Talk:Talk pages consultation 2019/Phase 1 report

Please find your name in this report

Hello, Akela NDE, Alexis Eco, AlvaroMolina, Andrzej19, AntonyB, Barcelona, Chnutz, Ciell, Charis, Designer1959, Dimaniznik, Edgouno, Enyavar, Fanaliceful, Floflo, Flugaal, Frigory, Geonuch, Grasyop, Gżdacz, H7, Jaluj, Jamain, Jckowal, Jeblad, Jpgibert, Jules78120, Lamiot, Liquendatalab, Louis H. G., MarioFinale, Matthiasb, Nattes à chat, Neitram, Niridya, O. Morand, Oleg3280, Oscar ., Pamputt, PaulSch, Reem Al-Kashif, Retired electrician, RonnieV, 百無一用是書生, Sumek101, Thieu1972, Tortliena, Tximit, Vriullop, Wargo, Wiesios, Wikilover90, Yellow Horror, Yodaspirine, Zellmer, そらたこ, リボンちゃん, and 無聊龍:

You have been quoted in this report. Please find (Ctrl+F or ⌘ Command+F) your name in the report. Some people were quoted more than once. Please tell me if the translation is not good. Whatamidoing (WMF) (talk) 20:12, 17 May 2019 (UTC)

Also Waldir, (from Flow/Research/Experienced User Responses) and Kippenvlees1, Prométhée, and Tom (LT) (from the 2019 Community Wishlist Survey on Meta): Your comments, from previous relevant discussions, are quoted in this report. I'm pinging you in case that might be surprising. Whatamidoing (WMF) (talk) 20:22, 17 May 2019 (UTC)

• Fixed minor errors. Feel free to edit if necessary. Jeblad (talk) 20:46, 17 May 2019 (UTC)
• Same. Best regards, Jules78120 (talk) 11:02, 18 May 2019 (UTC)
• another error (due to myself only) : mon avis sur la seconde phase a été mis dans les avis sur la première phase ; je consultais les avis déjà donnés et j'ai stupidement rajouté le mien. Cela prouve tout de même que les pages de discussion ne sont pas assez claires.
  • I thought about keeping a list of errors that I encountered during this process. Offhand, here's what I remember: Multiple editors forgot to sign at least one comment (one of which was a joke, I think, but the others were accidents). The use of non-matching signatures (e.g., "User:Example" for the real username vs "Example!" for the label in the signature) created repeated problems in writing this report. We had at least three quotations that had the wrong person's username in the notes. At least once, two consecutive comments, from different people, were initially counted as being from one person because it wasn't obvious where the first comment stopped and the second started. As we all know, there's no automatic way to keep track of discussions on 20 wikis, except by visiting those 20 wikis separately. There were probably more, but that's what I remember right now. Basically, I'd say that writing this report proved all of editors' complaints correct. Whatamidoing (WMF) (talk) 18:10, 24 May 2019 (UTC)

@Whatamidoing (WMF): Notifications work for 50 people at most. Here there are 57 users, so no notifications were sent. You must do two edits to notify us.

• First batch: Akela NDE, Alexis Eco, AlvaroMolina, Andrzej19, AntonyB, Barcelona, Chnutz, Ciell, Designer1959, Dimaniznik, Edgouno, Enyavar, Fanaliceful, Floflo, Flugaal, Frigory, Geonuch, Grasyop, Gżdacz, H7, Jaluj, Jamain, Jckowal, Jeblad, Jpgibert, Jules78120, Lamiot, Liquendatalab, Louis H. G., MarioFinale, Matthiasb, Nattes à chat, Neitram, Niridya, O. Morand.
• Second batch: Oleg3280, Oscar ., Pamputt, PaulSch, Reem Al-Kashif, Retired electrician, RonnieV, 百無一用是書生, Sumek101, Thieu1972, Tortliena, Tximit, Vriullop, Wargo, Wiesios, Wikilover90, Yellow Horror, Yodaspirine, Zellmer, そらたこ, リボンちゃん, and 無聊龍.

Frigory (talk) 14:31, 7 June 2019 (UTC)

@Frigory: I would note that based on the revision history almost all of the editors should have received a notification, as they were added in groups of 10 (although for some of the diffs I can't tell if any notifications would have been sent). Jc86035 (talk) 16:16, 7 June 2019 (UTC)

@Jc86035: For my part, I don’t remember I received any notification. Are you talking about the names of people joined to the quotations in the report? This won’t send notifications — see the conditions here. Oh… Actually I forgot to sign the first time, so it did probably not work either… Frigory (talk) 16:22, 7 June 2019 (UTC)

@Frigory: Yes, I was referring to the userpage links that Whatamidoing added in the first few revisions of this page. Jc86035 (talk) 16:23, 7 June 2019 (UTC)

Also, I don't know if Whatamidoing had notifications for successful mentions turned on. If she did then she would have known whether the notifications were sent correctly. Jc86035 (talk) 16:24, 7 June 2019 (UTC)

Oh, indeed, she put the usernames in batches… Sorry, I was stupid, I should have checked — . I won’t miss it next time.

Yet I don’t have the expected notification in my notif history. Too bad. Frigory (talk) 16:34, 7 June 2019 (UTC)

The translation of my comment seems good to me; I've also taken a quick look on other French comments, and the translation seemed more than decent there too. (N.B. : Gosh this topic is cluttered with usernames XD.).--Tortliena (talk) 16:20, 7 June 2019 (UTC)

I have made a quick look on all Chinese quotes, IMO, the translations are quite decent. --無聊龍 (talk) 03:29, 8 June 2019 (UTC)

I corrected translation of my comment. Best regards. -- Wiesios (talk) 10:13, 24 June 2019 (UTC)

Some thoughts

For the most part, I'm rather indifferent at this stage as it's high-level planning rather than specifics. But I do have some thoughts.

• I see several mentions of "new buttons" for accomplishing things, but no specifics of how those buttons would work.
  • If you're creating new special pages (like Special:MovePage) or index.php actions (like action=edit and action=history) for manipulating "threads" and such, keep in mind that your MVP must include ways to accomplish everything via the Action API as well.
  • If you're creating your new features using JavaScript, I'd guess that'll already be using the Action API. But remember that many of our users use browser extensions like NoScript or otherwise disable JavaScript and would like to be able to continue using talk pages. It's ok if the experience isn't as convenient (e.g. they may have to manually type some wikitext and remember ~~~~ on their own), but they should still be able to accomplish everything.
• As much as possible, make sure that someone (or some bot!) who uses the old syntax doesn't break everything. For example, if you introduce some new syntax for headings and someone adds a section using the old syntax. At the same time, forbidding the old syntax would likely be counterproductive.
  • Remember that major changes to the syntax go against #Stability. Switching from colons/asterisks for indentation to some new character would be less disruptive than wrapping every comment in XML-ish tags. Normal "==" headings would be less disruptive than some new syntax for headings, even "=/="; on every page I'm aware of, "a discussion" is almost always indicated by the same level of heading, so all you might really need is some way to know that e.g. each separate discussion on w:en:Wikipedia:Templates for discussion/Log/2019 May 16 is the "====" level rather than "==".
• At least on enwiki, people do sometimes use the Wikipedia_talk pages of Wikipedia-namespace discussion boards to separate discussion about the board itself from whatever topic is supposed to be being discussed on the board. While in my experience the meta-discussion tends to be infrequent, expect pushback if you try to force all boards to move to Wikipedia_talk. The wikitext magic word (like the existing __NEWSECTIONLINK__ magic word that adds the "add topic" tab) strikes me as more likely to be accepted.
• Regarding per-page history versus per-discussion history, I personally would prefer to keep the per-page history as that's how I keep up with discussion boards.
• Regarding moving the top-of-page templates, I'd expect pushback on that one along the lines of "if you hide them somewhere else, no one will ever see the important information!". Likely no matter how objectively useless the information actually is to a newbie trying to use the talk page.

