC)
- According to my quick test, when javascript is disabled (1) the Edit beta tab is hidden, and (2) if you get around that by following a URL with the
&veaction=edit
parameter, the page you reach isn't editable. Are you saying that you'd like the Edit beta tab to be visible at all times but when javascript is disabled it takes you to a page that contains a no script message? - Pointillist (talk) 16:52, 4 August 2013 (UTC)
Proposed merge with VisualEditor
I propose this article be merged into MediaWiki and WP:VisualEditor. Per the discussion at Wikipedia:Articles for deletion/VisualEditor, it seems most people agree that this content belongs on Wikipedia but it is not notable enough to deserve its own article. Beerest355 Talk 21:34, 4 August 2013 (UTC)
Internet explorer users
I have noticed all the recent adaptations of "edit" to "edit source" etc. and the header messages about Visual editor but am amazed that there are no clearer explanations about why the beta Visual editor is not available to Internet Explorer editors. I have pressed for a WYSIWYG approach to Wiki editing for years but have always been bulldozed down by the admins and crats. Now you introduce a beta version but only for those using What?--Ipigott (talk) 21:42, 4 August 2013 (UTC)
- You will probably get a better answer by asking your favorite web search engine, because the problems for VE are the same as the problems for everyone. The short answer is that IE8 and earlier were not standards-compliant (so you create one thing for the rest of the world, and re-write it just for IE users), and IE9 (when they re-wrote IE to make it compliant) is missing some necessary features and still has some quirks that need to b worked around. Whatamidoing (WMF) (talk) 20:08, 5 August 2013 (UTC)
Infobox deletions
For background, I go through the lists of Category:Orphaned non-free use Wikipedia files to see if all of the files were removed correctly and if not to add them back to the article with an explanation. In doing this since July 22, I have come across eleven edits from eleven different editors ranging from July 2 to August 2, where the editor makes an edit using VisualEditor and ends up removing the infobox.
[2], [3], [4], [5], [6], [7], [8], [9], [10], [11] and [12]
Maybe some of these editors wanted the infobox deleted, I cannot tell in the cases where no edit summary is left, but where there are edit summaries, they are about making a change to the lead paragraph(s). I was wondering how easy/hard it is to edit the lead paragraph(s) and accidentally delete the infobox. These infobox deletions just deal with fair use images, I do not know how we would know how many infoboxes were deleted that had no image or had a free image in it. Aspects (talk) 20:14, 5 August 2013 (UTC)
- Thank you for your note. It is extremely easy to delete an infobox: you just backspace once at the start of the lead paragraph. VE sees infoboxes as one "character", and they are normally placed immediately above the lead sentence. I think it would be safe for you to assume that 100% of these removals were accidents unless the edit summary specifically mentioned the infobox.
- This is a known problem, and I am collecting suggestions for how to address it. One idea was to label certain templates as requiring confirmation before deleting them. Whatamidoing (WMF) (talk) 20:38, 5 August 2013 (UTC)
- Rather than asking for permission after the user has started an action that would delete the box, it's better to provide feedback before a keypress would remove the content. An approach to this may be to shadow the whole template box with a grey overlay when the text insertion caret is located right before or after it, to inform that a Backspace or Del key would remove the box. Diego Moya (talk) 20:52, 5 August 2013 (UTC)
- It would be even better to treat a deletion key when the caret is besides a template as if it deleted a white space - i.e. instead of deleting the template right away, first select the whole box, and only then delete the box when Del is pressed for the second time. This is how the left-right cursor navigation keys work now, just do the same when a deletion key is pressed. Diego Moya (talk) 20:56, 5 August 2013 (UTC)
- Rather than asking for permission after the user has started an action that would delete the box, it's better to provide feedback before a keypress would remove the content. An approach to this may be to shadow the whole template box with a grey overlay when the text insertion caret is located right before or after it, to inform that a Backspace or Del key would remove the box. Diego Moya (talk) 20:52, 5 August 2013 (UTC)
Changing the user preference label
Hi. I've begun a discussion at MediaWiki talk:Visualeditor-preference-betatempdisable#Changing this message to discuss whether we should change the user preference label that completely disables VisualEditor. Please participate there. --MZMcBride (talk) 06:52, 6 August 2013 (UTC)
VE as WYSIWIG editor
VE is not described as a WYSIWYG editor. Why is that? I think using a term many people are familiar with (together with a link to the WYSIWYG-article) would help reduce misunderstandings of what VE is (especially by wikitext veterans). I've seen arguments that VE is not really WYSIWYG, but I don't buy that. If it looks and quacks like a duck, then it is a duck. --Frederico1234 (talk) 18:20, 6 August 2013 (UTC)
- It's technically a rich-text editor, which is sort of "what you see is almost, but not quite, what you get". Whatamidoing (WMF) (talk) 22:49, 6 August 2013 (UTC)
Should inability to edit ISBNs be listed in the limitations section?
Wikipedia:VisualEditor#Limitations is a bullet list of various known limitations. One of Mediawiki's features is if you enter ISBN 1-4013-0371-4 in wikitext it gets displayed as ISBN 1-4013-0371-4. If you use VE to edit that contains an automatic ISBN link and hover your mouse over the displayed ISBN then the following happens:
- The mouse cursor changes to the No symbol.
- A transparent graphic with alternating gray/green bars gets dropped over the ISBN text.
- A hovertext message pops up with "Sorry, this element can only be edited in source mode for now."
You can see this in Roger Ebert#Bibliography which I tried to clean up using VE but ended up using source mode. Clearly the VE developers know there's a problem but is it a large enough issue that it should be listed on the limitations of this article either as its own bullet or as part of the Incomplete editing functionality bullet? --Marc Kupper|talk 02:01, 8 August 2013 (UTC)
- Its in bugzilla as T54204. You can add it to Wikipedia:VisualEditor/Known problems which is a less discriminating list. I would not have it in the limitation page as ISBN's generally get covered by citation templates.--Salix (talk): 05:47, 8 August 2013 (UTC)
- I believe that you can add and remove these magic-word elements. I've done it with PMIDs. But you can't just change the number right now. Instead, you'd have to type a new (corrected) instance and then delete the old one.
- BTW, I pulled a couple of old items off the "known issues" list, since they're no longer issues, and updated a few bug numbers. I haven't gone through the long "details" section yet, but it should be done at some point (feel free, if you want to). Whatamidoing (WMF) (talk) 17:11, 8 August 2013 (UTC)
- I get different results depending on whether I'm clicking on a hyphenated or non-hyphenated (previously saved) ISBN. I've added comments to the bug report. Whatamidoing (WMF) (talk) 17:26, 8 August 2013 (UTC)
Survey on VE
Let's see how many users who use VE and find it useful. A survey like this is a great help for the project. LiquidWater 10:57, 16 July 2013 (UTC)
- By function, a survey here is not going to reach most editors, since few of them are actually visiting this page, so you will have a selection bias towards those who are actively interested in the deployment of the software, who are not likely to be neutral. :) Of course, there can be value in seeing how those people feel, but it won't offer much insight into "random IP editor Sally" or "dedicated wikignome George". (Just noting.) --Maggie Dennis (WMF) (talk) 11:44, 16 July 2013 (UTC)
- +1 to that.--¿3family6 contribs 17:49, 17 July 2013 (UTC)
- I agree with you, Maggie; however, I'd suggest that it's probably time to get some more scientifically collected results from the general community. The people on this page may be better informed about VE and its challenges than the average editor, but I'd be hesitant to say they're any less neutral than the rest of the community - at least without stats. Risker (talk) 18:10, 17 July 2013 (UTC)
- Not directly related, but I'm actually plotting out some research on precisely that point: how well metapedians' outlook represents wikipedians as a whole. Okeyes (WMF) (talk) 18:12, 17 July 2013 (UTC)
- Of course, stats make all the difference. :) But in my subjective experience, people who watch (and participate) in talk pages related to policies or initiatives tend to have stronger feelings about said policy or initiative (whether positive or negative). --Maggie Dennis (WMF) (talk) 19:03, 17 July 2013 (UTC)
- Maggie is right - I'm only here to register the very strongly held opinion that VE is an extremely bad idea. Unfortunately, this survey doesn't have a section for people who think that VE should not be rolled out, and indeed should be rolled back immediately. If I'd known about VE earlier, I'd have spoken out against it then. As it is, I'm just an occasional editor, so I didn't know about VE until it was way too late. There's no point in saying anything now, of course - too many people are invested in the project, so we're just stuck with it. Oh well... RomanSpa (talk) 20:46, 10 August 2013 (UTC)
- Of course, stats make all the difference. :) But in my subjective experience, people who watch (and participate) in talk pages related to policies or initiatives tend to have stronger feelings about said policy or initiative (whether positive or negative). --Maggie Dennis (WMF) (talk) 19:03, 17 July 2013 (UTC)
- Not directly related, but I'm actually plotting out some research on precisely that point: how well metapedians' outlook represents wikipedians as a whole. Okeyes (WMF) (talk) 18:12, 17 July 2013 (UTC)
I use VE for most of my article edits.
