Wikipedia talk:Talk page guidelines

IP user talk page

I'd welcome other opinions on this edit.

It seems to me that it's not what the guideline intends at all. But in that, as far as I can tell, IP user talk pages are in namespace 3, the same as the talk pages of registered users, perhaps this needs clarification.

See Wikipedia:Administrators' noticeboard/Incidents#User:185.59.158.22 for some of the background to this. Andrewa (talk) 05:11, 26 June 2017 (UTC)

@Andrewa: It's explicitly prohibited by WP:REMOVED, fourth bullet. --Redrose64 🌹 (talk) 08:05, 26 June 2017 (UTC)

So it is! Thank you. Andrewa (talk) 08:24, 26 June 2017 (UTC)

DTTR

I believe we should codify in these guidelines that WP:DTTR is not sufficient grounds to justify removal of a post on a talk page other than your own. Though DTTR is often treated as a policy or guideline, it isn't one, there's an antithetical one called Wikipedia:Do template the regulars, and user warnings are specifically written to not be personal attacks. Thoughts? pbp 14:03, 11 August 2017 (UTC)

WP:DTTR is just an essay, which can be boiled down to "templates often treat the editor as brand new, and provide helpful links - regular editors (should) already know about all the links and what's expected of them, so treating them otherwise isn't the best idea". Regardless of if someone agrees or disagrees with DTTR, it does not give grounds to remove a post from a talk page other than your own -- There'sNoTime ^{(to explain)} 14:15, 11 August 2017 (UTC)

• Oppose See WP:CREEP. While I agree this is a correct reading of existing rules, observe that 3 of us read the existing rules as already frowning on third party talk page reverts of this sort. NewsAndEventsGuy (talk) 17:16, 11 August 2017 (UTC)

Very old talk pages

On old talk pages, I find some weird layouts.

---

text
reply

---

text
reply

reply

---

etc

What was that? — Preceding unsigned comment added by Nixinova (talk • contribs) 20:42, 13 August 2017 (UTC)

Guidance against interleaving replies

Proposed text for introduction in "Editing others' comments" section:

"Generally you should not break up another editor's comment to reply to individual points. Interleaving comments like this confuses the layout of the page and obscures the original editor's intent, as well as potentially leaving text unsigned."

Version 2 of proposal (19:49, 14 August 2017 (UTC)):

"Generally you should not break up another editor's comment to reply to individual points. Mixing comments like this confuses the layout of the page and obscures the original editor's intent, as well as potentially leaving text unsigned. Instead, place your reply entirely below the original comment. You may wish to use the {{Talk quotation}} template to quote a portion of the material in question."

I have encountered a few times situations where to respond to seemingly itemized multiple points an editor interleaves their reply within the post they are replying to, for example as Andrewa did here (I'm inviting them to continue the discussion here out of courtesy). It may be particularly prone to happen when a post has bulleted points which I have seen a couple times and which made a real mess of the talk page. The biggest issue is that it leaves the original post's text broken up and without signatures. If signatures were added after the fact that would, to me, definitely constitute editing another's post and changing what they intended to convey and how they wanted it to look, without improving the clarity of formatting. I think instead the proper convention should be to say something like Regarding X, "Regarding X," whether X is a description or a numbered point in cases where there is one, or Quoted material: with the tq template, and to do this entirely below the post you are replying to. I am proposing that some guidance be added in this regard. —DIYeditor (talk) 00:26, 14 August 2017 (UTC) Edited to correct use of tq template. 19:08, 14 August 2017 (UTC) Updated with Version 2 of proposal. 19:49, 14 August 2017 (UTC) Underlined version 2. 02:06, 15 August 2017 (UTC)

The convention as I have followed it is not to break up existing paragraphs. It is in my experience easily learned and followed, and common elsewhere on the Internet. The indenting makes the authorship plain, and the interleaving makes the logic plain. Respecting others' paragraphs leaves their comments intact.

The proposed addition doesn't really make it clear that this is discouraged, and I'm not convinced it should be.

The convention I have followed is however easily messed up, either accidentally or deliberately, and when this happens it can get very messy.

I'd like a stronger statement on the mixing of colon and asterisk indenting. This is the most common way that the convention gets messed up, in my experience. Andrewa (talk) 01:02, 14 August 2017 (UTC)

No, please do not do that. Consider what might happen if someone wanted to reply to you, and then there was some back-and-forth. That leaves a dreadful mess. Talk pages are not just for the benefit of those currently participating who might know what is going on. In a year, people might want to work out why a particular decision was taken or not taken. Johnuniq (talk) 03:10, 14 August 2017 (UTC)

Exactly. But the mixing of asterix and colon indenting is depressingly common, and as you say often leaves a dreadful mess... I'll dig up some examples. Sometimes I suspect it is even deliberate rantstyle (I might not give examples of that as it raises behavioural issues) but other times it is, disappointingly, experienced and respected users, to the point I sometimes suspect I'm just being grumpy to criticise it. But if we could avoid it, it would greatly increase the value of the archives, as you say, as well as making it easier IMO to arrive at and assess consensus in the first place. Andrewa (talk) 06:34, 14 August 2017 (UTC)

This edit is a case in point - I fixed three problems there:

1. blank lines, contrary to WP:INDENTGAP (and which incidentally I have also fixed in this edit);
2. signature divorced from post by interspersed comments;
3. markup symbols inconsistent bwtween a post and its reply which caused the enumerated list to restart at 1 instead of continuing with 3.

Mixing the three styles (asterisk, colon and hash) is not a problem per se, the problem is when people mix them incorrectly. The general principle should be that if you reply to somebody, copy the markup from the start of their post, whatever combination of symbols that might be; and add one symbol (of any type) to the right hand end. --Redrose64 🌹 (talk) 09:01, 14 August 2017 (UTC)

Agree that the protocol proposed in the above post copy the markup from the start of their post, whatever combination of symbols that might be; and add one symbol (of any type) to the right hand end would work extremely well if followed consistently. But we need to deal with bold edits by inexperienced editors as well as considered edits by old hands, and as even the old hands often depart from the relatively simple current rule of Generally colons and asterisks should not be mixed for no obvious reason, there's reason to be very afraid of a new and more complex rule. I think on balance it would be worth a try. Andrewa (talk) 13:16, 14 August 2017 (UTC)

You know, Wikipedia possesses specific tools which allows one to reply to a specific section or paragraph of an other contributor's post. For instance, the template {{Talkquote}} allows one to quote the specific part of the post one wishes to reply to, complete with signature and linked timestamp, within one's own post below which one can then post one's own reply.Tvx1 17:21, 14 August 2017 (UTC)

Certainly... at the expense only of brevity. But that can also raise objections, in my experience. Andrewa (talk) 22:39, 14 August 2017 (UTC)

• Support Although long posts can be troubling in themselves, interrupting their flow to reply to a specific point takes any shortcoming with the fact its a long post and multiplies that by 10. So don't do that please. In the discussion I noticed two related and somewhat side issues to which I reply as follows -

Re A) on messy format.... see WP:SOFIXIT. These guidelines already encourage stand-alone edits that only clean up formatting problems. I usually do not do that with regulars unless I get their permission first. But for newbies, don't hesitate, just do it, and give them a friendly how-to-do-better-formatting note.