HTH. Anomie (talk) 20:51, 17 May 2019 (UTC)

Thank you for writing this down! Nemo 08:02, 17 June 2019 (UTC)

Once upon a time…

I was involved in a project to replace a really ancient and crappy system. We were trying to rebuild the system, to create something better and more usable. It should even have a graphical user interface! After several year the system emerged. Strangely, it had all the bad interaction gadgets, but now the exact same weird handles were recreated in the graphical user interface. The users still typed the actual numeric action codes. What had happen? We did one major error, we asked expert users what they wanted. Everybody thought they know best how to make an efficient system, but what they did was to recreate a system with all the original flaws because the did know how to work around those flaws.

The lesson learned; don't make an expert group of old users with bad habits. ;) Jeblad (talk) 21:44, 17 May 2019 (UTC)

• Charming. Tuvalkin (talk) 21:50, 27 May 2019 (UTC)
• I know exactly what you mean. El komodos drago (talk) 16:06, 28 May 2019 (UTC)
• Agree. I read most of the text about what is being proposed, but I did not see a differentiation between "What expert Editors might want" and "What new Editors might require", and it's an important distinction to make. Before doing anything, I think the developers need to make an explicitly stated decision on what their primary focus is, and prepare in advance what they are going to do if the interests of these two groups should come into conflict, rather than being "surprised" midway through development and have to make that decision after work has been expended in one direction, and making the decision by a small number of technically-oriented types on an emergency basis, rather than having that decision made in advance by long term policy types. You don't want the programmer to make this decision unilaterally when conflicts arise, because they are going to make their decision based on expediency.Tym Whittier (talk)

I agree w/(talk)73.11.63.20 08:46, 3 June 2019 (UTC)

Maybe there could be an third-party application with a graphical user interface that shows the Wikipedia talk page as if it were a forum with different topics. I don't mind the normal way in Wikipedia, but it might confuse some new users, as has been already stated. User:Chanc20190325 23:52, 28 May 2019‎ (UTC)

Comments by Alsee

I thought phase 2 would include more detail.

1: What do you think of the proposed product direction?

I am greatly relieved that the proposal is to add functionality to what we already have. The specifics will still be key, but the direction sounds very promising.

I would like to add some advice on how to conceptualize the project. It appears the team is still in the mindset of building talk page features. I think it may help if you can think from the viewpoint that "a page is a page". You're not building talk page features, you're building page features. In some ways it narrows down the options that make sense, and in some ways it may make some of the issues vanish.

2: Marking separate discussions

I would split this into two kinds of change. First, a style change such as =/=SECTION=/=. It is unclear to me why that would be needed, but I can see that being viable if it really is needed. The second kind of change is what I'd call "not human writable". Something like software-generated Globally Unique IDs (ge09834h534uo3g) is obviously something a human can't click edit and write. That raises a much more serious hurdle. I would want to carefully examine what's possible without that. I think there may be some flexibility in "expected functionality" that may help avoid that kind of approach.

3: Helping newcomers find the talk pages

I suspect that the Vector skin is part of the problem here. It makes Talk and other links-for-editors literally fade into the background. That aside, increasing the visibility of Talk pages for new users sounds fine. It's unclear what "discussion functionality connected to individual sections" would look like, so I have no view on that.

4: Where to show discussion tools

It's not completely clear yet what these would look like, but if you can follow the model that "a page is a page", that may solve itself.

One of them is to move all discussions to a talk namespace.

That's rather awkward. Sometimes the primary workflow-page is a discussion page, and we need a Talk page behind it for discussions about that workflow. For example Village Pump, requests for promotion to administrator (RFA), and deletion-discussion pages (AFD). This option would leave us with a useless/not-created non-talk page, and it would leave us nowhere to have discussions about workflow happening on the Talk page.

In a sense, the real problem here is that this violates the model that a page is a page. It tries to create types of pages, with technical differences between them. That breaks the "pageness" of a page. I'm sorry of that comes across as too zen or too abstract, but part of the simplicity and flexibility of a wiki is that a page is just a generic page.

5: History tradeoffs

I'd be willing to consider ideas here, but like you, I don't see how this would work. There have been very rare times when it would have been convenient if an archive page gave more direct access to a discussion edit history, but overall I'm not seeing much opportunity for a net positive. I'm pretty sure we need to retain the full history on the original page, and trying to duplicate the relevant entries onto the archive history would almost surely be a disaster.

6: Metadata location

Some templates really do need to be at the top of the page, and some are clutter that I'd gladly see moved out of sight. Some are a difficult call that would involve significant internal community debate&grumbling. I think for simplicity we can just call it "half" that could be moved. I don't think the Foundation should even ask which ones. That would be handled by the community.

Currently all pages come in pairs (a page and a linked Talk page). I was stunned for a moment by the idea that we might convert all pages into triplets, to have a place to put this kind of stuff. That is a bold and unexpected notion. I'm not sure what I think about it. However I would say it is effectively independent of the Talk page discussion. If you really want to consider that, I'd split it off as an essentially separate proposal.

Other than making pages into triplets, I'm not sure where else that stuff would go. However I would like to note that the community already has some of these templates default to collapsed mode, and in some cases we already put multiple templates behind a single collapse. Even without Foundation action, the community could immediately start moving all hideable templates behind a single show/hide collapse if we decide that is worth doing. Alsee (talk) 08:11, 18 May 2019 (UTC)

Comments by Jules

Hi,

1. I really like this orientation: not replace the wikitext discussion by something like Flow, but add a functionnalities that make easier the use of wikitext discussions. A bit like VisualEditor did not replace Wikitext, but came over wikitext interface. I think this orientation is the more likely to make consensus inside communities.
2. I think Wikimedians can adapt to minor changes in the way in which discussions are separated. It won't affect too much flexibility, if it is still possible to merge or split discussions.
3. Great Idea. The only disadvantage I can see is, as a side effect, an increase of vandalisms and/or irrelevant comments (remember the feedback feature?) on talk pages. If talkpages are more visible (for both readers and newcommers), it would probably be needed to think how to inform Wikipedia readers about what is relevant (discussing about article content) and irrelevant (how to contact this personnality, your opinion about this singer, etc. ;-))
4. The best thing would be to encourage to move all talk spaces on 'talk page' namespace, but also allow exceptions with a magic_word or template (as fr:Modèle:Page de discussion on French Wikipedia) which can be added on top of a page to give it talk pages features. This would keep the current flexibility.
5. As an experimented editor, I prefer complete page history, but I can understand that topics histories could be nice, in addition. But these topics histories should be like (look the same) the complete page history. Currently, Flow topics histories are not built in the same way classic histories are; it's a mess ^^. To implement those two histories in parallel, maybe just keep the "History" tab as it is now (complete page history is displayed) and add a "Topic history" tab near each topic title?
6. I'm not sure I understand this question. Metadata information should imho stay on top of talk pages (on article talk pages, there can be advices or warnings when the topic is hot/controversial; on community talk pages like Admin noticeboards, there can be explanations that should be read before publishing; etc.). But maybe could they be identified with markup/tags/template so, with the new discussion overlay (if we can call new features this way?), it would be explicit that it is not a discussion content, not a talk topic.

Sorry for mistakes, my English need improvements ;-). Regards, Jules78120 (talk) 11:32, 18 May 2019 (UTC)

MediaWiki.org is a multi-lingual wiki. You can use any language you're comfortable with.