Yes
- The only reason I don't use it as much now is because of wikitable and some template edits. But now that I'm getting the hang of inserting references, I use it the majority of the time.--¿3family6 contribs 17:49, 17 July 2013 (UTC)
- It's definitely my editor of choice now. WaggersTALK 11:41, 18 July 2013 (UTC)
No
- It does not work very well for me, giving both a lot of lag, and being particularly unsuitable for image work. Adam Cuerden (talk) 11:58, 16 July 2013 (UTC)
- I edit math articles. So, VE is completely useless to me (and will remain to be so in a near future.) -- Taku (talk) 20:05, 16 July 2013 (UTC)
- I'll try again in a few months. Looks like there are enough bugs reported to keep people busy for at least that long. EllenCT (talk) 23:23, 16 July 2013 (UTC)
- The only time it has opened was when I accidentally clicked on it. It looked confusing and I had no interest in trying it out. Apteva (talk) 23:25, 16 July 2013 (UTC)
- Not at all. I help out from time to time on the IRC-help channel, can't post helps/links on user talk pages with it. I also work on some fairly long historical biography articles and not having section-editing ennabled makes VE basically unusable for working on long articles. For me the processing time and hangtime is already at the breaking-point even when I work on a section, to then add VE's issues on top of the present wikitext system?...I just don't even want to try it out at this point. I rarely see VE pop up on any of the articles I have on my Watchlist & when I do it's placing unwanted <nowiki> tags within edits. Shearonink (talk) 23:42, 16 July 2013 (UTC)
- It's quicker to use "Edit Source" for adding links and spelling changes, or for building election results. doktorb wordsdeeds 09:52, 18 July 2013 (UTC)
- Most of my edits are to sections, and loading the section content only is way faster than reloading the whole article + waiting the VE to parse it. Diego (talk) 17:07, 18 July 2013 (UTC)
- It's much easier for me to use the source editor as it is what I am used to. Although I will keep an open mind about it, I have no interest in using it. --Breawycker (talk to me!) 05:12, 25 July 2013 (UTC)
- I don't find visual editor useful. It's a lot of "go around the barn 3 times and stand on your head," when plain old editing has the problem fixed and solved in the time it takes VE to load the page one time. Ellin Beltz (talk) 18:14, 25 July 2013 (UTC)
- 120.145.218.98 (talk) 02:42, 26 July 2013 (UTC)
- I've tried to use it, and sometime I still do, to perform trivial edits on very short articles, but it's so difficult to do anything more complex than a trivial edit, and so slow to open longer articles, that it's not worth even bothering to do so on anything more complex. -- The Anome (talk) 07:01, 26 July 2013 (UTC)
- Per all above. Too many bugs, too inefficient, too clunky. Resolute 15:14, 26 July 2013 (UTC)
- Mostly because of the reasons listed here. I don't want to kill it with fire or anything; it's just not efficient, especially when you're already familiar with the old way. --BDD (talk) 19:00, 26 July 2013 (UTC)
- Same as above.--Sinuhe20 (talk) 19:53, 27 July 2013 (UTC)
- Very confusing for image work, which I do a lot of. FunkMonk (talk) 16:14, 28 July 2013 (UTC)
- Slow to load and absolutely unnecessary. It only wastes my time. Thegreatgrabber (talk)contribs 01:15, 31 July 2013 (UTC)
- No per all of the above. Darkwarriorblake (talk) 08:36, 31 July 2013 (UTC)
- I prefer using the source editor (like right now) because I feel like I have more control over what I'm doing (likewise, I've always preferred writing HTML/CSS over a WYSIWYG editor).katacarbix (talk) 23:26, 31 July 2013 (UTC)
- My first reaction was, "What the @#$# is this"? I still don't understand how it works, and frankly don't want to bother. The system now works fine, why change? I still only use the new upload-to-Commons script once in a while, as it isn't easy to find a link to the copyright tags (I figured it out, but the old uploader had it right there at the bottom, not hidden under some menu). Don't change stuff, this works fine. Oaktree b (talk) 02:43, 1 August 2013 (UTC)
- I can't stand this thing, make a way to shut it off. --Jeremy (blah blah • I did it!) 07:03, 2 August 2013 (UTC)
- I think the VisualEditor is a great technical achievement, and I'm happy it's being developed. Just not for me. Over on OpenStreetMap, I'm very glad that the iD editor has been created, but I'll still use the not-particularly-beginner-friendly JOSM. People should be able to use whatever tool they find most comfortable. —Tom Morris (talk) 11:41, 2 August 2013 (UTC)
- The maon factor preventing me from using it except for testing is the slower speed it works. DGG ( talk ) 00:21, 3 August 2013 (UTC)
- I only use it for bug testing. For real edits, I highly prefer the legacy editor. --Cryptic C62 · Talk 13:05, 8 August 2013 (UTC)
- I did try to use it at first (I'd used visual editors on other wikis), but it was too much of a pain. The worst part was when I realized I couldn't finish what I'd started in the VE (or at least couldn't figure out how to finish it), and switched to the regular editor, only to find that switching cost you all your work. Bleh. --Aquillion (talk) 07:04, 9 August 2013 (UTC)
Comments
One of the reasons for introducing VE appears to be that it will make things easier for newbies: they are said to find it hard to use Wiki markup, so need WYSIWYG. I suggest we might make more mention to Newbies of the various markup Cheat Sheets available online. These can be down-loaded and perhaps laminated, ready for instant use when composing an article. Even tho' I've been an editor since 2004 I still need to refer to them for codings of edits I don't do very often. There does not appear to be cheat sheets for the more esoteric aspects of Wikiwork such as making templates, infoboxes, etc. These can be read on WP Help, of course. Apwoolrich (talk) 08:58, 9 August 2013 (UTC)
I find using VE easier than using the source editor,
Yes
- LiquidWater 20:01, 16 July 2013 (UTC)
- And it's improving all the time. WaggersTALK 11:42, 18 July 2013 (UTC)
- GerardM (talk) 15:57, 1 August 2013 (UTC)
- Yes, I do, because the markup is too confusing for me as I'm not a regular user. It is still pretty buggy though. 81.178.161.204 (talk) 10:07, 3 August 2013 (UTC)
No
- I can't even scroll in VE without severe lag. Adam Cuerden (talk) 11:58, 16 July 2013 (UTC)
- I have quite a slow connection, but VE is not lagging at all for me. LiquidWater 12:25, 16 July 2013 (UTC)
- My computer is not very high-end anymore; I suspect it's a processing issue. Adam Cuerden (talk) 17:33, 16 July 2013 (UTC)
- I have quite a slow connection, but VE is not lagging at all for me. LiquidWater 12:25, 16 July 2013 (UTC)
- Templates are impossible. I tried to change "Election box candidate with party link" to just "Election box candidate" the other day and was soon drowning in computer programmer language hell. VE is not easy to use if you're not a computer programmer. doktorb wordsdeeds 09:51, 18 July 2013 (UTC)
- There are a few citation templates (those with a high number of required parameters, and a small number of non-required parameters) that are improved by having the parameter list pre-filled. But for any other edit that would require wikicode, typing it directly is much less work and way faster. Diego (talk) 17:06, 18 July 2013 (UTC)
- I oppose the ideological framing of this question, which makes editors who prefer to edit in wiki markup out to be nay-sayers. Do I prefer wiki markup over a clumsy WYSIWYG editor? Yes! --85.197.3.203 (talk) 21:56, 20 July 2013 (UTC)
- Agree with previous comment. I too prefer wiki markup over a very clumsy WYSIWYG editor. Ellin Beltz (talk) 18:15, 25 July 2013 (UTC)
- Inability to edit substitutions is a deal breaker for this aspect. 120.145.218.98 (talk) 02:42, 26 July 2013 (UTC)
- No. It's too slow, too buggy, and lacks any equivalent for many more complex wikitext features. -- The Anome (talk) 07:01, 26 July 2013 (UTC)
- Not remotely close. Granted, I have eight years experience with the text editor, while this is a new and horribly flawed beta implementation, but I have a hard time imagining VE will ever match the efficiency of pure wikitext. Resolute 15:14, 26 July 2013 (UTC)
- What's with all the extra line breaks? I could see VE being useful for fixing a typo here and there, but I don't think it's up to the job of serious editing. --BDD (talk) 19:01, 26 July 2013 (UTC)
- --Sinuhe20 (talk) 19:53, 27 July 2013 (UTC)
- No, slow, bugged and less flexible. Darkwarriorblake (talk) 08:37, 31 July 2013 (UTC)
- No. Opaque, hard to use. WYSIWYG and all its associated flaws. I prefer the text mark up editor. Rwflammang (talk) 03:44, 1 August 2013 (UTC)
- No. I should mention that I have used visual-style editors a lot on other sites; but most of them consisted entirely or almost-entirely of pages that had been originally designed via a visual editor. The visual editor here feels crippled and generally frustrating to use due to the inability to modify the numerous vital places on Wikipedia that use complicated wikitext. --Aquillion (talk) 15:36, 1 August 2013 (UTC)
- No - I prefer the old way, but I am slowly turning into a curmudgeon... And get off my lawn. --Jeremy (blah blah • I did it!) 07:05, 2 August 2013 (UTC)
- I prefer editing wikitext. I'm perfectly happy for the VisualEditor to exist, and I want it to be the best possible editor it can be for people who prefer a WYSIWYG experience. I don't prefer a WYSIWYG experience, I prefer a plain text experience. I wish the VE team all the best, just not for me. —Tom Morris (talk) 11:40, 2 August 2013 (UTC)
- When it comes to user interfaces, there tends to be an inverse relationship between intuitiveness and power. Once you've become accustomed to using the powerful features of an interface, switching to an intuitive version becomes all loss and no gain. --Cryptic C62 · Talk 13:09, 8 August 2013 (UTC)
Comments
- I believe it is like any visual tool. Nice for beginners. You may start from there. But when it gets to real work, you open a text editor. --194.44.219.225 (talk) 11:49, 17 July 2013 (UTC)
- Depends on what I am trying to do. For simple content edits, yes. For wikitable stuff or complex templates, no.--¿3family6 contribs 17:49, 17 July 2013 (UTC)
How high have you found the number of bugs in VE to be?