Re B) on replying point-by-point.... hopefully my comment here shows how I do this. If the long post does not include numbers or letters so you can reply that way, just give the point you want to reply to a letter or number and say what you wish after the longwinded editor's signature.

In closing, I think the suggestion to not insert comments in the middle is a good one, but I don't care for the word "interweave". My brain stopped cold, I had to think, it was an obstacle. Better to just use simple third grade language, something only a bit more refined than "Don't butt in line". NewsAndEventsGuy (talk) 14:06, 14 August 2017 (UTC)

Perhaps "Mixing comments" rather than "Interleaving comments"? And how about some additional guidance like: "Instead, place your reply entirely below the original comment." and perhaps "You may wish to use the Talk quotation template to quote a portion of the material in question." I agree that keeping it simple would be good but maybe some clear advice on what to do in addition to what not to do would help. —DIYeditor (talk) 19:19, 14 August 2017 (UTC)

Positive guidance, rather than negative, is greatly to be preferred. It is both far more likely to be effective and adheres to the spirit of wp:AGF. Andrewa (talk) 19:45, 14 August 2017 (UTC)

I don't think those are side issues at all...

A) I would welcome strengthening the relevant guidelines to make that a bit clearer. In particular, In general... is vague. If the proposed more elaborate guideline (which is growing on me) is adopted, I hope the phrasing will be more to the point than that.

B) Yes, that works in cases like this. Another technique which I have employed is to start a new subsection on a particularly important point that is raised. I've received some criticism in the past for doing this, but generally from those who did not wish to hear what was said (at the risk of violating wp:AGF... sometimes the assumption wears a bit thin). Andrewa (talk) 19:45, 14 August 2017 (UTC)

Andrewa, as far as your A) on this this proposal maybe it should simply be "You should not" rather than "Generally". I wasn't sure if consensus would be behind a strong statement but it seems to be heading that direction. —DIYeditor (talk) 23:58, 14 August 2017 (UTC)

I must state that I cannot stand when an editor breaks up my comment to reply to individual parts. Any time that it is done, I either put my comment back the way it was or copy and paste my signature for each part of the broken up comment to make sure that others are not confused by who is commenting. And I ask the editor not to break up my comment like that again. Flyer22 Reborn (talk) 02:33, 31 August 2017 (UTC)

Modifying comments already replied to

This edit rather surprised me... wouldn't it be better to raise it as a new post, with a heads-up? Andrewa (talk) 01:11, 15 August 2017 (UTC)

It's got a time stamp. Perhaps should have been underlined if it is not clear by the time stamp that it was a change. —DIYeditor (talk) 02:03, 15 August 2017 (UTC)

Yes, it complies perfectly with the guideline on modifying your own comments (WP:REDACT) as far as I can see. But it seems to me far more confusing than my edit [1] which inspired this whole section. I think the indenting there makes the signatory of the original post quite transparent, but I concede there are other views on this. But I can't see how this edit can fail to tangle the logic of the discussion. The text to which I was replying is no longer there to see, you need to go into the page history to find it. How can that possibly be helpful? And yet it seemingly conforms to guidelines. Should it? Andrewa (talk) 04:10, 15 August 2017 (UTC)

Well, without software to manage talk page comments all we have are manually implemented guidelines. If users are to be able to edit their posts I think strike and insert work well enough, as well as the instruction to add a new timestamp when you've done it. On some forums where users can edit their posts the custom is to do something like "Edit: I did so and so" at the end of the edited post which I think is less clear than our strike/underscore method, but we also include a suggestion to offer an explanation if necessary in brackets. Or are you proposing that users not be able to edit posts once they are replied to? This is always an issue in forums; what it appears someone responded to may actually have been edited. Without strongly discouraging or prohibiting edits I don't see a way around it. My thinking was that it would be useful in that particular situation if the proposal were updated at the top which I believe I've seen done in other surveys. Maybe it would be useful to have clear guidance on that specific circumstance - what if you want to add another option to a survey. To me at the top makes sense as long as it is clearly marked as an edit with a time stamp. It also makes sense to include a "Survey" and "Threaded discussion" section which I wish I'd done to keep the two separate. Then any updates to the survey or !votes on it can be kept in the same area. —DIYeditor (talk) 06:22, 15 August 2017 (UTC)

Strike does not have appropriate semantics, see HTML5 documentation: The s element is not appropriate when indicating document edits; to mark a span of text as having been removed from a document, use the del element.; and underscore does not have any associated semantics. For accessibility reasons, we should be using <del>...</del> and <ins>...</ins> respectively, see HTML5 documentation. --Redrose64 🌹 (talk) 14:38, 15 August 2017 (UTC)

Version 2

Proposal:

"Generally you should not break up another editor's comment to reply to individual points. Mixing comments like this confuses the layout of the page and obscures the original editor's intent, as well as potentially leaving text unsigned. Instead, place your reply entirely below the original comment. You may wish to use the {{Talk quotation}} template to quote a portion of the material in question."