I believe that User:Yodaspirine has thought about the "metadata" problem. Some notices might be important, but some re-organization might be helpful. Whatamidoing (WMF) (talk) 18:19, 24 May 2019 (UTC)

Well, my point was more focused on notices that are thrown in the middle of a talk page (e. g. [1]) than metadata info on top of pages. Regarding metadata on top of pages, the only thought I have is that there can be a confusion when the templates push the ToC one scroll behind.--Yodaspirine (talk) 13:29, 2 June 2019 (UTC)

Preliminary thoughts

• This is a good summary and I think it's good that individual comments from the discussions have been presented.
• I somewhat doubt that no one defended signatures, considering how many users customize them. I think it could be beneficial to allow signature customization to live on in some way, even if they're removed from the discussion interface (e.g. showing them in popups when hovering over usernames).
• I think it would be a bad idea to give each section of a talk page its own history. This would almost certainly conflict with the desires to view all changes on a page at once and edit all of a page at once. (It also wouldn't take into account messages moved between discussions.) I think – if it would be possible – a better approach would be to automatically add diff links to comments and allow modified level 2 section headers to serve as "real" markers for watchlist notifications, possibly by generating short tags for each section and keeping them in the wikitext. This would also help for section links on wikis like Commons and Wikidata, where section headers appear differently for different users because translation parser functions and templates are used in them. (Perhaps by using an edit-tagging system based on this you could filter relevant edits in page histories, rather than keeping two copies of every thread's history.)
• Using an actual template, à la {{FlowMention }}, could be difficult to manage (why not use a software-defined parser function or magic word?). In general, syntax should probably be kept lightweight and unobtrusive.
• I would support something like a pair of __DISCUSSION__ and __NOTDISCUSSION__ magic words to enable and disable the talk page upgrades.
• As a Timeless user, I think many of the interface problems (e.g. contributors clicking on the wrong "Talk" link) can be attributed to Vector. The important links aren't emphasized enough, and there are so many links that almost no one looks at them unless they're really bored. Even in Timeless, where the "User talk" link is always hidden under a drop-down, I'm not sure whether ten new contributors would all find the "Talk" link.
• As a Wikidata user, I think changing the ‎<h1> to indicate more clearly that a page is a discussion page could be helpful interface-wise. It doesn't have to be exactly the same as the title indicated in the URL.
• I think a mechanism to notify more than 50 editors at once could be helpful. Perhaps transcluding a page where users can only add their own username but not someone else's, and user renames are automatically reflected? Jc86035 (talk) 11:46, 18 May 2019 (UTC)
• Supporting multiple discussion types is definitely necessary, especially if things like deletion discussion votes/expressions-of-opinion could be emphasized through CSS. A Twitter-style inline polling system (separate from SecurePoll) could be useful as well, although it could be difficult to implement this (e.g. a "poll creator" user group might have to be created) and it could fundamentally change how decision-making discussions are held (e.g. is a supermajority necessary for consensus? would ranked-choice voting have to be implemented?). Requiring some discussion (through policy) before a vote could help. Jc86035 (talk) 12:09, 18 May 2019 (UTC)

Following separate discussion

Of course people like to follow or watchlist separate discussions on the same talk page; be able to neatly archive discussion threads, and search/find them afterwards. One option is to have each discussion on a separate sub page and list the separate discussions on the talk page by transcluding the sub pages. Archiving is easily done by removing a transclusion of the sub page from the talk page (A button next to the title of the discussion on the talk page to archive/remove/suppress the discussion would be a helpful add on). Following a specific discussion is easily done by watchlisting the sub page. (An open five pointed start next to the title of the discussion on the talk page would be a helpful add on). Nearly all functionality already exists. The 'Add topic' tab needs to be redone. Hitting 'Add topic' should in this scheme create a new sub page which is automatically transcluded on the talk page. That could be entered at the bottom of the page, or at the top of the page. The 'Edit section' putting should let the user edit the designated sub page - and after saving they might want to return to the talk page where they came from. Ad Huikeshoven (talk) 12:39, 18 May 2019 (UTC)

Is a indentation the best way to indicate start of a post?

Thank you friends for this work to summarize the contributions to a report. I see only one point clearly to improve: the discussions about the way of how to do the indentation does keep out the question, if indentation is the best way to indicate the start of a post (comment) and relation to previous posts. A common clear start of a new post e.g. on facebook and other social media shows WHO postend WHEN together with a clear end this is a frame. Only with a clear start and end of a post I can see, if the post is related to previous level of indent - else an indent only shows that there is a new post but not, if related to previous post or pre-previous. Suggestion: Username and Date-Time can be created like a signature by shortcut e.g. Charis19052509 and the old current signature can be set as End. This can be enhanced as clear identifier by anchor to relate to a post, also with a short header: Charis19052509:Clear start indicator of a post...post-body-text...Charis (talk) 09:44, 25 May 2019 (UTC)

@Charis: As indentation was previously used to organize messages, now we have to preserve its features.

The point with indentation is that it does not only allow to add new messages, but it also allows to respond to any previous message. I.e. you can have:

Initial message

Answer A

Answer to answer A

And possibly more stairs of answers…

Answer B

Answer C

And maybe stairs there too

Therefore to preserve this feature I think keeping the indentation system is a fair choice.

Frigory (talk) 16:57, 9 June 2019 (UTC)

Hi @Frigory, yes, for short postings the indent can be well overviewed, and I don't want to substitute the system of indent + signature but to enhance where needed :-) This thread is a example how difficult it is, if 2 articles refer not to each other but to the previous (see "Answer C" I added) having the same level of indent and to find the begin of the different answers as I only have the signature of the previous on the right side to show it. For this a clear indicator for the start of an post would be helpful. Also it's difficult to identify in long texts the level of indents so it would be helpful to link to the related article spoken over - and to have the option to relate additional links to other related posts. Charis (talk) 20:32, 27 June 2019 (UTC)

[Really need a visual separator here!]

Pelagic: .......................................... [reply]

I agree with both Charis and Frigory. Indentation should mark whom you're replying to. Though referencing Frigory, I'm primarily replying to Charis, so I've indented only one level. (In mobile preview this seems to visually run on from Frigory's post, unless I put in two blank lines, which is frowned upon.)

A general reply to the discussion "should" do this (indent one level below OP), IMO, but it's common to see the other approach where everyone indents and the stair-step effect becomes ridiculous. We need strong agreement on an accepted convention before we can create UI/automation that implements it.

If indentation indicates the level of reply, then what delimits the actual post? Currently, it's the signature at the end.

That only has predictable structure if you use the magic number of tildes (to get the time-stamp) and your user sig isn't overly customised.

One approach would be to introduce new markup. We have section headings with edit buttons/links: I can't overstate the utility of being able to edit a section rather than a whole page (and I have the top-edit option turned on in my pref's – sorry can't remember the proper name). Why not add a special comment heading that has a reply button instead? I've simulated that above, but it doesn't have to be formatted as a plain header, it could be box-and-colour or some other visual combination.

We might need an end-comment mark to protect against cases where the next poster is editing manually and forgets their own start-comment heading. Sure, the boxing could be achieved with a template or pair of templates, but doesn't that mean maintaining separate templates across multiple wikis? Having an end tag means how do you deal with the case where somebody else edits between those tags? Dealing with that could be worse than dealing with users forgetting their comment header, though it would be less frequent.

In a long thread, you don't want to scroll up to the OP to hit reply, so we'd want some kind of "add to this discussion" button that's at the end or, even better, always available.

I know I'm mixing discussion of markup and GUI, but whilst people want the option to edit manually, the markup is the UI for many users.

What if we could repurpose the tilde (or choose another easily-typed symbol) and combine that with the user-date-time identifier that Charis mentioned ? And have it automatically be an anchor the way current headings do.

+++Pelagic-20190615-0755
::... comment/post here ...
::[[User:Pelagic|Pelagic]] ([[User talk:Pelagic|talk]]) 07:55, 15 June 2019 (UTC)+++

or

^^Pelagic-2019-06-15T07:55 Optional heading/subject display text
... comment/post here ...
[[User:Pelagic|Pelagic]] ([[User talk:Pelagic|talk]]) 07:55, 15 June 2019 (UTC)
^^^SomeoneElse-2019-06-15T08:58
Indented reply.
[[User:SomeoneElse|SomeoneElse]] ([[User talk:SomeoneElse|talk]]) 08:58, 15 June 2019

Tools and bots that deal with normal headings could remain blissfully unaware of the comment header. Ones that deal with sigs might run into trouble with closing marks, if they were to be adopted.

For manual editing, the Reply button could pre-fill the the codes that get expanded in to the header and sig. The leading colons are messy, but we might need to keep them for backwards compatibility (rather than, say, multiple levels of comment headers). Maybe the tooling could automatically add the correct number of colons for you.