Very high, not compatible with being available to the general public
- VE keeps damaging articles. That isn't compatible with a general launch, although I'd accept it in a test run. Adam Cuerden (talk) 12:00, 16 July 2013 (UTC)
- Only supported on some browsers. Relatively impossible to fill out cite ref tag fields. EllenCT (talk) 23:24, 16 July 2013 (UTC)
- To this day we are seeing major bugs. 120.145.218.98 (talk) 02:42, 26 July 2013 (UTC)
- It's still too buggy to use for serious editing. -- The Anome (talk) 07:01, 26 July 2013 (UTC)
- Over 600 open bugs are too high.--Sinuhe20 (talk) 19:53, 27 July 2013 (UTC)
- ϢereSpielChequers 21:57, 31 July 2013 (UTC)
- There are too many major bugs. Johnny Au (talk/contributions) 00:12, 1 August 2013 (UTC)
- VisualEditor is terrible right now and completely unready. It's producing so many bugs that it feels like I'm spending more time cleaning up mangled-syntax VisualEditor edits from new users than doing any productive new editing. —Lowellian (reply) 03:08, 2 August 2013 (UTC)
- Even if newbies don't experience the bugs themselves, their edits are still getting mangled. So no, not compatible with the general public. --Cryptic C62 · Talk 13:13, 8 August 2013 (UTC)
High, but still usable
- It's functional, just annoying.--¿3family6 contribs 17:49, 17 July 2013 (UTC)
Relatively low, but not optimal
- LiquidWater 20:00, 16 July 2013 (UTC)
- It does the basics very well and what bugs there are tend to be fixed quickly. WaggersTALK 11:44, 18 July 2013 (UTC)
Very low, close to perfect
Comments and other opinions
I believe that I will prefer VE over the source editor once it contains fewer bugs.
Yes
- Assuming bringing citation support up to regular levels is considered fixing bugs. EllenCT (talk) 23:25, 16 July 2013 (UTC)
- Yes. I already prefer it for article body content.--¿3family6 contribs 17:49, 17 July 2013 (UTC)
- I already do. WaggersTALK 11:45, 18 July 2013 (UTC)
No
- Not for me, since I do so much image work, which, ironically, is far easier to do with filenames than any sort of GUI could handle. I think it will be good for others, though. Adam Cuerden (talk) 12:01, 16 July 2013 (UTC)
- VE contains a lot of transclusions for templates that a baby could use, and with time, I believe that it will make image work easier too. LiquidWater 12:27, 16 July 2013 (UTC)
- Yes, that doesn't make it suitable for someone who often has to use the Find command to look for a filename, because I've just restored a larger version. I can't keep multiple image sizes in copy paste, but I can leave the "300px" or "50px" or whatever in place. Further, I want the captions to remain exactly the same, I don't want to have to copy them over. Cntrl-F, enter, Highlight, Cntrl-V is probably always going to be faster than anything VE can make, unless someone makes a tool specifically for my use. I'm not rejecting VE because I think it's a bad idea; I'm rejecting it because I have very specific editing requirements that no VisualEditor could reasonably be expected to handle.Adam Cuerden (talk) 17:37, 16 July 2013 (UTC)
- So don't use it. Simple. Okeyes (WMF) (talk) 18:45, 16 July 2013 (UTC)
- I did say that I thought it would be good for others. Had the WMF not thrown a fit about not wanting anyone to hide the user preference, I'd have shut it off and got on with things. Adam Cuerden (talk) 18:51, 16 July 2013 (UTC)
- So don't use it. Simple. Okeyes (WMF) (talk) 18:45, 16 July 2013 (UTC)
- Yes, that doesn't make it suitable for someone who often has to use the Find command to look for a filename, because I've just restored a larger version. I can't keep multiple image sizes in copy paste, but I can leave the "300px" or "50px" or whatever in place. Further, I want the captions to remain exactly the same, I don't want to have to copy them over. Cntrl-F, enter, Highlight, Cntrl-V is probably always going to be faster than anything VE can make, unless someone makes a tool specifically for my use. I'm not rejecting VE because I think it's a bad idea; I'm rejecting it because I have very specific editing requirements that no VisualEditor could reasonably be expected to handle.Adam Cuerden (talk) 17:37, 16 July 2013 (UTC)
- VE contains a lot of transclusions for templates that a baby could use, and with time, I believe that it will make image work easier too. LiquidWater 12:27, 16 July 2013 (UTC)
- Not bugs, but I need to be able to edit math formulas.... --- Taku (talk) 20:07, 16 July 2013 (UTC)
- Text editor and text editor only for that. --194.44.219.225 (talk) 11:55, 17 July 2013 (UTC)
- Admittedly I'm a special case here: I know Wikimarkup cold yet I don't expect to do a large amount of editting in the foreseeable future. At least not an amount to justify learning a new interface. If I am unable to edit the Wikimarkup directly for whatever reason, I will most likely stop contributing. (That's not a threat, just a simple statement. And some people might prefer I stop contributing.) -- llywrch (talk) 17:30, 17 July 2013 (UTC)
- No because I don't see any reason to use one markup system to work on talk pages and a totally different one to work on the articles. Ellin Beltz (talk) 18:16, 25 July 2013 (UTC)
- Unless they include editing of substituted templates, infoboxes and such 120.145.218.98 (talk) 02:42, 26 July 2013 (UTC)
- The bugs aren't the problem for me. The simple inefficiency is. VE is for novice users. Many experienced users will always prefer simpler + more powerful. Resolute 15:14, 26 July 2013 (UTC)
- I don't like clicking around (3 or 4 times for a small edit), I prefer editing in a textbox.--Sinuhe20 (talk) 19:53, 27 July 2013 (UTC)
- Bug are far from the only problem. V/E is designed to work on new fast CPUs, if your kit is a few years old then it is unacceptably slow. That is a pain for people like me who are otherwise happy with our four year old machines but can with some annoyance afford a few hundred quid to upgrade. But as a movement we have a global mission, that mission is undermined if the techies deliver a system that disproportionately disadvantages editors in parts of the world where the necessary upgrades are more than a months average income. ϢereSpielChequers 20:32, 27 July 2013 (UTC)
- No, bugs aren't the only issue, the only thing VE is good for is letting IPs make bad edits more efficiently than usual. Darkwarriorblake (talk) 08:38, 31 July 2013 (UTC)
- No, I'm still not 100% on all the ins and outs of the present editor with regards to citation tags, I don't need to learn something new when I haven't figured out all the knobs and buttons on this one yet... Oaktree b (talk) 02:45, 1 August 2013 (UTC)
- No. Far far too complicated, frustrating to do even the simplest tasks, and it's impossible to learn as you go by seeing how something was done elsewhere. It's an irredeemable mess. --Aquillion (talk) 15:37, 1 August 2013 (UTC)
- No. It is a really, really awful experience. GimliDotNet (Speak to me,Stuff I've done) 19:41, 1 August 2013 (UTC)
- It's not for me. It's great that it's being developed, but I don't see myself using it. —Tom Morris (talk) 11:43, 2 August 2013 (UTC)
- No. I was willing to give it a shot, but ultimately I'm utterly disinterested in using a visual editor for editing Wikipeida. --Aquillion (talk) 07:05, 9 August 2013 (UTC)
Comments
- Maybe, but probably not. I don't want to be the type that whines about every new change, but for now, VE is much harder and less efficient for the sort of editing I do. I can't really even conceive of it ever being more efficient for my uses, but it's the sort of thing I'd be willing to try revisit every few months to see if improvements can catch up. --BDD (talk) 19:03, 26 July 2013 (UTC)
I am not using VE because it is too slow for my computer/connection.