1) Replaces "interleaving" with "mixing" per NewsAndEventsGuy's concern. 2) Adds a brief description of what to do in addition to what not to do. 3) A question from Andrewa is whether this should start with "Generally you should not" or stronger wording like "You should not". I thought the stronger wording may not cover every possible situation which is why I started with "Generally". 4) I felt that only a brief mention of the quote template that is most often used would be best and we should avoid putting a detailed style guide for replying in the "Editing others' comments section", but conceivably we could start a new section about quoting/replying. It could for example cover numbering or lettering points (if that's not obvious) and {{Talkquote}} (a different template from {{Talk quotation}}). That is more than I wanted to get into originally and even if that were added I think a statement in the "Editing others' comments section" against splitting another editor's post would still be appropriate. —DIYeditor (talk) 03:09, 15 August 2017 (UTC)

It's rearranging deckchairs. The more we look at the guideline the worse it gets, see #Modifying comments already replied to. Total rewrite required, incorporating the "add one of anything to the right" suggestion for more sophisticated users, and a far simpler protocol for beginners, example-based. And I still think that a brief interspersed comment is helpful on occasions, but I will of course go with the consensus on this. Andrewa (talk) 04:19, 15 August 2017 (UTC)

There may be some issues with this page as far as providing enough detail but I think the main places to address that are Help:Using talk pages and Wikipedia:Indentation (and any other topic specific locations). Providing examples and detailed style guides here would totally change the nature of the page. It is supposed to be dos and don'ts more than detailed instructions. We can make sure the reader is pointed in the right direction for more information. —DIYeditor (talk) 06:00, 15 August 2017 (UTC)

This proposal is a very welcome clarification. But I'm concerned that it is sufficiently severe in its impact as to require wider discussion. As it stands it seeks to ban or at least discourage what is a very common and IMO clear and helpful convention, one that is long in use far beyond Wikipedia. But this is a convention that I acknowledge is poorly documented on Wikipedia and often ignored here, leading to some very messy talk pages. Andrewa (talk) 17:30, 16 August 2017 (UTC)

Anchors (Version 3?)

@Andrewa, DIYeditor, NewsAndEventsGuy, Tvx1, Redrose64, and Johnuniq: In replying to or commenting on points in preceding comments, I have occasionally inserted anchors in those comments and linked to them. This way there is no need to quote in full the point I am responding to. More than that, when the discussion is already long and complex, with many replies-to-replies-to-replies-..., a reader can find the point being replied to without a potentially long and distracting search: for example, aXXXX this reference to Andrewa's comment about the "dreadful mess" that can result from mixing asterisk and colon listing. And unlike interleaving, anchors do not affect the display at all, but are visible only in edit mode.

I guess I'm making a proposal for a better (imho) way of replying to specific comments without interleaving or necessarily quoting, so I'm adding "Version 3" to my section header. I'm not ready to turn this into a guideline proposal, so I invite you all to please have a go at it (thus the question mark).

Please {{Ping}} me to discuss. --Thnidu (talk) 18:23, 27 August 2017 (UTC)

Thnidu, I have never used anchors in that way but can't see any reason not to. I instead use a diff to refer to the comment to which I am replying, but generally quote the relevant text in italics as well.

There are many acceptable ways of structuring a discussion.

I have been involved in Internet discussions since before public ISPs were available in Australia (we used permanent dialups to form APANA that's redlink and shouldn't be, see http://www.apana.org.au/ I see it still exists, and before that there was FidoNet which also still exists of course), and was frankly astounded that the interleaving that provoked this discussion caused anyone any stress or confusion at all. My belief was (and is) that this convention is still the most common and easily followed method of structuring a complex discussion on the Internet generally. Many if not most email clients provide it automatically.

But some do have problems with it obviously. So the questions are (1) is there a better way and (2) can and should our guidelines be improved (one way or the other depending on the answer to (1)).

I have had experience before with people objecting to this convention, but previously it has always been in the context of the low-level disruption I call ranting... for example, some users will punctuate a long post with p HTML tags or with no paragraph breaks at all. Either tactic prevents interleaving, and in my opinion should be discouraged for exactly that reason. Andrewa (talk) 22:48, 27 August 2017 (UTC)

You would not be astounded that many editors find interleaving to be disruptive if you had observed discussions where they were common, and where it was necessary to refute the interleaved comments. If people cannot make their point in a digestible manner they should not comment. A comment has to be made in a way that replies to the comment could reasonably occur. Further, discussions are not just for the benefit of the current participants; future editors may need to review old discussions to see why certain conclusions were reached. Johnuniq (talk) 23:06, 27 August 2017 (UTC)

Strongly agree with most of this.

The attempt to personalise this discussion (read the link, and discuss the contribution not the contributor please) is just plain ridiculous, as you would know had you bothered to do any check of my edit history. Enough of that please.

And I'm afraid I remain astounded. Refutation of the interleaved comments is exactly what the convention makes easy and transparent, and easy for others to follow later. Of course there comes a point where indentation is excessive, but for the first two or three indents it works very well. If it goes beyond that, probably best to start a new subsection, IMO... or outdent sometimes works well, sometimes not.

But for the rest, good points all, and I think they support the proper use of interleaved comments in Wikipedia, for the same reasons as it is standard practice elsewhere. Andrewa (talk) 23:29, 28 August 2017 (UTC)

@Andrewa and Johnuniq: How do you even find relevant comments in a long and heavily interleaved discussion? Especially if they're not “addressed” with {{ping}} (or one of its numerous aliases), or with simple linked mention as in {{u|Susannah Q. User}}.

In a long discussion, it is hard work to follow the threads at the best of times, and from time to time users even use this to advantage. But is there any doubt what I'm replying to here? Does it make the above post look unsigned, or this one? I don't think so. But then I'm an old hand at this, since long before Wikipedia.

So I propose to add a brief description of this anchoring method, explicitly stating that this is not a guideline but an available alternative to interleaving. If there are no strong objections I will do so. Please {{Ping}} me to discuss. --Thnidu (talk) 18:30, 30 August 2017 (UTC)

I have no objection at all, provided it is not claimed that this is a preferred method. But I still have doubts that interleaving should be in any way discouraged. See this two-part reply as an example of a case in which I think it works well. How would you make the logic clearer? Is it necessary to do so?

But we do I think need to update the guidelines to give some help to new hands who have not seen it done previously. In particular, it's not standard practice on mobile devices AFAIK. There may even be an argument to discourage it for the benefit of mobile users, I'm not one so I would not know.

On conventional web browsers it works well IMO, if properly done. Here we're now five levels deep (perhaps we should recommend a limit to the depth of indenting), so pushing the limits, but it still works well on my browser.