Sorry for the long post, Pelagic (talk) 23:06, 14 June 2019 (UTC)

@Pelagic: About signatures, I had proposed (in French) a few technical solutions here. To summarize: use a syntax {{Signature:User name|Timestamp (in the standard Unix format: number of seconds spent since the very beginning of year 1970)|Potential prefix such as ‘--’|Potential use of a ‘forgotten signature’ template}}, and this {{Signature:User name}} would possibly be just a shortcut for a somewhat special user page {{User:User name/Signature}} that would have a default content [[User:User name|User name]] ([[User talk:User name|talk]]). To quickly insert the {{Signature:, the four tildes would still work.

To my mind, the Reply button is correctly placed at the end of the message (possibly the one that started the topic), just like it’s done in Structured Discussions (but with them, we can’t have several stairs below one message). You should reply to a message only once you’ve read it entirely, preferably.

In my opinion still, the wikitext shouldn’t be changed for more than signatures, and maybe adding a meaningful tagging system.

To avoid the annoying ‘stairs’ graphic, I think it would be a good idea to use tabs, e.g. you would have:

Charis’s message

Frigory Pelagic

Frigory’s message

Subsequent answers to Frigory (possibly with tabs if several people reply to my message)

If you click on ‘Pelagic’ it should show your answer and the replies to it (but I think I can’t implement it here).

Frigory (talk) 16:36, 15 June 2019 (UTC)

Personally, I don't like either of those "closing marks" suggestions. I'd rather see a new indentation character for comments that behaves more sanely when combined with other wikimarkup, e.g.

Original post. [[User:Example|Example]] 01:00, 8 July 2019 (UTC)

+ Reply to the original post. [[User:Example2|Example2]] 02:00, 8 July 2019 (UTC)

++ Reply to the reply.
++ {| class="wikitable"
++ |-
++ | This table should render correctly despite not being at the start of a line.
++ |}
++ [[User:Example|Example]] 03:00, 8 July 2019 (UTC)

+ Another reply to the original post.
+
+ This has multiple paragraphs. [[User:Example3|Example3]] 04:00, 8 July 2019 (UTC)

+ A third reply to the original post. [[User:Example4|Example4]] 05:00, 8 July 2019 (UTC)

That would render each comment with appropriate CSS (e.g. a divider or a bar to the side) to make division of posts obvious, and the blank lines between posts at the same level would be required (otherwise they'd be seen as part of the same post, and probably run together) and wouldn't cause a w:WP:LISTGAP. Anomie (talk) 14:02, 8 July 2019 (UTC)

Hi @Pelagic, I would not recommend to substitute the system of indent + signature but to enhance, where needed, so it would be upward compatible ;-)

Wouldn't an automatic bold identifier at the begin of a post by username and date/time be sufficient? CharisJJMMDDhh see my initial post or CharisJJMMDDhhmm or any other time format that is short and can be sorted well Charis (talk) 20:32, 27 June 2019 (UTC)

Other discussions

@Whatamidoing (WMF) and Trizek (WMF): Is feedback that falls outside the six given categories necessary or desirable? I personally think it could be useful, e.g. to hold straw polls. Jc86035 (talk) 14:16, 18 May 2019 (UTC)

Well... our own home wiki has historically been the most visible exemplar of the m:Polls are evil mindset, so it might not be appropriate there. Other communities take a very different approach and settle everything by a simple majority vote. I would say that you should follow the local culture, whatever that is. If you think that a straw poll would produce useful information and would be normal in that community, then that's okay.

That said, it should go without saying that the outcome of this consultation isn't a vote. The team is taking in information and will make the final decisions. Some of that information might come in the form of "13 of the 18 editors in this one discussion voted to...", but that information is equivalent to all the others. Rationales, details, and examples matter a lot. (And even the end of this consultation won't be the end of decision-making; if "popular" and "logical" changes fare poorly in user testing, then they'll have to change their plans. I've been living with the "just what I asked for, but not what I want" situation in VisualEditor these last few weeks, and I would hate to see that kind of outcome on all of our talk pages.)

Something related that's been on my mind is prioritization. The foundation has gotten futher along in its strategic planning while this consultation has been running. At this point, I think that convincing them to take on a multi-year "massive" project is going to be a hard sell. I'd help you make that case, of course, if that's what people decide they need, but realistically, I think we're looking at what can be done in the space of a year or two, and then the team (which is still unassigned) will stop and let the dust settle. So if you have ideas about which thing to do first (wikitext changes first, or moving categories into a proper metadata space first?), then that could be useful information. Whatamidoing (WMF) (talk) 21:17, 18 May 2019 (UTC)

@Whatamidoing (WMF): I think straw polls were a bad example; the main reason I would be starting other discussions would be to further advance the discussion so that that information could be taken into account in phase 3 (e.g. doing basic design demos by messing with HTML; asking about other topics which aren't covered by the six questions like signature customization and different indent/discussion types).

On prioritization itself, changing the interface on discussion pages would probably be my first priority, but this would mean doing a lot of the mildly disruptive changes as groundwork (e.g. changing syntax and the parser, introducing magic words). The disruptive changes could also allow for section watchlisting (depending on implementation, of course). Alternatively, if this is shot down, it would probably be passable (but not by any means a good solution) to translate reply-link and enable it globally after changing the mainspace interface. Jc86035 (talk) 06:30, 19 May 2019 (UTC)

Realistically, would it be possible to, basically, put the Flow interface over wikitext comments in the time available? I have no idea how those work under the hood so I don't know whether this would be possible. In my mind you could get a fairly sophisticated interface (diff links, user hovercards, etc.) just by dumping all the necessary metadata onto the talk page upon signing/saving a new-style comment or adding a new-style level 2 section header, but I don't know how feasible that would be. In theory, at least, you could generate something that looks like a Reddit comment out of a bunch of wikitext comprising the comment, the username, the date and the revision ID (or even just the revision ID).

The reason for doing the aforementioned design demos (essentially, taking four divs from Flow and putting some more CSS and fake buttons on them...) would be to get some early feedback about various possible interface features (e.g. how cluttered should the interface be, how much padding is too much padding, should hovercards be used, etc.), assuming that revamping the interface is something that's probably going to be done and that it would be done without using hacky tacked-on scripts that don't always work (i.e. the reply-link band-aid). I don't really know if the team would find this sort of thing helpful. Jc86035 (talk) 08:56, 19 May 2019 (UTC)

Is it possible? Well, in theory, it's all just bits, so probably. Would we be happy? Well, it would put some restrictions on the people who would opt out of the changes. Their personal cost:benefit ratio would be negative. For example, imagine that making that interface work required a sort of AbuseFilter that won't let you save the page if you've broken the ==Section heading== like I did in this reply. The people who used the new tools wouldn't be affected (because a typo or stray click couldn't break the section heading), but the people who used the old editors might be frustrated. As with most changes, I assume that some people would accept this, and others wouldn't.

The analysis you wrote above ("On prioritization itself") is likely to be very useful. I'd be happy to see something similar from lots of people. :-) Whatamidoing (WMF) (talk) 22:21, 19 May 2019 (UTC)

@Whatamidoing (WMF): Thanks. I hadn't considered that some users might be opposed to even small changes like warnings, but I think it would probably be fine if it doesn't outright prevent the edit from being saved (e.g. "you're deleting a section header; are you sure you want to do this? [publish/cancel]").

I agree that the changes could even be a net negative for those who continue to write comments as source (not that I can see that happening very often if the inline reply interface works exactly like a standard wikitext editor with a visual/source button), and this could actually have a fairly significant impact on user warning templates and the like, but I imagine that you could probably make something work if you were to change how four-tilde signatures work on talk pages (while keeping three- and five-tilde signatures the same). Jc86035 (talk) 07:02, 20 May 2019 (UTC)

Comments by Cedders

I followed the invitation from the top of a Wikipedia page and added my comments on the six questions at Talk_pages_consultation_2019/Individual_feedback. To summarise, I think any change should be incremental, but general across MediaWiki use cases, and wonder about testing more newbie-friendly discussion tools based on additional 'suggest' or 'reply' buttons, prefilled signatures, and an enhancement to the TOC so it can function as a more useful thread index. --Cedders (talk) 08:32, 28 May 2019 (UTC)

Mobile website has a talk page browser.