Agree
- Well, that's the main reason I haven't made a single edit in it. I probably wouldn't use it much, but I'd have tried it out a lot more if it was at all practical. Adam Cuerden (talk) 12:02, 16 July 2013 (UTC)
- Agree. It's slow. Instead of waiting for Visual Editor to load, I finish the edit by clicking "edit source". Kavas (talk) 01:44, 17 July 2013 (UTC)
- Agree. I'd enjoy using it for minor or text-only edits, but it's too slow for that. Diego (talk) 17:03, 18 July 2013 (UTC)
- Way too slow, even on a fast computer on a fast Internet connection. Large articles are almost uneditable. -- The Anome (talk) 07:01, 26 July 2013 (UTC)
- It's too slow for large articles.--Sinuhe20 (talk) 19:53, 27 July 2013 (UTC)
- According to the devs the biggest bottleneck is CPU rather than IO - though there is an overhead there. I have a pretty fast connection but a four year old CPU so for me the new editor is several times slower than the classic one. People with new kit should find that the new editor is almost as fast as the classic one. ϢereSpielChequers 20:39, 27 July 2013 (UTC)
- The CPU cannot be the bottleneck, it's the slow Javascript technique or maybe it's a programming fault. A 100-page Word document can be edited much faster, even on a very old PC.--Sinuhe20 (talk) 09:37, 28 July 2013 (UTC)
- WereSpielChequers is right: speed at the moment is largely a function of how fast your computer is. They are fixing this, but it requires re-writing a lot of the Mediawiki code. If memory serves (and it may not, of course), they are going to have to re-write both Vector and Monobook skins in the process. Whatamidoing (WMF) (talk) 21:15, 31 July 2013 (UTC)
- The CPU cannot be the bottleneck, it's the slow Javascript technique or maybe it's a programming fault. A 100-page Word document can be edited much faster, even on a very old PC.--Sinuhe20 (talk) 09:37, 28 July 2013 (UTC)
- Yep I'm in the same boat on this one. Whether VE is better or not isn't the issue. It'll surely detract more editors having a slow interface than an inefficient one. ErdoS (talk) —Preceding undated comment added 15:52, 1 August 2013 (UTC)
Disagree
- My Internet connection is fast, so the speed isn't an issue. -- Taku (talk) 20:08, 16 July 2013 (UTC)
- It is not slow for me. I might use it for minor corrections on small articles, where scrolling does not take too much time. It is just one of possible instruments. Otherwise, it is a kindergarten. --194.44.219.225 (talk) 12:00, 17 July 2013 (UTC)
- It sometimes is slower than the wikitext editor, but for the most part about the same. I just have an inconsistent internet connection.--¿3family6 contribs 17:49, 17 July 2013 (UTC)
- Hasn't been an issue for me. WaggersTALK 11:46, 18 July 2013 (UTC)
- I have been able to load it just fine... until I disabled it for being a buggy POS with a reaction from the developers I would expect more from EA than WMF. 120.145.218.98 (talk) 02:42, 26 July 2013 (UTC)
- I'm not using it because I prefer editing source text to WYSIWYG. That's the only reason. —Tom Morris (talk) 11:46, 2 August 2013 (UTC)
Comments
- Well, I do use it, but it takes an average of 4x longer to load, even on my "fast CPU" computer, and I regularly can't load anything bigger than 70kb on my least fast computer. I think the resource debt is just too high right now. I do not believe that this is only a user-end issue, because the same users complaining about the slowness with this do not have the same experience with source code. Risker (talk) 21:28, 31 July 2013 (UTC)
No, not only for three browsers
The article says:
- VisualEditor currently works only in the most modern versions of Firefox, Chrome and Safari.
Not true. I've just now used it (two test edits to Austin Maestro) in Iceweasel 20 (which I presume is a rebranded version of Firefox 20, and if so is not the latest version of Firefox), and Chromium (version 28.0.1500.71). -- Hoary (talk) 08:22, 6 August 2013 (UTC)
- I have also used it (happily!) with the Android in-built browser, but right now those mentioned are the browsers that are fully supported, others might still have issues, we hope to get more info about this soon. Thanks, --Elitre (WMF) (talk) 15:07, 10 August 2013 (UTC)
What it's disabled for
We're told:
- Attention Internet Explorer (IE) users: At the moment, VisualEditor is disabled for all IE web browsers.
It's also disabled for rekonq (which is a bit of a surprise). -- Hoary (talk) 06:44, 7 August 2013 (UTC)
- There's a list of supported browsers in the FAQ, if it is not there, then we can't confirm VE works with that browser, but we do hope to have more details on this subject sometime soon. Thanks, --Elitre (WMF) (talk) 15:03, 10 August 2013 (UTC)
Trade-offs in speed improvements
The software team for VisualEditor is asking for community input on a question about improving VisualEditor. One of the main priorities for improving VisualEditor from a technical standpoint is improving the speed: Despite recent improvements, it's slow, especially during busy times and on long articles. The devs are planning to invest several more months in improving the speed. When the work is done, they expect VisualEditor's speed to improve even more than it already has.
After this work is completed, the expectation is that VisualEditor will be much faster in all areas than it is now. However, in a few cases, optimization will require them to make a choice about which direction to optimize the program: is it better to be somewhat faster "here" or "there", because making both faster isn't possible?
The devs working on VisualEditor have specifically requested information from the editor community about which type of speed-related improvement is most important to you when you're editing in VisualEditor. They expect that most of these trade-off choices will involve a choice between two general categories:
- Speed of operation: How quickly VisualEditor works after it has opened the page. Choosing to favor speed of operation means that you might wait an extra second or two before you can start editing a page, although once you've started, it's quicker and more responsive than it otherwise would be (e.g., for scrolling or editing). This might be best for people who make large or complex changes to pages.
- Speed of starting: How quickly VisualEditor opens the page. Choosing to favor speed of download means that you will wait less to get started editing a page, although once you've started, you might occasionally need to wait briefly for VisualEditor to catch up with you (e.g., for scrolling or editing). This preference has a lower bandwidth requirement for starting to edit, and it might be best for people who make tiny changes, such as typo fixing.
Remember that, if it is possible to make both faster, then the devs will do both. This question is only about when one or the other could be made faster, but not both. Which category of improvement do you believe is most important for VisualEditor? Whatamidoing (WMF) (talk) 17:14, 7 August 2013 (UTC)
- I'd go for the speed of starting. I just went to the Barack Obama article to check the relative speeds, and normal editing seemed acceptably responsive at current speeds, although perhaps a little sluggish at times. The speed of starting, though, was very slow, and Firefox even came up with one of those "this script is not responding" messages halfway through loading. (Part of this is due to the fact that I really need a new computer, though.) — Mr. Stradivarius ♪ talk ♪ 17:44, 7 August 2013 (UTC)
- IMO There are a couple of things we could do but I agree speed of loading would be good. I also think we need to expedite the ability to edit sections. If the app didn't try and load the entire page when editing one section, that would be lovely. I think it would also be useful too look into creating an offline toolbar to do somethings. It would have to be optional to do that but it could free up VE from doing somethings and allow it to load faster. Another thing I think would be benefitial would be to modularize some of the code. Some may be able to be moved into LUA modules or some other thing and shared with other apps. For example, AWB has separate pages for fixing typo's and other things. These could be optin for VE and likelwise creating submodules of VE for some of its custom edits would allow some of us with technical experience to help the developers expand and fix some of the functionality. Not all of the coding should be open to us IMO but there are parts that would benefit extra eyes and hands willing to help. Kumioko (talk) 18:34, 7 August 2013 (UTC)
Speed of starting, but both opening up editing and saving edits takes way too long in VE. I just added 2 wikilinks to the Barack Obama article, an article I have edited before. One wikilink with VE, and one with the wikitext source editor. It took 19 seconds for the VE section edit to start up and open up for editing. It took 81 seconds for VE to save my edit. It took 4 seconds for the wikitext source editor to start the section edit, and 30 seconds to save my edit.