Thnidu, pinging as requested. But I'm surprised that is necessary... do you use watchlist and contributions? Andrewa (talk) 22:15, 30 August 2017 (UTC)

Sorry, Andrewa. Yes, I do use them. The trouble is that on some pages that have many unrelated discussions going simultaneously, like the Teahouse, I get too many notifications about edits on topics I'm not interested in. This page is not such, but the habit stuck with me. Also, as I believe I mentioned above, navigating such a long discussion as this on a smartphone creates other difficulties. --Thnidu (talk) 23:02, 30 August 2017 (UTC)

Making the articles available to mobile users is definitely a good thing. I am yet to be convinced that mobile editing of articles or discussions is a good thing overall. It has obvious advantages but there seem to be some drawbacks to it. Andrewa (talk) 02:47, 31 August 2017 (UTC)

@Andrewa: Edits like this are precisely the problem we are trying to avoid. Without knowledge of that specific edit diff, can anybody tell from the above thread that the paragraph beginning "In a long discussion, it is hard work to follow the threads at the best of times" was not written by Thnidu (talk · contribs)? It's misattribution, plain and simple. --Redrose64 🌹 (talk) 08:25, 31 August 2017 (UTC)

QED. EEng 14:08, 31 August 2017 (UTC)

This is the heart of the matter, and apologies if I offended anyone by this post, but I did so as an example, and it's proving to be a good one.

Yes, I think that the indentation makes it quite obvious that the paragraph in question was written by me and not by Thnidu. I can't see how anyone can miss it, in fact. But obviously you have difficulty following the thread, so we need to do something. If there's consensus that interleaving is to be discouraged, then of course I'll abide by that decision. But I think it's the wrong way to go.

Strongly disagree that it is misattribution. That is over the top. There is no intent to mislead, and the convention I'm using is clear and unambiguous. The problem is just that some people apparently have difficulty in following it. Andrewa (talk) 14:18, 31 August 2017 (UTC)

(Thnidu here.) Between the 2 paragraphs of one of my comments, andrewa inserted a paragraph including the question

But is there any doubt what I'm replying to here? Does it make the above post look unsigned, or this one? I don't think so. But then I'm an old hand at this, since long before Wikipedia.

Redrose64 responded

Without knowledge of that specific edit diff, can anybody tell from the above thread that the paragraph beginning "In a long discussion, it is hard work to follow the threads at the best of times" was not written by Thnidu (talk · contribs)?

To which I will add: Yes, as a matter of fact, it does leave the preceding paragraph (not "post", since your [​andrewa​] paragraph interrupted my post) not just "look" unsigned but be unsigned, since your interposition separated that paragraph from my signature. And your paragraph there is also unsigned, since the reader must scroll five paragraphs down to find your signature. You could have avoided the latter problem by typing four tildes after your interruption, but the "un-signing" of my first paragraph would be much harder to fix, if at all doable. And we certainly couldn't rely on new users, who often neglect to sign their own posts, to handle such complications.

As you say, you're an old hand at this, and that's part of the reason for our differences here. In such conversations finding the correct attributions is not a simple task at all. To make an analogy, being an experienced driver does not qualify one to teach driving, and one reason is that there are so many actions that by now are reflexive and unconscious to the "old hand" that they need to learn that the novice needs to consciously learn the stimuli (e.g., car a short distance in front suddenly hits the brakes) and responses (brake immediately but not hard at first, while checking side view, mirror and corner-of-eye direct, to see if it's safe to swerve that way; if not, check other side while braking harder). --Thnidu (talk) 16:24, 31 August 2017 (UTC)

You say that Yes, as a matter of fact, it does leave the preceding paragraph (not "post", since your... paragraph interrupted my post) not just "look" unsigned but be unsigned, since your interposition separated that paragraph from my signature. And your paragraph there is also unsigned, since the reader must scroll five paragraphs down to find your signature.

That is true if but only if we ignore the indenting and interleaving convention, correct?

Or conversely, if we do not ignore the convention, that statement is quite simply false, is it not? The signatures are intact provided the convention is understood to apply here.

Agree with many of the points made in that post. But some of them are splitting hairs and ignoring the issue. We have had this convention for many years. You (and others) want to change it. That's the issue. And there may be a case.

But it seems to me that it would be much easier to discuss this and seek consensus on this if we followed the convention for now. I am refraining from doing so, reluctantly but at your implicit request. It seems to me for example that this series of edits let to an impenetrable mess and the points you make there could have been far better presented, and more easily answered, by using the indenting convention with interleaving.

I have problems with this recent edit (not by you) too, but perhaps that's enough for now. Andrewa (talk) 00:14, 1 September 2017 (UTC)

How many experienced editors have disagreed with the views you have expressed on this page, and how many have agreed (put me down in the former group)? Your "recent edit" link shows Redrose64 reverting a change to their comment—why would you "have problems" with basic common sense? That is indeed enough for now, and actually it is enough forever at Wikipedia. Please do not refactor other people's comments to suit your style, and definitely do not break-up other people's comments with interleaving that the community has rejected. Johnuniq (talk) 03:41, 1 September 2017 (UTC)

@Andrewa: You have problems with my recent edit to my slightly-less recent post when all I did was restore my intended version to how it had been as I had left it in my immediately-previous edit to this page? Get out of here.

There is an indenting convention, but occasionally people will use one symbol (colon or otherwise) too many (or one too few), perhaps as a simple typo. Sometimes, in a post having three (or more) indented paragraphs they will indent one of the intermediate paragraphs to one level deeper than the rest, again perhaps it's a typo, or perhaps it's to emphasise it. Maybe they want to indicate that it has been copied from elsewhere: not everybody uses (or is aware of) tags like <blockquote>...</blockquote> or templates like {{tq}}. It might be an example of proposed wording for some guideline or other, there are at lease three such instances on this page alone. So the extra indent level of one paragraph will not necessarily indicate that the particular paragraph was added by somebody other than the person who added the ones above and below. --Redrose64 🌹 (talk) 10:15, 1 September 2017 (UTC)

Agree that when the convention is not followed it causes problems. Disagree that this is a problem with the convention. Andrewa (talk) 15:21, 1 September 2017 (UTC)

In fact this edit above may be a classical example. It appears by the content to be replying to me, but by the indentation it appears to be replying to the post immediately above it. Probably it is simply indented one level too many. But best not to fix it now that I've replied to it.

That's not necessarily the fault of the convention. But perhaps the convention can be made clearer (either by simplifying it or documenting it more clearly or both) so that such mistakes can be reduced. Andrewa (talk) 02:44, 3 September 2017 (UTC)

@Andrewa: You're darned right it was a reply to you. It should have been obvious from my use of the phrase "You have problems with my recent edit" which was a straight person-reversal of your phrase "I have problems with this recent edit", and since that was an edit that I had made, I am clearly the person involved. Don't claim that you didn't know that.