I am not sure whether it has already been mentioned, but the mobile website does have a topic browser for talk pages. --Chanc20190325 (talk) 23:50, 28 May 2019 (UTC)

Women

Design is not neutral. I am convinced that design could be a differentiator between connecting to specific groups of people and putting them off. As we seem to have much less women than men editing Wikipedia content, and we would like to have that more even, any design change should explicitly be tested for and by female test groups. Titusmars (talk) 09:45, 31 May 2019 (UTC) (man, actually)

Also, breakdown the female group into new and experienced users. Geraldshields11 (talk) 15:05, 12 June 2019 (UTC)

+1 that--Mkouklis(2) (talk) 17:51, 12 June 2019 (UTC)

Annoying? Annoying …

Introduction: "[…] annoying for experienced editors. Counting colons to indent a reply properly, using tildes to sign your name, having to watch an entire talk page instead of the section you're participating in, not having an easy reply link – these are headaches for everyone."

Well, not for me – and possibly others. So, please refrain from speaking for people you are not speaking for: "everyone" – how presumptuous.

Annoying was – in opposition to many talk pages I've seen – a sheer endless page of content … which was simply 'too much' to read through or understand its meaning(s) (due to not having read through).

So much for talking … CompleCCity (talk) 22:42, 31 May 2019 (UTC)

@CompleCCity: This is not presumptuous, this is respect. Non-automatic, non-semantic features are always tiring. Here we are talking about talking, and in this context, counting colons, writing the tildes, etc., it’s boring, and venerating all this is autistic. By saying ‘autism’ here, I’m referring to the non-fatalist schizophrenic behavior (not the neurodevelopmental condition which is completely something else and is all-life determining). I feel that when indenting, you’re like: ‘Oh! Writing eight colons for the eighth message! And eight more for each new paragraph! That’s so genius!’ — No, it’s not genius. It’s autistic. Everybody should dream of having clean, visual, semantic and efficient computer interfaces, instead of the dirty, autistic computer code. So it’s not presumptuous, it’s respect to assume you’re not schizophrenic as long as we don’t have any proof that you are schizophrenic.

Well, be sure that the fact you’ve got a frank expression is not a proof that you are schizophrenic — . Expressing yourself is something good (and schizophrenia is bad). Frigory (talk) 17:32, 9 June 2019 (UTC)

@Frigory: Now, if I understand you right, my "presumptuous" was specifically about being addressed by the term "everyone" in connection with "experienced editor" (which I would call me) and attributed with opinions I don't actually have, while your "respect" is about the topic itself and in general. I suppose …

I did not say that these mechanics are "genius". But if not getting a "headache" or being annoyed by them makes me autistic or schizophrenic, well then … I'm still not "everyone". And although this perhaps makes me a good candidate to actually read through and understand that article, I'm perhaps also the wrong person to invest their time (and odd mind) to participate in this project. This article … I mean, actually the focus of what I meant to tell was on the discrepancy between trying to simplify things and making them easier to use – by setting up a page that by its pure length and layout makes reading it one of the most "annoying" things that may happen on a wiki (besides admin elections, perhaps), and guessing by other comments, I'm not the only one who thinks this way. (By the way, I'm not that involved with the big and official wiki sites, more on gaming related ones.) How much time did it took you to read it in its entirety? CompleCCity (talk) 19:12, 9 June 2019 (UTC)

@CompleCCity: It took me many hours, because I translated it into French. So you mean you don’t like the report and its layout? It’s quite a professional work, so it’s a bit long, but nobody has to read it entirely. I just hope they didn’t write it to fake caring about people’s speeches but really care about all that.

There is no ‘discrepancy between trying to simplify things and making them easier to use’; I mean it’s not a discrepancy: simplifying things often makes them easier to use. I don’t see what you mean with that… Actually I don’t think that the proposed direction will lead to a simplified system: on the contrary, it will be more complex, because we will have both wikitext and a modern interface. But the idea is that you will be able to ignore wikitext and recognize the standard and efficient forum conventions. Sometime, wikitext may disappear and we will be able to really simplify things by structuring the data in databases, which will make them closer to the algorithms we use to provide the interface.

Well, my respect was for ‘everyone’ as the ‘experienced editors’. If we have a proof that you are schizophrenic, then we won’t immediately write it in a non-medical report, because mental illnesses are something that touch the intimacy, and medicine is the only domain that can touch intimacy without being explicitly authorized. So to avoid telling everyone about your intimacy, we say you’re just like the normal people, who get headaches. If you are schizophrenic and you wish to reveal your intimacy on such reports, then you can say it, and then we’ll have to write: ‘These are headaches for everyone except some schizophrenic people.’ If you want that, then please also show us your diagnosis so that we’re sure you’re in position to speak as a schizophrenic person.

Frigory (talk) 20:22, 9 June 2019 (UTC)

@Frigory: It's fun – if I understand you correct – talking with you, but also a bit complicated – because you don't understand me always correct. Is this now because we're both non-native speakers (although I can't say of me that I work in the translation branch) or just for my clumsy English? Hm … no, French, that's ages ago for me … I almost knew that that "and" could be misunderstood, and because the actual second thing in the discrepancy should also start with an and – like it's usually done –, I switched to the dash. So, read: "discrepancy = (trying to simplify things and making them easier) : (by setting up a page …)" (you're good in (pseudo-)math?).

Now, here comes my disadvantage: due to not having read (and I don't think you will convince me to do so) the article, I can't tell if "the proposed direction will lead to a simplified system" or not, though I have interpreted your comment as if the system then is simpler in terms of easier to use for … everyone. But, if you take a look at my name, I'm (again) probably not the person to have difficulties with a more "complex" system.

Hey, wasn't it you who diagnosed the oddities (not to be confused with diagnosing odditis) of my mind? CompleCCity (talk) 21:20, 9 June 2019 (UTC)

@CompleCCity: OK, now I understand it for the dash – . It’s not much a discrepancy either, because it’s actually quite a complex work to have a debate that will determine the best way to simplify and ease things and that will also convince the concerned people that it’ll be cool. Once this work is done, anybody will be able to benefit from the simplification and easing, even though they didn’t participate to the debate. So even though you think the report is annoying, hopefully you will be able to ignore it and to just enjoy the simple and easy discussion system that will result of all this, thanks to passionate people like me who deeply involved in it. It’s quite an optimistic theory, but as I’m here this time, it should be realistic – . But maybe you were just ironical from the beginning.

Actually I’m not qualified to say that someone is schizophrenic. This can only be diagnosed when you are in front of a medic, knowing s·he’s a medic. The way you talk to a medic is needed to tell whether you’re schizophrenic or not, because you might not want to reveal some parts of your intimacy with someone else than the medic, and those parts may determine whether you’re perfectly conscious and greatly reasoned in your behaviors, or you’re just a fool seeking domination. But it’s a bit out of subject. Also note that when I imagine you saying ‘That’s so genius,’ I’m not saying you think about such an expression, but it’s expressing (or trying to express) what is not expressed in you.

I’m autistic in the second sense; I mean I have the neurodevelopmental condition. So this might explain the communication problems – . I didn’t feel your English is particularly clumsy. But in my autistic case, it takes me much energy to try to determine what you actually wanted to say when the construction of the phrase allows several possibilities. In accordance with this, I always try to write phrases whose construction allows only a single interpretation — so I’m coherent and this distinguishes my autistic case from a schizophrenic case.

I was skilled at math until the end of high school, and not much after. I don’t know what pseudo-math is.

Please take care to continue with the colons: you should only add one more colon at most at each newline. If you break our exchange with an empty line without indentation, it will also be done in HTML by closing/reopening tags, so it’s semantically evil and it causes me to have headaches. Moreover, if answers are styled just like here, it will break the layout. Frigory (talk) 22:52, 9 June 2019 (UTC)

@Frigory: I wasn't aware of the semantical importance of colons for blank lines – thanks for pointing this out: I will regard your advice in the future. However, they remove the empty line intended by the blank line …

Regarding the clumsiness of my English: I should have used "correctly" when it was about understanding each other, rather than "correct".

Okay, I'm not familiar with your condition, and what I 'know', I know from movies – so much for that. (Well, with a few exceptions perhaps, from some stories by an ex-colleague about her son.) Could it be that you have issues to understand irony? (Which is also said about children, but I'm not absolutely sure it's right.)