Editor | Time to open up editing (seconds) |
Time to save edit (seconds) |
---|---|---|
Visual editor | 19 | 81 |
Wikitext source editor | 4 | 30 |
See bugzilla:48429, "VisualEditor: Support editing of sections inside a page". Tim Starling made a comment there a couple days ago. That bug is about implementing true section editing with VE. Not the fake section editing VE currently does. True section editing by VE is the only real solution for both problems discussed here. No matter what it costs it needs to be done, and soon. --Timeshifter (talk) 19:04, 7 August 2013 (UTC)
I think that in general, optimising the speed while editing is preferable, but we need to set an acceptable performance target. Agree on a target system representing a relatively minimal system and set a limit of 30 seconds for open and 90 seconds for save on some agreed upon set of large articles. So long as the open and close performance meets the target, optimise for speed while editing. It's things like sticking cursors and unresponsive clicks that cause the most user frustration.—Kww(talk) 19:25, 7 August 2013 (UTC)
- I would say that the tolerable amount of delay varies by activity, and that overall we would want the editor to appear snappy when being used. Activities that are perceived to be simple, including basic typing, scrolling the page, and clicking on the pane to move the cursor or select basic elements, should always appear essentially instantaneous to the user. Whatever one needs to do, those basic editing tasks need to be fast or people will quickly get frustrated. For tasks of intermediate complexity, such as opening dialog boxes, a short delay (e.g. 2 seconds) would be okay. I'm generally not a fan of animation in user interfaces, but if populating a dialog box (or similar function) is often going to take a couple seconds, then masking that delay with a quick and subtle appearance effect would be acceptable. In my opinion, the only time that it is acceptable to have long delays (e.g. more than about 5 seconds) is at well-defined entry and exit points, i.e. page loading and saving. So, to that extent I favor optimizing for execution. That said, I consider delays beyond about 15 seconds at the start or end of an edit to be basically unreasonable (and yes that includes the slow saves in the current source editor). In addition, anything that triggers a "slow script" warning message, should be considered completely unacceptable. Personally, I would say that a noticeable one-time delay when moving between sections would be a tolerable trade-off if it means that editing starts more rapidly. In other words, preparing the editable content a section at a time is probably a good idea. Page saves for long pages are slow in the source editor because Mediawiki has to reparse the entire page and recreate all the link table entries and associated minutiae. A truly successful visual editor should avoid most of that by restricting updates to only those sections of the page that have actually changed and by providing an immediate answer to what the parsed content looks like. I realize it is a tall order, but ultimately I would expect to see that saves from VE become faster than saves generated by the source editor, though we are very far from that right now. Dragons flight (talk) 23:17, 7 August 2013 (UTC)
- I will note that even in the plain wikitext editor, saving a page can be very slow indeed - because the page doesn't come back as "saved" until it's been rendered at least once, and that rendering can take 30 seconds or more, even if you're just saving one section. (Example: OpenOffice, which has a couple hundred references, mostly using {{cite web}}.) So slowness on saving a complicated page is something we're already dealing with - David Gerard (talk) 00:54, 8 August 2013 (UTC)
The data from this experiment may be helpful for this discussion. --Cryptic C62 · Talk 01:57, 8 August 2013 (UTC)
As most of my wiki work involves making small changes (typo fixes and the like), speed of starting is more important to me than speed of operation. Thryduulf (talk) 10:42, 9 August 2013 (UTC)
- To me, working from a medium-to-high end machine, I find the editing speed currently tolerable, and the starting speed not. So I have the feeling I would most benefit from better starting speed. That said, to accomplish that, the editing speed shouldn't degrade much. Would that happen, it might just cross over to intollerable, and I wouldn't have any incentive to use VE at all anymore. I also noticed that there seems to be a tremendous difference in warm starting and cold starting VE, whether or not it has been cached. When people are testing for performance, they should acknowledge the difference, and note that both are important. Martijn Hoekstra (talk) 10:50, 9 August 2013 (UTC)
Start speed is important for me. I click edit because i want to edit, not because I want to wait. If I then need to wait for a particular piece to load at a later time, then that is no problem for me. BUT feedback is the most important in any wait step. Where possible try to see about doing actual progress bar instead of 'busy' indicators. It doesn't have to be highly accurate, but you need to be able to assess: ah this is going to take about 4 times the time that I waited so far. —TheDJ (talk • contribs) 12:22, 9 August 2013 (UTC)
Infoboxes
What is it with visual editor and infoboxes? Every time I see an IP edit a page using visual editor the infobox gets deleted, even though they're editing a section.—Ryulong (琉竜) 07:13, 12 August 2013 (UTC)
- Obviously they're doing it wrong, and this is not a severe problem with anything to do with VE - David Gerard (talk) 10:58, 12 August 2013 (UTC)
- If somehow innocuous edits to a page suddenly delete a 2000 character infobox I think that's a problem that needs to be investigated.—Ryulong (琉竜) 11:57, 12 August 2013 (UTC)
- I think the VisualEditor ate a <sarcasm/> tag in David's comment. :-P There was notice by one developer that the bug is supposed to be solved, and it will be available in a new release soon. Diego Moya (talk) 12:31, 12 August 2013 (UTC)
- If somehow innocuous edits to a page suddenly delete a 2000 character infobox I think that's a problem that needs to be investigated.—Ryulong (琉竜) 11:57, 12 August 2013 (UTC)
Visual Editor is buggy in more ways then one. When I see the tag, I expect trouble to be involved, and usually not intended by the user. Example. I really hope there get fixes to prevent accidental content removal, whether it be section titles, infoboxes, or other content.(Hiya Ryulong, I swear I'm not stalking you, I legit just came here to complain about the editor) Blake (Talk·Edits) 03:35, 20 August 2013 (UTC)
- It's bug 51043, marked "critical". I don't think it's been fixed yet.
- Sometimes, the problem might actually be that people are intentionally-but-not deleting the infobox. When you open a page with an infobox, it often looks like there's a stray blank line at the top. That's actually the "blank" line that contains the infobox. So you think that you're just removing a stray line, and you're actually deleting the infobox. There are presumably other problems, too, but that one is known to happen and easily reproducible. One idea for solving it is to add a confirmation dialog ("Do you really want to delete Infobox Foo?") or to require pressing backspace twice, but I don't know what will happen in the end. Whatamidoing (WMF) (talk) 16:54, 20 August 2013 (UTC)
- On a related point, T54488 is supposed to be fixed, but I'm not sure that it's arrived here (insert devs waving their hands and invoking the word "cache" repeatedly). There's supposed to be an update in ~5 hours that should make that one live. Whatamidoing (WMF) (talk) 17:48, 20 August 2013 (UTC)
Next update is next Tuesday
The long-awaited post-Wikimania update has been deployed at Mediawiki and is scheduled for general release next Tuesday afternoon PDT (Tuesday evening for all those in Europe). mw:VisualEditor/status#2013-08-15_.28MW_1.22wmf13.29 has a list of the most interesting bug fixes. Whatamidoing (WMF) (talk) 21:51, 16 August 2013 (UTC)
Download the source?
Is the server side source of the VisualEditor available for download somewhere? I'd like to use the Visual Editor in my local wiki. A git repository would be ideal. But other formats would be fine too.-----<)kmk(>--- (talk) 14:10, 11 August 2013 (UTC)
- I'm not sure if you can get is separate from the rest of the MediaWiki software, but mw:Download is the place to start. There are instructions for git there.--Salix (talk): 15:26, 11 August 2013 (UTC)
- You can download VisualEditor - the front end component and Parsoid, the backend parser.Ckoerner (talk) 16:50, 28 August 2013 (UTC)
Loading slowness, etc.
The Slow entry re VE limitations in this project page needs work. Loading time is impacted as much or more by available effective connectivity bandwidth as it is by the speed of the editor's computer. I just timed the loading time via to view the Philippines article (not the time to load it for editing with VE) via a residential a DSL connection at my current location in Bacoor, Cavite at just over one minute. I haven't been able to measure the loading time for editing by VE; I gave up on that after several minutes struggling with complaints from my browser that the page had become unresponsive.
Perhaps this has implications which should be considered by VE developers and by those managing the deployment of VE. I would suggest that testing of VE releases for deployment suitability include testing over connections with bandwidth-limited connectivity. Wtmitchell (talk) (earlier Boracay Bill) 01:44, 2 September 2013 (UTC)
- According to the devs, the main bottleneck really is CPU time. Obviously, if you've got a limited internet connection, then you're going to have both problems (and it's possible that they multiply, rather than merely adding to each other), but that doesn't mean that CPU time is less important.
- For reference, loading that article (I'm on a medium-quality DSL line in California, with what I assume is a medium-quality computer) takes me about 20 seconds (the first time), and opening it into VisualEditor takes almost 30 seconds. Opening it in the classic wikitext editor requires only about 8 seconds. Whatamidoing (WMF) (talk) 05:42, 3 September 2013 (UTC)
- I'm presently located in Bacoor, Cavite, Philippines. Most of my editing is done from either Romblon, Romblon or Boracay island, both of which have generally slower connectivity than I have from this location. I just surfed over to the Philippines article, and it took 1m45s to load completly. Once loaded, it took 1m44s to open Edit Source in a new tab, and 2m09s to open Edit (VE) in a new tab. Once I've got the article open in VE (this time -- one of the few times I've struggled with it), clicking in the article text produces no visible indication of cursor position, enough up/down key operation to require scrolling produces no reaction, pgup/pgdn produces no reaction, my browser eventually complains that the page is not responding and offers to kill it. This is with a Win7 Home Premium OS with nothing much else going on, running on an off-brand netbook with an Intel(R) Atom(TM) CPU N455 @ 1.66GHz and 2GB of RAM, and using a current or near-current Google Chrome browser. I'm guessing that keyboard operations with VE might require byte-by-byte user/server interaction, and that VE is not dealing well with the longer-than-expected client/server data interchange delays, but that's just a guess. If my guess is anywhere near correct, I suggest that VE developers and deployers do some testing using test environments having artificially-delayed connectivity to simulate usage by editors not near internet backbone gateways. Wtmitchell (talk) (earlier Boracay Bill) 06:22, 4 September 2013 (UTC)
- "I suggest that VE developers and deployers do some testing using test environments having artificially-delayed connectivity to simulate usage by editors not near internet backbone gateways." This has been asked a number of times - is anything to this effect actually done by the VE developers? Is it at the performance optimisation stage? - David Gerard (talk) 06:42, 4 September 2013 (UTC)
- Wtmitchell, I opened a page in VisualEditor and then completely disconnected from the Internet. I could still scroll, type, and add links (although not with the search feature). So I kind of doubt that there is any ongoing communication necessary for at least simpler editing operations. I can ask, though. Whatamidoing (WMF) (talk) 21:13, 4 September 2013 (UTC)
- "I suggest that VE developers and deployers do some testing using test environments having artificially-delayed connectivity to simulate usage by editors not near internet backbone gateways." This has been asked a number of times - is anything to this effect actually done by the VE developers? Is it at the performance optimisation stage? - David Gerard (talk) 06:42, 4 September 2013 (UTC)
- I'm presently located in Bacoor, Cavite, Philippines. Most of my editing is done from either Romblon, Romblon or Boracay island, both of which have generally slower connectivity than I have from this location. I just surfed over to the Philippines article, and it took 1m45s to load completly. Once loaded, it took 1m44s to open Edit Source in a new tab, and 2m09s to open Edit (VE) in a new tab. Once I've got the article open in VE (this time -- one of the few times I've struggled with it), clicking in the article text produces no visible indication of cursor position, enough up/down key operation to require scrolling produces no reaction, pgup/pgdn produces no reaction, my browser eventually complains that the page is not responding and offers to kill it. This is with a Win7 Home Premium OS with nothing much else going on, running on an off-brand netbook with an Intel(R) Atom(TM) CPU N455 @ 1.66GHz and 2GB of RAM, and using a current or near-current Google Chrome browser. I'm guessing that keyboard operations with VE might require byte-by-byte user/server interaction, and that VE is not dealing well with the longer-than-expected client/server data interchange delays, but that's just a guess. If my guess is anywhere near correct, I suggest that VE developers and deployers do some testing using test environments having artificially-delayed connectivity to simulate usage by editors not near internet backbone gateways. Wtmitchell (talk) (earlier Boracay Bill) 06:22, 4 September 2013 (UTC)
Opt-in or opt-out? Confused
Page reads:
After a huge consensus in favor of an opt-out, the WMF developers have implemented a user preference to disable VisualEditor. Go to Special:Preferences#mw-prefsection-editing, and check the box labeled "Temporarily disable VisualEditor while it is in beta", then click "Save".