Now, it is clear to several people here that you are getting tedious: therefore, it is time for you to WP:DROPTHESTICK. --Redrose64 🌹 (talk) 09:07, 3 September 2017 (UTC)

Not quite sure what you're getting at in the first paragraph. Yes, I understood it as a reply to me, and my post did refer to your edit, explicitly, I can't see where I claimed not to know that. It's just the indenting that is wrong, and I wondered whether that might even be deliberate, to prove a point. Was it? (Redundant signature added to allow interleaving.) Andrewa (talk) 19:55, 3 September 2017 (UTC)

As to the second paragraph, see #Where to 2. Andrewa (talk) 19:55, 3 September 2017 (UTC)

@Andrewa: I'm fed up with trying to explain things to you. You seem quite unable to link a reply back to its question. Or are you being deliberately difficult, pretending not to understand in the hope that I will switch to using interleaved replies? It ain't gonna work, feller. --Redrose64 🌹 (talk) 22:15, 3 September 2017 (UTC)

I'm certainly not trying to be disruptive, and if you think I am then the place to raise it is on my user talk page, not here.

And I'm finding it a bit frustrating too. You seem to see a problem with this edit. But what exactly is the problem?

Yes, I did add a redundant signature as an experiment, to see whether you or anyone else would like to try interleaving. Was there any harm in that? And, with this redundant signature in place, would there have been any harm in interleaving? I'm not really surprised you didn't, but I think it would have done you no harm to try it.

But I'm not being deliberately difficult, pretending not to understand. I'm assuming good faith on your part and think you should do the same for me.

And I still do not understand what you're getting at in the first paragraph, and you seem to have made no attempt to explain. You said You're darned right it was a reply to you. It should have been obvious from my use of the phrase "You have problems with my recent edit" which was a straight person-reversal of your phrase "I have problems with this recent edit", and since that was an edit that I had made, I am clearly the person involved. Don't claim that you didn't know that. In reverse order, I did know that, and made no claim not to. Yes, it was clearly an edit you had made, and it was your reply to me, and I said that too, and so your indenting was incorrect. Wasn't it? We all make mistakes. Let's get back to the issues. Andrewa (talk) 01:53, 4 September 2017 (UTC)

Four subsections, four sub-subsections, and 70K have been invested in this. Care to go for 70K more? EEng 02:13, 4 September 2017 (UTC)

Current guidelines

From above: Please do not refactor other people's comments to suit your style, and definitely do not break-up other people's comments with interleaving that the community has rejected. [2]

I have no intention of doing any of that, I still think that a brief interspersed comment is helpful on occasions, but I will of course go with the consensus on this. [3] But there are several suppositions there that I want to question.

The main one is that as far as I can see the guidelines do not currently ban interleaving, so I think it's over the top to claim the the community has rejected it. But agree that the editors involved in this discussion, apart from myself, are strongly of the opinion that it should be banned completely. I suggest therefore that an RfC should be raised with a specific proposal to do so. I think this discussion has probably gone as far as it can.

I will probably oppose this RfC, depending exactly what it says and what arguments are advanced. But if it succeeds then as said above I'll abide by it of course. I may find it difficult to walk the line between allowing others to ignore this new consensus and being pointy in enforcing a ruling I don't agree with! But cross that bridge when we come to it. Hopefully we can reach a strong consensus, that will help a lot. Andrewa (talk) 20:19, 1 September 2017 (UTC)

How about if you stop wasting everyone's time and drop this? EEng 20:23, 1 September 2017 (UTC)

Agree this has gone about as far as it can. Disagree it was a waste of time, and there are still a few things to tidy up, see below. Andrewa (talk) 17:07, 2 September 2017 (UTC)

I was wrong about it having gone about as far as it can, you have now significantly changed or clarified your position below as to the changes you would like to current guidelines, and perhaps we can also make progress with the question of what they currently say, see #Where to 2 below. Andrewa (talk) 19:28, 3 September 2017 (UTC)

Where to now

There's obviously no desire above to test consensus on this at an RfC. There's a clear majority wanting to ban interleaving completely, but this is not the same thing as a consensus. It's not a head count, despite a recent edit summary suggesting one.

The argument seems to be that they just don't like it. Claims that this well-established convention misrepresents the original post and/or violates existing guidelines are unsubstantiated, IMO.

But obviously, these editors can boldly change the guideline and under the 3RR I can't singlehandedly stop them... and so I won't even revert once. I may support others who do, and cross that bridge when we come to it.

It would be good to clarify the guideline, but obviously I don't think that we should change it to ban interleaving. Rather it should be explicitly permitted, and some restrictions considered to control its abuse, and the abuse of indenting in general, particularly to accommodate mobile users. That's just my view of course.

I will try to avoid interleaving comments in posts made by editors who object to the practice, as I have above (with one exception deliberately used as an example). Perhaps we should set up some sort of register, either opt-in or opt-out, so that editors like me who like the practice can use it without offending those who don't.

Apart from that we may be finished here. If so we can let the discussion archive in due course. Thanks to all who have contributed. Andrewa (talk) 17:07, 2 September 2017 (UTC)

Do not avoid interleaving comments in posts made by editors who object to the practice. Just do not do it, period. If you want, create an infoboxuserbox that only you will use: "This user doesn't mind if you interlard you comments inside his posts." EEng 17:13, 2 September 2017 (UTC)

I was wrong, it seems there may be a lot left to say.

Just do not do it, period. I'm afraid that's a ridiculously sweeping request IMO. For example, if someone interleaves their comments with mine, I think I should feel free to follow their lead, and even that it would be a bit rude not to. Isn't that fair enough? I really think you are overstating both the problem and the solution.

The userbox (you said infobox but I think that's what you mean) is a good idea, although many of us already feel we have too many userboxes and there may be better ways of achieving the same goal.

And yes, if I'm the only one who ever uses it, then you'll have made a point.

An "opt-out" userbox (or something that better achieves the same) that says something along the lines of This user does not use the indenting convention to intersperse their comments between paragraphs of other users' talk page posts, and requests others not to do so too would also be a good idea IMO. Andrewa (talk) 02:11, 3 September 2017 (UTC)

No one should have to opt out to keep others from intruding on the integrity of their posts. If you and some other editor are in a discussion and agree to interleave your comments for some special reason, knock yourself out; I obviously didn't mean to restrict the right of consenting adults to indulge in whatever perverse personal behavior together they want, as long as children aren't exposed to it and you don't frighten the horses. EEng 02:40, 3 September 2017 (UTC)

I obviously didn't mean to restrict the right of consenting adults to indulge in whatever perverse personal behavior together they want... That's a welcome clarification of what you actually meant by Just do not do it, period.