Then let me clarify some things: I was not ironical from the beginning or later … I mean, sure, there were parts of my comments being this, but certainly not the whole thing and per se.

Saying, you were the one "who diagnosed" was ironical, as I believed, many of your talk about schizophrenia would it be as well; now it seems to me that you were very serious with that topic. Let me assure you: as far as I know I am not schizophrenic; and I had many sessions with psychologists and even some psychiatrists.

I might show some minor symptoms of autism, however, but that hasn't been diagnosed either.

"Pseudo-math" was ironical: I tried to clarify what I wanted to say by using parentheses – as is done in mathematical formulae. Poor attempt, however, because no better symbol came to my mind then a colon (and I don't think there is a mathematical symbol for "versus" or similar). So, to understand the parentheses, you needed to know a bit about math, but due to the lack of a matching symbol and the poor construction of that formula, I called it "pseudo".

I have problems to understand the meaning of your "when I imagine you saying ‘That’s so genius,’ I’m not saying you think about such an expression, but it’s expressing (or trying to express) what is not expressed in you". I should – I'm not dumb … mind to explain, anyway?

I have tried to write this without ambiguities – did I achieve that goal? And nothing's ironical in this latest reply.

Oh, and to keep this discussion true to the topic: perhaps I do read the article or parts of it some day. And if this thread's number of colons becomes too annoying, we could use {{Outdent }}. (Okay, using the term "annoying" in this sentence has some irony. ) -- CompleCCity (talk) 15:04, 10 June 2019 (UTC)

@CompleCCity: You can keep the ‘empty line intended by the blank line’ by removing just one colon (instead of removing all the colons). It’s a fair solution to me. This is what I did for your previous message.

As there is an answer below, I can hardly use {{Outdent }}…! Well, this is just related to this discussion above.

As long you don’t explicitly say that you’re ironical, I must act as if you weren’t, even if I know you are ironical. There is no other universal sign to certify that you were ironical. In justice, in trials, we can’t easily say if a person is ironical, unless if it was explicitly acknowledged, or if the person would otherwise be considered as a criminal person. E.g. if I say ‘I’ll kill you’ it will usually be considered as ironical (even if nobody said it explicitly), because otherwise it would be an illegal threat. But quite often, justice considers that some murders are cool and does not punish (I have heard of it in the subject of feminicides), so in that case you may consider that the threat is not ironical and is illegal, even though it’s a bit absurd to criminalize the threat and not the murder, but at least it might avoid that the person is killed.

I had perfectly understood that you were ironical when you said that you could be schizophrenic – . But imagine that we had a judicial conflict on this subject. Then you could say you were not ironical at all. If I supposed you were, it would be bad for me. But from there, I can say I hadn’t understood that you were ironical and it might save me. Moreover, I like the subject of schizophrenia and autism, so it was cool to talk about it.

About your clumsy English:

• indeed you wrote ‘discrepancy between’ which implies ‘between … and …’ so I misunderstood; you should have written instead: ‘discrepancy: simplifying and easing – by setting up a complex page,’ then I would have understood (so keep in mind that ‘between’ goes with ‘and’ while the colon ‘:’ fits right with a dash ‘–’, the discrepancy was not the problem);
• I think you were talking about my piece of advice and not all my advice, otherwise from now you should blindly listen to all my advice, beware – .

About ambiguities, you did it right, your message is very clear – .

I had understood the tip about the parentheses; you gave two ‘mathematical’ terms. It was natural for me, because I’m actually quite good at math and computer programming (I don’t have much more than a high school diploma and a small IT diploma, but I’m sure I’m much better than most people).

For the ‘genius’ thing and the need of expressing, I have to reflect a bit on what I said… Please wait a moment… Well, I’m ready. Your speech showed that you accepted that the colons were a cool system (‘cool’ does not mean ‘good’ but ‘neither good nor bad’). But in our society (at least in mine, maybe you’re not in the same one – ), there is actually not much consideration for the neutral point of view, so as long that you consider that something is ‘cool’ then you might accept that it’s ‘good’. You should have an opinion on things. E.g., do you think murders are cool – ? No, they’re not, they’re bad. You can say it’s bad because you have arguments to prove it: the death will be terrible for the people who liked the murdered person. It’s quite the same thing for the colons: you’ve got arguments to say it’s a bad thing (i.e. it requires some effort) but you continue to think it’s cool. Maybe you think it’s more ecological if people type ‘:::::::’ compared to if the computer adds it automatically, but it’s not, waiting for the colons to be typed will actually use more electricity than if they were added automatically. I try to be a bit short; I hope it’s clear, but I can write more if it may help.

Frigory (talk) 17:44, 15 June 2019 (UTC)

Wikipedia in its many incarnations has always had a fairly strong bias toward OCD and/or autistic editors, who on the one hand are the core backbone of keeping the network relevant and ever-expanding, while at the same time extremely resistant to any kind of change whatsoever, regardless of who or what it benefits. (This has held true through many changes over the years, with newer editors becoming resistant to changes to features that older editors never accepted in the first place.) No matter what, there will always be those who take their ball and leave, and that's OK, if it gets enough others contributing; the history of Wikipedia is littered with much drama and many angry exit tirades, yet here it is, bigger and vaster than ever. Because such a large portion of WP's core fits that bill, though, changes have to be measured and have to garner enough buy-in (like the popularity rankings of potential features), lest there be a real community revolt. I do believe the WP leadership understands this, though, which is why they haven't just imposed wholesale changes on the site and told everyone to live with it, like Quora does all the time. SilverbackNet (talk) 10:27, 14 June 2019 (UTC)

Maybe this is working?

I got a link on the top of my page about the Wikimedia Foundation wanting to make it easier to "talk to other editors." I clicked through and it took me to the a completely unexplained page labeled "Talk pages consultation 2019/Phase 1 report." After skimming the report, it aligns with "wtf is this, I want to participate but this is too byzantine." It literally linked me to a page that has no instruction or follow up with the general pitch of "click this button because we want to make it easier to communicate" and then not making it apparent what's going on or making it easier to communicate. And the sad thing is, I'm probably an edge case of almost participating but not willing to learn an archaic coding language just to understand the total BS talk-page coding that I've clicked on a few times in my life and given up. Anyone else with less experience, but more to contribute than me? Good luck getting them. I wish you all the luck in the world, and sorry for probably (?) breaking the rules just by making this edit. — Preceding unsigned comment added by Nindustrial (talk • contribs) 06:04, 1 June 2019 (UTC)

I completely agree: the proposed changes page is baffling. I have no idea what is actually being proposed after reading it.ThatMontrealIP (talk) 15:36, 1 June 2019 (UTC)

Please do not let Wikipedia become like social media

The aesthetic and design changes on some icons and such were enough to leave me morose as the entre internet is swept away in this wave of modern change that abandons reason and stability for the sake of being sleek and open , and in general has caused me anguish for various reasons including physical painful stimulation. However I and many others in this case especially are perfectly fine with the good old days of navigating discussions and saying what needs to be said when necessary on the traditional channels. I actually wanted to write a long rant about all this, but instead I'll just close with this observation and a maxim: 1. Many Wikipedians suffer from mental disorders which make dealing with significant changes in sturdy sections of life quite toiling and painful and a true struggle. 2. If it ain't broke, don't fix it.--Sigehelmus (talk) 07:00, 2 June 2019 (UTC)

We do have a working minimal viable option in place that many long time editors are hesitant to change. As a relatively new editor, I have to agree that Visual Editing is much more user friendly and allows for editors to dive into the work rather than search for formatting information. Still, I fail to see how signing you name is difficult. That is a unique WIKI method that has character and is the same across the board for all- four simple taps on one keyboard button... if we forget, the bots add it for us anyway. Adding the name of another user, that is difficult.

• mental health

To specifically address the mental disorders comment, I am part of that challenged group and yet believe just like many change their anti-depressants every few years, it is good to change things up and allow us to see life with new perspective, hopefully without as many side effects. Enhancements to our minimal viable option should be welcomed. Also as we age, while we dislike change, small changes are actually brain testing options that we need. The ability to continue learning should never become for the young only. Think of changes as sudoko.