However, I understand that the RFC concluded with opt-in, not opt-out. Suggest changing the page to:
After a huge consensus in favor of an opt-in, the WMF developers have implemented a user preference to opt-out from VisualEditor. Go to Special:Preferences#mw-prefsection-editing, and check the box labeled "Temporarily disable VisualEditor while it is in beta", then click "Save".
Or am I just missing something here.. --hydrox (talk) 15:44, 3 September 2013 (UTC)
- Suggestion. Consensus is determined from policy-based opinions. Since the policy is to provide an opt-out, opinions not based on this policy are discounted. Of the policy-based opinions, the huge consensus was to support an opt-out. Thincat (talk) 16:56, 3 September 2013 (UTC)
- Thincat, I have no idea what you referring to. No RFC has ever resulted in a decision that opt-out was acceptable. It's opt-in solely because it is within the WMF's power to do what it wants without regard to consensus.—Kww(talk) 18:27, 3 September 2013 (UTC)
- The discussion at VPT that was linked in that sentence was a request for an opt-out. VisualEditor is currently "opt-out" on the English Wikipedia. Whatamidoing (WMF) (talk) 18:32, 3 September 2013 (UTC)
- That discussion did not compare opt-in vs opt-out. There's a clear consensus that it should be opt-in. I agree that it is currently opt-out.—Kww(talk) 19:23, 3 September 2013 (UTC)
- You're talking about two completely different discussions. The statement there about the first discussion (the "huge consensus") is misleading, because that discussion probably had little or no effect on the decision. (I'm honestly not sure that the people making that decision were even aware of the discussion linked there.) I'll go change it to a simple statement of the facts: A user preference was implemented. Whatamidoing (WMF) (talk) 17:19, 3 September 2013 (UTC)
- Both statements are correct and the following is not going to be an unbiased statement because I am so fed up with this application and the WMF at this point. There was an RFC, where there was an overwhelming consensus that VE should be opt in for now, not opt out as it is currently. However, while that RFC was ongoing and presumably because the WMF could see the writing on the wall they jumped in and said they could do this opt out solution, which by the way they had initially opposed. So, in fact, both are true. There was a consensus at VPT to make it opt out to give those of us who wanted to be rid of it a way to get rid of it (but its really not gone, just hidden) and the RFC had a consensus that it should be opt in only because of all the problems and missing functionality. That RFC has yet to be implemented because it needs a closer. With that said, the WMF aslo needs to follow it, which frankly they have no obligation nor the desire to do. So, regardless of the outcome of the RFC, we are left with the Opt out function and a daily mess that needs to be cleaned up from all the problems left in the wake of VE. Its been broken since they released it nearly 2 months ago and even with all the fixes its still months if not years from being a net benefit. Kumioko (talk) 18:46, 3 September 2013 (UTC)
- Again, you're confusing completely unrelated discussions.
- The "huge consensus" is a discussion at VPT started by Doc James on July 1. The title is "Opt out" of VE needed under preferences. It was not a request to make VisualEditor opt-in. (It was also not an RFC.)
- The prefs option to disable VisualEditor during the beta was added on July 24. (As I understand it, the prefs option really does remove it. It's Matma's gadget that only hides it.)
- The "overwhelming consensus that VE should be opt in for now" is your interpretation of the RFC started by Kww on July 30th, which seems to be in the process of being closed now.
- It seems to me that actions taken on July 24 are not likely to have been influenced by an RFC that started on July 30. I believe that when the RFC is closed and any resulting requests for changes are filed, that the WMF staff (mostly James F) will consider the best response at that time. Whatamidoing (WMF) (talk) 20:18, 3 September 2013 (UTC)
- Again, you're confusing completely unrelated discussions.
- Both statements are correct and the following is not going to be an unbiased statement because I am so fed up with this application and the WMF at this point. There was an RFC, where there was an overwhelming consensus that VE should be opt in for now, not opt out as it is currently. However, while that RFC was ongoing and presumably because the WMF could see the writing on the wall they jumped in and said they could do this opt out solution, which by the way they had initially opposed. So, in fact, both are true. There was a consensus at VPT to make it opt out to give those of us who wanted to be rid of it a way to get rid of it (but its really not gone, just hidden) and the RFC had a consensus that it should be opt in only because of all the problems and missing functionality. That RFC has yet to be implemented because it needs a closer. With that said, the WMF aslo needs to follow it, which frankly they have no obligation nor the desire to do. So, regardless of the outcome of the RFC, we are left with the Opt out function and a daily mess that needs to be cleaned up from all the problems left in the wake of VE. Its been broken since they released it nearly 2 months ago and even with all the fixes its still months if not years from being a net benefit. Kumioko (talk) 18:46, 3 September 2013 (UTC)
- In all honesty, I somewhat doubt the RFC has any real point anymore. It dealt with a very different, more buggy software. Adam Cuerden (talk) 06:39, 4 September 2013 (UTC)
- I agree that it has, to a significant but not complete extent, been overtaken by events, but I'm not the (non-admin) editor closing it, so my view isn't the one that matters. It's hard to guess whether, if it were started today, we would get the same responses. I wish that the WMF did routine user surveys so we could have a better idea of what editors believe over time. Whatamidoing (WMF) (talk) 21:08, 4 September 2013 (UTC)
- This is exactly why I encouraged people to close the RFC as soon as the consensus was clear. If it takes us 40 days to process an RFC and WMF only listens to things that are brand new and fresh, they will always have an excuse to reject every RFC. VE is still far too buggy to encourage inexperienced editors to play with: that hasn't changed.—Kww(talk) 21:15, 4 September 2013 (UTC)
- I was just going to add that my guess is that most of the participants there would have approximately the same end result, but that the divide might be somewhat less sharp—perhaps more like 60-40 than like 80-20. Did it seem to you that later respondents tended to be somewhat less opposed to VisualEditor than the earliest ones? Whatamidoing (WMF) (talk) 21:19, 4 September 2013 (UTC)
- For very slight values of "somewhat". It seemed like later support was from people supporting the idea of a visual editor that hadn't thought through the impact of encouraging newcomers to use this particular visual editor.—Kww(talk) 21:36, 4 September 2013 (UTC)
- However, VE is now marked as beta, which also reduces the encouragement. Adam Cuerden (talk) 06:58, 6 September 2013 (UTC)
- Is there any way we can do a snap poll? Just a random sample of, say, 100 users who agree to take a survey? Because we could really use a tool like that. Adam Cuerden (talk) 06:59, 6 September 2013 (UTC)
- I think so—something like SurveyMonkey, you mean, or whatever company it is that does their annual editor survey? It would probably require help from the devs to set up, but if they can do it for the editor survey, then they could do it for this. Actually, I wish that they had been doing a little poll for 50 or 100 editors a day for the entire summer: Which editor did you last use (or maybe answer that automagically)? How happy are you with it (happy/neutral/unhappy)? If you wanted to get fancy, you could rotate a few questions, so that nobody gets more than two or three at a time. Whatamidoing (WMF) (talk) 17:21, 6 September 2013 (UTC)
- For very slight values of "somewhat". It seemed like later support was from people supporting the idea of a visual editor that hadn't thought through the impact of encouraging newcomers to use this particular visual editor.—Kww(talk) 21:36, 4 September 2013 (UTC)
- I was just going to add that my guess is that most of the participants there would have approximately the same end result, but that the divide might be somewhat less sharp—perhaps more like 60-40 than like 80-20. Did it seem to you that later respondents tended to be somewhat less opposed to VisualEditor than the earliest ones? Whatamidoing (WMF) (talk) 21:19, 4 September 2013 (UTC)
- This is exactly why I encouraged people to close the RFC as soon as the consensus was clear. If it takes us 40 days to process an RFC and WMF only listens to things that are brand new and fresh, they will always have an excuse to reject every RFC. VE is still far too buggy to encourage inexperienced editors to play with: that hasn't changed.—Kww(talk) 21:15, 4 September 2013 (UTC)
- I agree that it has, to a significant but not complete extent, been overtaken by events, but I'm not the (non-admin) editor closing it, so my view isn't the one that matters. It's hard to guess whether, if it were started today, we would get the same responses. I wish that the WMF did routine user surveys so we could have a better idea of what editors believe over time. Whatamidoing (WMF) (talk) 21:08, 4 September 2013 (UTC)
Anyway, don't get me wrong. I do think VE has quite a way to go before it's really newbie friendly. But once it got marked as a beta.