And this is progress. As on all talk pages, we are working towards consensus. So can I ask, do we have consensus that a blanket ban on interleaved discussions between consenting adults would be, as I said above, ridiculous? I accept that's not what you meant to say, but it seemed to be to me what you said, and what you and others have suggested above. So it's a very significant step IMO.

No one should have to opt out to keep others from intruding on the integrity of their posts. Agree. But I do not believe that the indenting and interleaving convention is quite that bad. Like your Just do not do it, period that is over the top. At worst, it makes the discussion hard to follow to those who are (for whatever reason) not comfortable using the convention.

And they should be considered, and I've said I'll avoid using the convention to reply to these people, and I am doing so. But give us a break. Wouldn't it be good to give some warning that you're one of these people? Andrewa (talk) 03:54, 3 September 2017 (UTC)

I'm partial to the out-of-the-ordinary!

Sorry, I have to feed the cat, change the water in the fish tank, and brush up on my differential equations before the dominatrix gets here. I hope you work something out. EEng 04:14, 3 September 2017 (UTC)

Please could people observe WP:INDENTGAP, at least. This is an accessibility issue. --Redrose64 🌹 (talk) 09:07, 3 September 2017 (UTC)

Very good point, and I have been guilty of forgetting that from time to time. It would be good to incorporate more on accessibility into the talk page guideline IMO, MOS:Accessibility is not often cited and I'm guessing not often read. Andrewa (talk) 11:06, 3 September 2017 (UTC)

I don't see how INDENTGAP applies; I don't notice anyone trying to leave blank lines. —DIYeditor (talk) 20:19, 3 September 2017 (UTC)

I took it simply as a pre-emptive heads-up... and a useful one. I hope I haven't blundered in that way here, but I have in the past, partly at least for reasons given in #Accessibility below. Andrewa (talk) 20:30, 3 September 2017 (UTC)

@DIYeditor: See this post; blank lines between every paragraph. @Andrewa: No use hoping: the evidence clearly shows that it was you who blundered. --Redrose64 🌹 (talk) 21:54, 3 September 2017 (UTC)

@Redrose64: Ah, thanks. I actually didn't understand the accessibility reasoning or details of INDENTGAP; I had not read it carefully. —DIYeditor (talk) 22:05, 3 September 2017 (UTC)

That was indeed a mistake on my part, and refactoring to remove it was entirely in order. And I have made that mistake before, for reasons set out in #Accessibility. I do make mistakes. I try to learn from them. My definition of an expert is someone who has already made most of their mistakes. That's part of the reason Wikipedia encourages mistakes. Andrewa (talk) 02:07, 4 September 2017 (UTC)

Accessibility

I've just realised that there's a trap with indenting... at the first level (no indenting), you do need to leave white space. Just going to a new line doesn't make a new paragraph.

To do that, you need to leave a blank line. Andrewa (talk) 20:19, 3 September 2017 (UTC)

On the other hand, once you indent, you must not leave white space, as it confuses accessibility aids.

And it's not necessary. Just going to a new line does give a new paragraph once you are at first indent.

Perhaps this should be more explicitly stated, somewhere? Or is it already and I've missed it? Andrewa (talk) 20:19, 3 September 2017 (UTC)

No, once you're at the first : level, just going to a new line starting with : does not give a new paragraph. The right way to break paragraph once you're in the : regime is

:Blah blah 1st paragraph

:

:Blah blah 2nd paragraph

EEng 22:21, 3 September 2017 (UTC)

I see... what I described above seems to work on my browser, but perhaps not on others. I think this is important... is it documented anywhere in guidelines etc? Andrewa (talk) 23:18, 3 September 2017 (UTC)

No, if you try it with and without the intervening extra line with : alone, you'll see a definite difference. EEng 06:32, 4 September 2017 (UTC)

I take it from that that as far as you know this is not documented anywhere.

I've set up a little experiment in the community sandbox. I see five paragraphs, with no significant difference between the spacing of paragraphs three, four and five. What do you see? Andrewa (talk) 00:09, 5 September 2017 (UTC)

@EEng: You may be thinking of how vertical spacing used to work. However, MediaWiki:Common.css was changed in December 2016 (diff) so indented and unidented paragraphs on talk pages have the same vertical spacing. The issue of correct indenting is part of MOS:INDENTGAP where the colon on an otherwise blank line is necessary so the result is a single definition list for screen readers. Johnuniq (talk) 04:07, 5 September 2017 (UTC)

Crikey, you're right! And I was just getting the hang of remembering to include the otherwise-blank-line-with-only-colon-on-it, and here they go and make it unnecessary. My apologies to all for sowing confusion. EEng 04:21, 5 September 2017 (UTC)

Where to 2

You're like a dog with a bone – except it's not a bone, it's a stick. Drop the stick.

–EEng 19:46, 3 September 2017 (UTC)

As I have said above, the goal of talk pages is to work towards consensus. There is some progress above on this.

In particular, one editor has changed or clarified their position, as I see it from wanting a complete ban on interleaving to being happy for consenting parties to use it. They still want the default to be not to use it, and perhaps that can work, I'd need to see a definite proposal and I'm a bit sceptical and wary of instruction creep. But happy to work on it.

The other bone of contention is the assertion above that guidelines already ban interleaving. I think that's also over the top, but hope we can similarly resolve that one too. Andrewa (talk) 19:37, 3 September 2017 (UTC)

You're like a dog with a bone, except it's not a bone, it's a stick. (Note that the post concerned was added after my comment above but above it... probably not a good idea, but again don't change it now I've replied.) Nothing will be decided here IMO, it's a big enough issue to require an RfC to change it, and there's no response to my suggestion above that one might be raised.