• social media

On the comment title itself, "do not let Wikipedia become like social media" unfortunately it is already in many ways. Discussion sections of articles are not the only place editors post comments. Many Talk pages are never used and actual article work taking place outside of this discussion area is barely seen. A number of editors prefer to head to 'common spaces' like the tea room for immediate assistance. Their Talk page question may never be seen by others also investigating a page and its archived history. Is there a way to search for conversations across wiki that includes each individual request for input no matter where on Wiki the conversation occurred? Social Media is known for its many backdoor conversations. A Facebook post is commented on in a private messaging conversation or an instsgram post is berated on Facebook but not on Instagram. That leaves many clueless about the dialogue and from these alternative discussions, entire positions of mass thinking can be formed which directly effect the tenor the original post is seen as well as the kind of responses that are next generated. These sidebar conversations have ramifications.

• Where do we talk and where should we talk?

Should one post in the tea house for expediency or to gauge consensus before adding to a Talk page. I think it should be the other way around. Talk on the talk page first so all may benefit from the conversation that impacts the article in question. Mrphilip (talk) 11:29, 3 June 2019 (UTC)

About signatures:

• The fact that you need to sign – that you need to manually click a button at all – is not known when people make their first edits. The relatively few people (like the three of us) who get this right on their first talk-page edit are often suspected of having prior experience (or of having read the directions).
• It's only "four simple taps on one keyboard button" if (a) you know you need to do that, and (b) your keyboard actually has that button, which is not the case for multiple languages around the world.
• Very few communities can rest assured that "the bots add it for us anyway". Special:CentralAuth/SineBot serves exactly two wikis, and there are almost 300 separate Wikipedias. Unsigned comments on this wiki, for example, are not fixed by bots. Whatamidoing (WMF) (talk) 17:59, 3 June 2019 (UTC)

@Whatamidoing (WMF): @Mrphilip: I know this is literally about four years late and I'm SO sorry and don't even remember making this post (well, maybe a bit, but not with such melodrama, forgive me). All I'll say is that while since 2019, there's been a bit more change in that social media-ish direction, and I don't want resurrect this thread from the dead, but I've also seen some good changes though I believe there can be a synthesis of a more "traditional" wiki style with smooth and digestible presentation for the masses and all types of wiki users. It doesn't have to be a dichotomy of following Web 1.0 legacy or Instagram/TikTok. Overall, to paint a lucid and simple symbol: There is still nothing wrong with serif fonts, hard edges, and the like and it in fact helps the aesthetic "tone" for what Wiki strives to be in design psychology, and it still can be arranged with elegance that isn't stiff/outdated. So, welcome to 2023!--Sigehelmus (talk) 00:12, 11 May 2023 (UTC)

@Sigehelmus, I wonder what you think of the changes described at Talk pages project/Usability. Whatamidoing (WMF) (talk) 22:40, 20 May 2023 (UTC)

Thank you very much for the reply + tag. Honestly, I could do barely more than glance because self-sabotagingly I almost don't want to know whatever is being proposed that might turn Wiki more Corporate Memphis. Pls no. I'm not calling for a return to Frutiger Aero style per se, but all I want is serif fonts and hard ages not to be obliterated. If anything, a more "authoritative" visual design helps Wiki seem more stable/reliable/timeless and less mutable to the whims of whatever current editor. Right now for example as I type this reply: I think the current layout isn't terrible, but it needs more organizing lines and boxes. Maybe a sepia color option, actually, not just white or dark. And the symbols for "Add Topic", user, magnifying glass, etc. should have at least minimal shading and again, more hard edges. The "View the content page" symbol shouldn't be this rounded-edged tablet but be like a page perhaps slightly-turned with a bit of depth and color. The Watchlist Star and Page History symbols are half-decent if Minimalism is forced; the Edit pencil definitely should have a splash of color/depth too that invites the User to actually, y'know, contribute considering how few people actually edit wiki and are afraid to dive in. In essence: I still believe the traditional Wiki symbols style is best, but it can be made more sleek and simple WITHOUT becoming soulless slop. People should see Wiki and not think it's a careless wasteland of random junk, especially considering our current rep as well as the buzz around "AI"-generated content. Wiki should feel human. Does this make any sense? :( Sigehelmus (talk) 02:38, 22 May 2023 (UTC)

It sounds like you're concerned about the overall look and feel, rather than being specifically about communication between editors. I'll leave a note on your talk page. Whatamidoing (WMF) (talk) 18:24, 23 May 2023 (UTC)

Mobile users

Every time I search for the talk page on mobile I open the sidebar accessible via the button with the 3 lines in the top right, because that is where I expect it to be every time. As for the usefulness of not (1) having to scroll down the entire page, (2) change to desktop view, (3) go to discussion tab, (4) changing to mobile view, (5) remembering you cannot add topics in mobile view, (6) changing back to desktop view, (7) clicking add topic, (8) changing to mobile view:

If it took half as much input & time to access a talk page, I may actually use the system. Strikefinger (talk)

User:Ckoerner, this feedback on the mobile site sounds like it might be interesting to you. Whatamidoing (WMF) (talk) 20:44, 3 June 2019 (UTC)

Strikefinger, I have some good news to share. The Readers Web team is developing what we're calling Advanced mobile contributions. First thing up on the list is moving the talk page link to the top of the page on the mobile site. It's only on a few wikisfor now. More to come soon. :) CKoerner (WMF) (talk) 20:43, 5 June 2019 (UTC)

Good to hear of these improvements Newystats (talk) 04:26, 15 June 2019 (UTC)

Yeah, basically you'll need a few years to reverse every customisation made on MobileFrontend until it finally reaches (back) the functionality of the normal website. Nemo 08:07, 17 June 2019 (UTC)

Comments by Hawstom

I'm an old user and admin (2003) on English barely active any more. I like the Proposed product direction as far as the Talk pages go, and I agree with its conclusions. But since harnessing the dilettantism (dabbling and passing interest) of new users is key to Wikipedia's genius, I was strongly impressed by the New contributors section of the report. And I think that (to better harness dilettantism) the product direction (in addition to its focus on Talk pages (How to comment)) must also include some small design changes to article pages (Where to comment). Hawstom (talk) 01:13, 3 June 2019 (UTC)

Here are some possible ideas:

1. Put all the "My pages" links including Hawstom | Alerts | Notices | Talk | Preferences | Beta | Watchlist | New messages | Contributions | Log out into a vertical dropdown below username. Hawstom (talk) 01:13, 3 June 2019 (UTC)
2. On each section of an article, include these links [ edit | discuss ] or [ edit section | discuss section ]. Make the "discuss" link open a new comment section either appropriately titled under the old system or using the new talk system. Hawstom (talk) 01:13, 3 June 2019 (UTC)
3. After the main, large article title, include the same links as at each section or expanded for clarity [ edit article | discuss article ]. Hawstom (talk) 01:13, 3 June 2019 (UTC)

The above seems rather obvious. But it was somehow omitted from Proposed product direction. Hawstom (talk) 01:13, 3 June 2019 (UTC)

@Hawstom: The Timeless skin already uses a vertical dropdown for those user links. You might want to try that (you can change your default skin at Special:Preferences#mw-prefsection-rendering, or you can append ?useskin=timeless to a page URL).

Adding more links could be a good idea, although it might end up inviting off-topic discussion that's not related to improving the article if it's not implemented properly. Jc86035 (talk) 09:02, 3 June 2019 (UTC)

Translations

Happened to look at the last comment, in French, just above the Conclusion section; and éventuellement is not "eventually". This is an elementary translation mistake of the false friend type. Do you need translators? Mathglot (talk) 10:18, 3 June 2019 (UTC)

Mathglot, have I ever not needed translators? Please be bold and fix the translation(s). Whatamidoing (WMF) (talk) 20:48, 3 June 2019 (UTC)

I made the fix. Frigory (talk) 14:23, 7 June 2019 (UTC)

Wikipedia is not broken.