...Huh. Now, there's a question I never thought of. How well-understood is "beta" to the average newbie?
Well... regardless, it also now pops up a box explaining it's beta software on first use. So a lot of effort has been made to identify it as a feature in development. This effort to mark VE as experimental does a hell of a lot to mitigate the problem of showing it to newbies.
In the end, when the RFC ran, we were looking at software being presented as the default, with no warnings. Now, there are several warnings (there's another warning at save), VE is marked as being in development... and, were all this pointed out to the voters, I'm not convinced this would go the same way.
As I said, VE does have quite a way to go. I did change my mind on VE, but the change wasn't to thinking it was perfect; indeed, I think it still has some serious flaws - It can't edit tables, it currently relies too much on form-based editing, and, while it's improving, there's obviously still some bugs. What I changed my mind about was the development team: Now that I understand the limitations, I now think they pull this off, given time. But, getting back to the point, whilst we may not have got exactly what we said we wanted from the RFC, the warnings, in the end, serve substantially the same purpose when it comes to letting newbies know that VE may still be somewhat difficult. Adam Cuerden (talk) 08:37, 6 September 2013 (UTC)
- I think the current state of affairs is as good as it can be at the present. All users have access to both editors, the old editor is still the default, following the principle of least surprise, and the VE is correctly labeled as beta-quality software. If there aren't enough testers for the VE, I suggest running a site banner campaign suggesting users try the VE as testers: perhaps something on the lines of "We need testers to give us feedback on our new experimental visual editor: try the 'edit [beta]' links to try it". When the VE reaches release quality: which I think is probably still at least six months to a year off, we can move to labeling the two editors as "[edit source | visual editor]" instead of [edit source | edit beta]: and, because both editors will continue to be available for all users in a backwards-compatible way there won't be any need for anyone to opt either in or out. Once that's done, and everyone's then been happily using it that way for another six months or so without significant pushback, showing that the VE is finally unquestionably 100% usable and robust, should we then consider changing that to [visual edit | edit source], and then, in due course, to the final state of [edit | edit source]. -- The Anome (talk) 09:28, 6 September 2013 (UTC)
- I have to say that I am pretty disappointed (but not surprised) at the general attitude here that the WMF doesn't care about the consensus formed by the community. Its even more discouraging considering how rarely the community actually comes to a consensus on anything and the fact that never before has 500 of the community voted on anything. I said in the beginning that the RFC was a waste of time but I hoped I was wrong and I hoped the WMF would do the right thing...I guess I was right. All the WMF did was waste the communities time...again, with regards to Visual editor. It would have been better if you would have just said in the beginning that the English community does not have the power to overrule a WMF decision. That would have been better than letting us have an RFC that you had no intentions of following.
- I have agreed from the begging that Visual Editor has potential and it has come a long way. But its still a beta, it still never should have been released to the community in the horrible state it was in and its still got a long way to go before its ready to be available to every newbie and casual editor. It still doesn't work with any version of Internet explorer. That's ridiculous, its one of the most widely used browsers and it doesn't work at all. Anyway, this is exactly why I stopped using it or helping to find problems with it. Because the WMF doesn't care what we have to say, so they don't really need our help. There just going to do what they want to do anyway. All we are is a free lab for the WMF to test their new changes for free before releasing it in the for profit side of the software. Kumioko (talk) 10:44, 6 September 2013 (UTC)
- I'm inclined to assume good faith on the WMF's part here; I think this was purely an artifact of the forces of internal politics within the WMF, and not caused by any sinister or ulterior motive. While the WMF does not have any obligation to consult the community in matters of software, WMF management are not completely unresponsive, and the level of outrage and protest on both the en: and de: wikipedia in this case was enough to make them reconsider its policy. I think the current state of affairs is much more promising than a few weeks ago when the unilateral change was made, and the WMF is now taking a much more measured approach to VE development and deployment, which is good for the VE, good for the userbase, and good for the encyclopedia. -- The Anome (talk) 12:13, 6 September 2013 (UTC)
- I agree that its getting better and that there wasn't anything sinister intended. What it was though was callous anddisrespectful of the community, and frankly, just plain dumb...and that continues. The WMF insisted on releasing software they knew (partially because they knew and several of us told them back in May and even earlier) that it wasn't ready, they released it anyway. Its bad enough that it isn't finished and has flaws, its absolutely unacceptable that it actually damages the articles by breaking things, or removing or inserting things. What's even worse than that, they have the attitude they don't have to fix the problems, they expect the community to fix the problems they created. What's more is that I know they use WP as a lab to test stuff before putting it into the for profit software, most editors don't know that and that would be fine, if they didn't expect us to clean up their mess because they were in a rush. That is what bothers me and that is what causes me to feel the WMF doesn't care about this community. They took a calculated risk that we wouldn't complain and or wouldn't get a consensus to turn it off. They expected it!, and even after the consensus was formed have no intention of complying with the community. I again agree that VE has potential and that some day when its better developed and the bugs have been mitigated we can release it to the world...but we aren't there yet. If the WMF would have been patient and allowed this rollout to be calculated, informed and had actually thought it out a little, we probably could be releasing it soon. Since the WMF decided to release it months before it was ready though its not just the software they need to fix, its their credibility. Kumioko (talk) 14:02, 6 September 2013 (UTC)
- What "for profit" software would this be? As far as I'm aware, the WMF is a 501(c)(3) charitable organization with a quite specific remit, and does not create for-profit software. -- The Anome (talk) 14:08, 6 September 2013 (UTC)
- Yes but the core software that runs Wikipedia and most or all of the other pedias is the same for profit software used for Wikia (the one that runs Wookiepedia and the like) and a variety of other things. The 2 organizations are fundamentality different (which isn't always a good thing BTW) but the core software is the same. Unless Wikia and the others stopped using the Mediawiki software, which I suppose is possible. So changes made here to the core portion generally end up in the other. Kumioko (talk) 14:53, 6 September 2013 (UTC)
- Following that reasoning, we can call any software that doesn't have a non-commercial use clause in its licence (AFAIK none do) for profit software. Linux, Gnu, Git, you name it, it's all being used in for-profit organisations. Calling the software for-profit because it is used by a for-profit organisation seems odd. Martijn Hoekstra (talk) 15:02, 6 September 2013 (UTC)
- In any case, the WMF's release of Visual editor was extremely poor. Even with all the bug fixes and work that's been done to it, the application still isn't ready for prime time release and shouldn't be available to people as the default. Expecially Ip's and new users. Kumioko (talk) 15:08, 6 September 2013 (UTC)
- I don't think anyone's arguing with you on that last point. Fortunately, that's already been addressed, and it's no longer the default. -- The Anome (talk) 15:49, 6 September 2013 (UTC)
- Well that's good at least, but it still should be opt in not opt out until its more trustworthy and that's the consensus of the community not just me. Its just a matter of whether the WMF is going to listen to the 480+ editors who voted for the VE app to be opt-in vice opt-out or just ignore the wishes of the community. Kumioko (talk) 17:09, 6 September 2013 (UTC)
- I still wonder whether the opinion of a couple of hundred people, heavily skewed towards highly experienced editors, actually represents the wishes of the 129,675 registered editors, most of whom are relatively inexperienced and thousands of whom are actually using VisualEditor. (If you just want to talk about mere vote counting, then let's talk about people voting with their feet: The number of people who used VisualEditor during the last month substantially exceeds the number of people opposing it in that RFC.)