So feel free to just drop out if you feel it's a waste of time, and allow the discussion to be archived... as I also suggested once above. But IMO, there is progress. Painful progress, but this was never going to be easy. Andrewa (talk) 20:08, 3 September 2017 (UTC)

Sorry, I didn't see that you were continuing with this. I did take the discussion as consensus against interleaving so I added guidance about it to the page. I don't personally see how an RfC is necessary. It doesn't seem likely that anyone is going to support the "convention" of editing others' comments to break them up for replies. No one else has expressed support for that and several have opposed (sometimes strongly). Feel free of course to start an RfC if you think that there is some support out there for this. —DIYeditor (talk) 20:27, 3 September 2017 (UTC)

Aha, I missed this edit. A bit bold perhaps but a perfectly acceptable action IMO. It has certainly been discussed here, and your edit avoids the excesses that others have (in my view at least) implied.

The next step then is to see how this affects discussions... will people even notice it, will it help or hinder discussions. And I may add something to my sig encouraging interleaving in replying to my own posts. Andrewa (talk) 20:41, 3 September 2017 (UTC)

So feel free to just drop out if you feel it's a waste of time – Some of us have this page on our watchlists in case a productive discussion arises, and we'd prefer not to have our attention repeatedly diverted by this discursive trip to nowhere. EEng 22:25, 3 September 2017 (UTC)

I think that's another example of incorrect indenting. It replies to this post, so a single colon would have been correct. Or am I missing something? Andrewa (talk) 02:18, 4 September 2017 (UTC)

In all seriousness, you're an admin? EEng 02:21, 4 September 2017 (UTC)

Yes. And a lapsed member of Australian Mensa, and a drummer, and lots of other things.

I guess the reason you ask is that you find by behaviour here below par. And it's true that admins are held to a higher standard of behaviour, even when not using any of our sysop powers. But the rules are the same. We're just expected to follow them a bit more strictly and knowledgeably.

And one of the rules is comment on content, not on the contributor. This is not the place to discuss behavior. TIA Andrewa (talk) 22:50, 4 September 2017 (UTC)

Please tell me the bit about Mensa is intentional self-parody. EEng 04:35, 5 September 2017 (UTC)

It is quite accurate. And you are on really thin ice. Please cease the personal comments here. If you have complaints about my behaviour, make them in the appropriate place and fashion. If you think I'm stupid, tough. Get over it. Andrewa (talk) 06:54, 5 September 2017 (UTC)

I wouldn't dream of doubting you belong to Mensa. It's the offhand name drop I'm struggling to understand. EEng 12:42, 5 September 2017 (UTC)

OK, it was a reply to this personal attack which didn't seem worth raising on your user talk page. But strictly we should take it up there, and can if you wish to discuss it further. Andrewa (talk) 01:47, 7 September 2017 (UTC)

Would it be a personal attack to say that if you think the post you just linked is a personal attack, there's something wrong with you? What in the world are you talking about? I think it would be best if you stopped trying to police others' behavior, because you seem incapable of interpreting normal human interactions. This came back onto my watchlist because I answered a ping, but I'm unwatching again; I hope when I return you'll have found some useful way to occupy yourself. EEng 04:14, 7 September 2017 (UTC)

This is a bad, BAD idea

Many users find this practice highly offensive, and it makes discussions very confusing to follow. We should not be encouraging it in any way. Drive a stake through this. EEng 22:37, 30 August 2017 (UTC)

If it's true that Many users find this practice highly offensive, then I'd have to agree it should be banned. End of story.

But I must ask why is it so offensive? And why is it confusing? Was anyone confused or offended by my two-part post above? Why and how?

Walls of text are one of the key techniques of rantstyle. The ability to reply concisely and clearly to a long post is vital in order to get ranting discussions back on track, and work towards consensus.

If people find this offensive, I think perhaps they're in the wrong place. In a sense even your signed contributions don't belong to you here at Wikipedia. All text is available for reuse and refactoring. There are restrictions, of course, and we should not for example misrepresent others, or deny them their chance of a fair hearing, and the guidelines seek to ensure this. But interleaving, properly done, does neither of those things.

To the contrary, it allows arguments to be easily, clearly and concisely answered. And this is not always welcome! Andrewa (talk) 00:45, 31 August 2017 (UTC)

All text is available for reuse and refactoring – Reuse, yes. "Refactoring" that in any way even slightly tampers with the context, import, or connotation of an editor's post, no, and that includes interlarding your own comments in a way that neuters the original thrust. Quote a bit of what someone said, and respond to it – as I did in this very post. More than one person may want to respond to the same post, and if they all try to interleave it becomes a complete mess. EEng 00:56, 31 August 2017 (UTC)

Agree with all of the first two sentences except the unstated implication that interleaving properly done even slightly tampers with the context, import, or connotation of an editor's post or that it neuters the original thrust.

And if a second editor inserted a comment here, it would not affect the flow or logic of my post in the slightest. Andrewa (talk) 02:26, 31 August 2017 (UTC)

It does neither. Did my example above do either? It doesn't seem to have to me. Do you have examples that have? Andrewa (talk) 02:26, 31 August 2017 (UTC)

I'll tell you why I hate it. It derives from an email/usenet practice, where correspondents inserted their comments directly into the comments of the person they responded to. In that context, it works fine, for two reasons.

First, in the email/usenet context, each response was a separate document, an email message itself. It wasn't a single document being iteratively edited, with the end result being an intermingled mass of commentary by multiple editors, where finding each editor's comments and viewpoints is much more difficult.

Second, in email/usenet, one almost always trimmed away the parts not being responded to. It was a matter both of courtesy and to keep the note from becoming ungainlily long. In contrast, on Wikipedia talk pages, we obviously don't want other editors' comments trimmed by the act of responding; again, because it's a single document, not a series of individual documents.

What worked very well for email and usenet works very badly in an interatively edited document like a Wikipedia talk page. It causes attributions to be masked or difficult to figure out.

You say "it allows arguments to be easily, clearly and concisely answered", but I don't think that's the case. Easily and concisely, sure; but clarity is a casualty of the practice. I love it in email (top-posting is the bane of modern email communication) but hate it for talk pages. TJRC (talk) 00:59, 31 August 2017 (UTC)

It seems to be a matter of taste to me. I have no problem following properly inserted comments; I find it easy to follow the logic (or lack of same) in arguments presented in this way. Again, have you any examples where clarity has suffered? Do you think that the example I gave above, or the original one that started this whole string, are unclear?

I'd like to respond to the detailed points you make, some of which I agree with, but others are I think at least questionable. But without doing what you hate I think it would be unworkable, so I'll just leave it at that. Andrewa (talk) 02:26, 31 August 2017 (UTC)

andrewa, in reply to your first ¶ above (I refuse to interleave), my comment in the previous section about the "old hand" applies here as well. As a extremely experienced editor, you are unable, without effort, to comprehend how what is so easy for you can be difficult to a newcomer. --Thnidu (talk) 16:37, 31 August 2017 (UTC)

That's certainly not what I meant to say. I am unable to understand how you and other old hands have such trouble with it.