It's not broken. There's no need to fix it. Certainy catering to the lowest common denominator isn't going to lead to better articles or editing practices. It takes wit to edit an encyclopaedia. 92.237.223.240 23:49, 3 June 2019 (UTC)

While I don't think the person who wrote the above comment is going to see this, I think it's worth noting that this is clearly a massive oversimplification of the situation. There are almost certainly people out there who would currently be long-term Wikipedia editors if not for its systemic problems and technical deficiencies (even assuming that "it takes wit to edit an encyclopaedia"), but it currently seems almost impossible to figure out much anything about those people or usefully approximate the number of people who fit into this category. I think a lot of experienced editors don't think that either or both of those factors skew the distribution of editors in a significant way, or have just never thought about it.

Because it's not known how large this group really is, there doesn't seem to be a way for the WMF to even vaguely objectively determine how much resources should be allocated towards improving things for active editors (e.g. improving/maintaining/rewriting volunteer-developed tools, automating tasks) relative to the amount of resources allocated to improving new editor retention (e.g. communication aids), at least if increasing overall productivity is considered more important than getting more people to edit (regardless of how much they contribute). Presumably, if the pool of people who would potentially become long-term active editors (if various issues were resolved) were insignificant to begin with, then there wouldn't really be any productivity advantages to improving new editor retention in the first place. (The number of long-term editors is now fairly stable on most projects.) I think this is an area that would deserve some actual study, particularly to somewhat quantitatively discern why people stop editing, never started editing or don't contribute as much as they ideally would, since otherwise any improvements targeted at editors are basically just throwing spaghetti at the wall in terms of increasing overall productivity. Jc86035 (talk) 06:17, 7 June 2019 (UTC)

Well said Newystats (talk) 05:40, 15 June 2019 (UTC)

I didn’t mean to illustrate that point by putting my comment in the wrong place, but I did. Newystats (talk) 05:43, 15 June 2019 (UTC)

If the given tools are improved and made better, then the current editors may have reduced burnout. That will help keep wikiprojects going. Geraldshields11 (talk) 15:10, 12 June 2019 (UTC)

About Template:Quotation with original

In the original report, quotations were provided with inline styled <blockquote> tags. I cleaned this using a template, especially for translation. Here I explain a few design choices of this template.

• The source parameter is put at the top, so that the translators will see its value in the translation interface. Then if the following quotation comes from a wiki in their language, they’ll know they mustn’t translate (if they understand the point).
• We removed the [en] specifications before the quotations’ translations, because on translated versions of the report, this will be English (if not translated), the target language (if translated) or even a mix of both (if partially translated). See here what it looked like before.
• Two quotations have a ‘top note’ to inform they are repeated. As the diversity of quotations brings value to such a research, I feel it’s nice to tell when one of them (in its substantial content) is reused. (Note that the quotations that were used several times didn’t have originally the same translation in English. I fixed this.)

You should always think about creating a template (or using an existing one) for such things. It’s cleaner (as it avoids repeating long style descriptions and some other info)! Just use the power of MediaWiki and its wikitext – ! It would have avoided me introducing the template using search&replace with regular expressions, not fully sure of what I was doing, fearing to destruct people’s quotations. And it would also have avoided me the need to discuss the structure of the template with an administrator (which lasted several days – ).

All that speech of Wikimedia users was very interesting – . And (as the time spent translating it suggests) I really liked the report!

Frigory (talk) 17:22, 7 June 2019 (UTC)

Too much text

It seems to me that most users are not finding the links to the discussion pages because of how much text there is to read above the links (this is evidenced by the English Wikipedia page views, which had a massive spike during the period that the banner linked to the phase 2 landing page). Perhaps one of those big buttons could be used to direct users straight to the links? Jc86035 (talk) 16:32, 12 June 2019 (UTC)

I would make a similar comment. I think a major reason why the average person does not participate in talk page discussions is because they often are extremely long and someone with limited free time is not going to commit 20 or 30 minutes to reading through a Wikipedia talk page. Directing people by use of a banner to this page which is 180,000+ bytes does not help. Am I expected to read all of this? Hrodvarsson (talk) 03:14, 13 June 2019 (UTC)

This sounds like it would dovetail with a more standardized Archiving feature, perhaps combined with collapsing longer talk discussions in general; is that somewhere I missed? It's too bad that Wikipedia never adopted a standard template to have a collapsed section full of section titles that link back to full archived discussions; instead the talk page is essentially blanked of old discussions, no in-between. SilverbackNet (talk) 10:10, 14 June 2019 (UTC)

Good direction

Speaking as one of the described people who would like to participate but finds the whole process a little mystifying, this direction sounds promising. I agree it makes sense to build on the existing talk pages rather than to build something completely new. So long as the underlying text is reasonably structured, other tools can make editing it more beginner-friendly.

Rattic (talk) 21:42, 12 June 2019 (UTC)

Thoughts from starsandwhales

I've been following this discussion for some time but ironically, because of the design of the talk page system, I've found it awkward and confusing to contribute. The current system is great on articles with only 1-4 people editing it at a certain time, but when the page has dozens of edits a day or when it is extremely controversial? The talk page gets messy and barely navigable.

• In my opinion, indentation is fine. The real problem is "threads" of conversations being easy to lose in the many replies to one statement.
• Having an easier way to navigate threads goes hand in hand with having easier ways to reply. Flow makes some sense, but it's not that great for editors wanting discussions and people who want to revert changes.
• There should be a more natural way to archive and close discussions.
• Notifications on specific threads is something that I didn't know I wanted, but now definitely need.
• Signatures are the least complicated part of the talk page.
• We should implement changes slowly, lest we confuse newcomers even more.

Starsandwhales (talk) 02:51, 14 June 2019 (UTC)

Addition of a voting feature in ‘Extension:StructuredDiscussions’

I am working on a MediaWiki project. I have installed the Extension:StructuredDiscussions on my wiki & it is working successfully. If it is possible, I want to have a voting feature in ‘Extension:StructuredDiscussions’, so that users can vote the replies of other users & the total score appears.(like in ‘Extension:Comments’) It would be very helpful, as the number of votes can be a clear & direct indication of how many users agree with a particular reply among others without them unnecessarily needing to say so in a reply. It would be very efficient & neat.This should be added to 'Extension:StructuredDiscussions'. Also, it can be a optional feature i.e users should be able to configure this functionality for their talk pages etc

And if no developers are interested in this ,can someone please guide me?,so that I can integrate this functionality in my wiki.... --Falcopragati (talk) 16:58, 21 June 2019 (UTC)

A voting module was planned, but never built.

Given how many wikis decide at least some things by a vote (e.g., ArbCom cases and RFA at the very "anti-voting" English Wikipedia, but much more at other wikis – Commons Picture of the Year is a quite elaborate voting system), I have been surprised how few editors even mentioned voting during this consultation. Whatamidoing (WMF) (talk) 05:26, 26 June 2019 (UTC)

@Whatamidoing (WMF): Will something like a voting module be part of the proposal for phase 3? (I think one reason that there's so little feedback on this might be that users were never prompted to discuss polls in either phase 1 or phase 2.) Jc86035 (talk) 08:57, 11 July 2019 (UTC)

Jc86035, I think this (temporary) group is giving up on a separate Phase 3, and turning the whole thing over to the (permanent) Editing team for further development. Unless my team decides to rearrange assignments internally, it looks like I'll still be involved. Whatamidoing (WMF) (talk) 17:23, 16 July 2019 (UTC)

Phase 2 report

The Talk pages consultation 2019/Phase 2 report is being drafted. You can put the page on your watchlist now, or check back later (hopefully by the end of this week).

If you are going to Wikimania:, then several people from the consultation team and the Editing team will be there, too. I encourage you to attend the related session and to talk to them directly. Whatamidoing (WMF) (talk) 22:50, 6 August 2019 (UTC)

Pinning the "About this board" on the right

Sorry for being late to the thread (feel free to move if a different talkpage is now more appropriate) The ability to mark some the right hand panel of some pages to by default be position:sticky; so that it stays visible during scrolling would be a useful option for long pages. Similar to the formatting of this example. T.Shafee(Evo﹠Evo)^talk 11:04, 1 November 2019 (UTC)