- As for your complaint that the WMF was "wasting the community's time", please note that the WMF did not start or encourage the RFC. The WMF does not censor discussions: if the community wants to discuss something like this, or if an individual wants to start an RFC about the software, then the WMF will not prevent the discussion. In fact, the WMF will even spend time reading the responses to find useful feedback and areas for compromise. But you're right: the WMF does not blindly comply with the demands of 0.3% of active registered users merely because the demand was made. Whatamidoing (WMF) (talk) 17:36, 6 September 2013 (UTC)
- Well that's good at least, but it still should be opt in not opt out until its more trustworthy and that's the consensus of the community not just me. Its just a matter of whether the WMF is going to listen to the 480+ editors who voted for the VE app to be opt-in vice opt-out or just ignore the wishes of the community. Kumioko (talk) 17:09, 6 September 2013 (UTC)
- Following that reasoning, we can call any software that doesn't have a non-commercial use clause in its licence (AFAIK none do) for profit software. Linux, Gnu, Git, you name it, it's all being used in for-profit organisations. Calling the software for-profit because it is used by a for-profit organisation seems odd. Martijn Hoekstra (talk) 15:02, 6 September 2013 (UTC)
- Yes but the core software that runs Wikipedia and most or all of the other pedias is the same for profit software used for Wikia (the one that runs Wookiepedia and the like) and a variety of other things. The 2 organizations are fundamentality different (which isn't always a good thing BTW) but the core software is the same. Unless Wikia and the others stopped using the Mediawiki software, which I suppose is possible. So changes made here to the core portion generally end up in the other. Kumioko (talk) 14:53, 6 September 2013 (UTC)
- What "for profit" software would this be? As far as I'm aware, the WMF is a 501(c)(3) charitable organization with a quite specific remit, and does not create for-profit software. -- The Anome (talk) 14:08, 6 September 2013 (UTC)
- I agree that its getting better and that there wasn't anything sinister intended. What it was though was callous anddisrespectful of the community, and frankly, just plain dumb...and that continues. The WMF insisted on releasing software they knew (partially because they knew and several of us told them back in May and even earlier) that it wasn't ready, they released it anyway. Its bad enough that it isn't finished and has flaws, its absolutely unacceptable that it actually damages the articles by breaking things, or removing or inserting things. What's even worse than that, they have the attitude they don't have to fix the problems, they expect the community to fix the problems they created. What's more is that I know they use WP as a lab to test stuff before putting it into the for profit software, most editors don't know that and that would be fine, if they didn't expect us to clean up their mess because they were in a rush. That is what bothers me and that is what causes me to feel the WMF doesn't care about this community. They took a calculated risk that we wouldn't complain and or wouldn't get a consensus to turn it off. They expected it!, and even after the consensus was formed have no intention of complying with the community. I again agree that VE has potential and that some day when its better developed and the bugs have been mitigated we can release it to the world...but we aren't there yet. If the WMF would have been patient and allowed this rollout to be calculated, informed and had actually thought it out a little, we probably could be releasing it soon. Since the WMF decided to release it months before it was ready though its not just the software they need to fix, its their credibility. Kumioko (talk) 14:02, 6 September 2013 (UTC)
- I'm inclined to assume good faith on the WMF's part here; I think this was purely an artifact of the forces of internal politics within the WMF, and not caused by any sinister or ulterior motive. While the WMF does not have any obligation to consult the community in matters of software, WMF management are not completely unresponsive, and the level of outrage and protest on both the en: and de: wikipedia in this case was enough to make them reconsider its policy. I think the current state of affairs is much more promising than a few weeks ago when the unilateral change was made, and the WMF is now taking a much more measured approach to VE development and deployment, which is good for the VE, good for the userbase, and good for the encyclopedia. -- The Anome (talk) 12:13, 6 September 2013 (UTC)
I don't understand: the "edit source" link is back where it should be, in the place that muscle memory expects to finds it when editing: no-one is making you use the other link. Visual editing is thus effectively off by default, except for those editors who consciously opt in by clicking the "edit beta" link. What's the problem? -- The Anome (talk) 17:25, 6 September 2013 (UTC)
- To state it rather plainly, the complaint from some editors is not that they're forced to use it, but that other people are choosing to use VisualEditor. Those other people (generally inexperienced people) do not always check to see whether their edits worked perfectly and correct any perceived problems themselves. Sometimes this produces what we call a dirty diff (a change that looks perfectly fine to the reader, but produces odd changes to the wikitext, like a pointless nowiki tag). Sometimes it produces a visible mess.
- Of course, the same is true for wikitext: inexperienced people make all kinds of mistakes there, and they often don't notice or clean up after themselves. User:BracketBot sends a couple hundred messages a day about improperly paired brackets, because people screw up manual [Link formatting]] all the time. People making links in VisualEditor never produce that mistake. If you're curious, it looks like there are slightly more people producing improperly paired link brackets each day than there are VisualEditor users producing unwanted nowiki tags. Whatamidoing (WMF) (talk) 17:49, 6 September 2013 (UTC)
- First you are making this as though I am the voice of the community in this. I am not, I am just one vocal voice among hundreds. Most of which would rather edit than discuss it which is understandable. So this just goes back to wether the WMF is going to follow the community consensus or show the community they don't care what the volunteers think. Which it sounds like the decision has already been conveyed pretty clearly above. Also, your arguments that .3% of the communuty voted are utterly and completely irrelevent. That is not, nor has it ever been, how we gauge a consensus. Would that work in an RFA (well 129, 000+ people didn't vote oppose, so they really support). I can only imagine the comments that would ensue if that were to happen. Maybe I'll try that when I submit again in a few months and see what they say. How about with AFD, MFD or CFD, etc? It has long been stated that the number of page views doesn't justify the inclusion of content over a consensus. So really, the argument that .3% of the community is all that want it changed is absurd. Also, its irrelevent how it displays for me because I disabled it; because its still a crap app. If that changes then I'll turn it on and start using it. Further I completely ignore any and all errors to content I see if done by Visual Editor. And I have found quite a few even as of this week that haven't been fixed. I will continue to do so as long as the WMF continue to allow broken software to be used that they and everyone else knows is creating errors in content and then not cleaning up the mess. Yes, those errors are reducing, but it still occurs and until those errors are reduced to a nuisance instead of an ongoing pattern, then it should only be opt-in. Further, a beta release should never be a full release. That is counter to every common practice and common sense for software development. I would also add that I would bet that 99% of those vaste majorityof editors don't know that the VE app causes problems and probably wouldn't know if it did. Or if they did notice they wouldn't know who to tell or how. So, my guess would be that a large number of those folks either wouldn't use it if they knew. But its impossible to prove either theory without a survey and that's a waste of time too unless someone is actually willing to use the data objectively. Kumioko (talk) 18:06, 6 September 2013 (UTC)
- Who'd have thunk? I am the 1%. Martijn Hoekstra (talk) 19:29, 6 September 2013 (UTC)
- So am I.
- Kumioko, I don't know what the decision will be. I know that the comments and suggestions are being considered. What I do know is that it is possible to have a big difference between the viewpoints of the people who commented on any given page and the viewpoints of the entire community. The WMF is supposed to pay attention to all the editors, not just the ones who comment on a given page. The English Wikipedia is free to choose the narrower focus of only paying attention to the people who commented in a particular discussion. Whatamidoing (WMF) (talk) 21:00, 6 September 2013 (UTC)
- Who'd have thunk? I am the 1%. Martijn Hoekstra (talk) 19:29, 6 September 2013 (UTC)
- That may be true for some, Whatamidoing (WMF), but I think it misrepresents most of us. I have no problem with a skilled editor choosing to use VE. I object to VE being presented as an option to editors that are too inexperienced to recognise when the edits they have made have caused problems or to recognize that the edit they want to make is impossible in VE. Right now, the use of VE essentially requires editors to check the diff to make certain nothing went wrong and to be willing to revert, re-edit using the source editor, and write a problem report when something does.—Kww(talk) 19:40, 6 September 2013 (UTC)
- First you are making this as though I am the voice of the community in this. I am not, I am just one vocal voice among hundreds. Most of which would rather edit than discuss it which is understandable. So this just goes back to wether the WMF is going to follow the community consensus or show the community they don't care what the volunteers think. Which it sounds like the decision has already been conveyed pretty clearly above. Also, your arguments that .3% of the communuty voted are utterly and completely irrelevent. That is not, nor has it ever been, how we gauge a consensus. Would that work in an RFA (well 129, 000+ people didn't vote oppose, so they really support). I can only imagine the comments that would ensue if that were to happen. Maybe I'll try that when I submit again in a few months and see what they say. How about with AFD, MFD or CFD, etc? It has long been stated that the number of page views doesn't justify the inclusion of content over a consensus. So really, the argument that .3% of the community is all that want it changed is absurd. Also, its irrelevent how it displays for me because I disabled it; because its still a crap app. If that changes then I'll turn it on and start using it. Further I completely ignore any and all errors to content I see if done by Visual Editor. And I have found quite a few even as of this week that haven't been fixed. I will continue to do so as long as the WMF continue to allow broken software to be used that they and everyone else knows is creating errors in content and then not cleaning up the mess. Yes, those errors are reducing, but it still occurs and until those errors are reduced to a nuisance instead of an ongoing pattern, then it should only be opt-in. Further, a beta release should never be a full release. That is counter to every common practice and common sense for software development. I would also add that I would bet that 99% of those vaste majorityof editors don't know that the VE app causes problems and probably wouldn't know if it did. Or if they did notice they wouldn't know who to tell or how. So, my guess would be that a large number of those folks either wouldn't use it if they knew. But its impossible to prove either theory without a survey and that's a waste of time too unless someone is actually willing to use the data objectively. Kumioko (talk) 18:06, 6 September 2013 (UTC)