I admit I haven't checked what I said, and I don't find your references terribly helpful in doing so. Interested in other views on this. But I suspect you are, perhaps unintentionally, misquoting me.

There is a learning curve for newcomers, but the convention is easily learned and so useful in many situations that it should not be generally discouraged.

There should be guidelines discouraging its excessive use. But the proposed ban on using it even one level deep seems ridiculous to me. Andrewa (talk) 00:31, 1 September 2017 (UTC)

I think "generally" is a good compromise wording. It acknowledges that there may be circumstances where it is acceptable, although good examples elude me aside from by mutual explicit consent. Even then to me it is a bad idea. It's messy and confusing. —DIYeditor (talk) 20:33, 3 September 2017 (UTC)

Agree it's a good wording. Still disagree with the principle. I could be wrong. So let's see how it works in practice. Andrewa (talk) 23:24, 3 September 2017 (UTC)

WP:MOS is a good guideline where "generally" is excellent advice for two reasons. First, there are lots of exceptions in language, and one set of rules is unlikely to cover all cases. Second, the consequences of conflicting with MOS are small—the point of disagreement would often not be noticed by most readers. However, comments at a busy noticeboard (or an article talk page where conflict applies) are different. First, experience in such places shows that interleaving is a ghastly procedure—it makes it hard for those who need to follow the discussion (perhaps in a year) to see who-said-what and why a decision was made. Second, while there are a few ILIKEIT votes on this page, there are at least as many strong objections. Talk pages are where collaboration is tested, and procedures that annoy a significant portion of the participants are an immediate fail. The people here might understand that generally means "almost never", but those who like interleaving will use generally as a green light. If challenged they will reply that WP:TPG permits the procedure and their convenience required interleaving. Using generally will end up encouraging interleaving, and that would need a major RfC. Johnuniq (talk) 04:26, 4 September 2017 (UTC)

I've just gotten involved here and I'm not entirely sure what is meant by "interleaving" at this point, but your concern that we should assume bad faith and therefore misuse a guideline to make statements in conflict with fundamental principles (the right to edit and no ownership of text) is concerning.  There are only two rules needed, (1) we are here to build an encyclopedia and (2) don't change the meaning.  I think the second is where you've missed the departure, because it is typical when inserting text in someone else's comment to begin the insert with [insert begins here].  The big problem when inserting text in someone else's comment is when they are newbies (which in this context includes senior admins) who think they own the text.  So yes, it is important that TPG make correct neutral statements that protect good edits from ownership conflict.  Unscintillating (talk) 11:41, 4 September 2017 (UTC)

Might be best to find out what we're talking about before commenting. EEng 13:51, 4 September 2017 (UTC)

That is a refusal to discuss.  Unscintillating (talk) 15:57, 4 September 2017 (UTC)

No, it's a disinclination to humor the comments of someone who starts by declaring he doesn't know what he's talking about, then goes on to talk about it at length. EEng 17:02, 4 September 2017 (UTC)

According to my search, your first edit to WT:TPG was on 27 August 2017.  Is that correct?  I suggest that you and others review Archive 10 and Archive 11.  Also Archive 8 has relevant material.  Unscintillating (talk) 18:27, 4 September 2017 (UTC)

No, it's not correct, and no one's going to plow through three archive pages you vaguely wave your hand at, looking for something unspecified. EEng 19:46, 4 September 2017 (UTC)

@Unscintillating: It would help if you indicated what specific sections you are referring to as the archives are long. Luckily "interleaving" returns Archive 10 and 11 in a search so I understood that you meant they explicitly used the term. What section in Archive 8 are you referring to? Wikipedia talk:Talk page guidelines/Archive 10#Reverting interruptions seemed to conclude that consensus was against interleaving and that the exceptions should not even be mentioned to avoid giving approval for the practice. Wikipedia talk:Talk page guidelines/Archive 11#Posting within another editor's post doesn't have clear consensus. Wikipedia talk:Talk page guidelines/Archive 11#Non talk page reviews and posting style did not seem to reach consensus about whether WP:TPG should cover project pages like WP:FAC where interleaving is common and mention those practices/exceptions. I am happy to leave this with the "generally" wording and even to make mention of specific exceptions (although they aren't exactly for talk pages). —DIYeditor (talk) 20:18, 4 September 2017 (UTC)

Johnuniq makes some good points in their {(longish) single paragraph above but there's also some questionable stuff mixed in. First, experience in such places shows that interleaving is a ghastly procedure—it makes it hard for those who need to follow the discussion (perhaps in a year) to see who-said-what and why a decision was made - similar sweeping statements have been made, including of course in this section head. But we have had no examples of this experience, just strong statements that it's ghastly, a bad, BAD idea, and so on. If this experience is so persuasive, surely it shouldn't be hard to find specific examples? Andrewa (talk) 20:56, 4 September 2017 (UTC)

• Here is an example I posted in Archive 8:

[T]he two priorities are, we are building an encyclopedia, and "don't change the meaning".  User B, or for that matter other editors, can do more by identifying where the inserts begin and end, such as with:

:Point 1 by user:A

::[insert begins here]

::Reply 1 by user:B, sig

::[insert ends here]

:Point 2 by user:A

::[insert begins here]

::Reply 2 by user:B, sig

::[insert ends here]

:Point 3 by user:A, sig

Unscintillating (talk) 03:09, 19 August 2011 (UTC)

Posted by Unscintillating (talk) 22:13, 4 September 2017 (UTC)

Exactly. But some users think that this leaves points 1 and 2 by user A unsigned; The bot has no trouble finding the sig however. On the other extreme, I find the first sig by user B redundant (but harmless), and again so does the bot AFAIK. Andrewa (talk) 22:23, 4 September 2017 (UTC)

@Andrewa: surely it shouldn't be hard to find specific examples I already gave an example - in my post of 08:25, 31 August 2017 (UTC). Clearly, WP:IDHT applies here. --Redrose64 🌹 (talk) 20:27, 5 September 2017 (UTC)

Yes, in this post you referred to an example I had deliberately provided, and which is the subject of ongoing discussion, and I'm keen to discuss it some more. It's proving to be a very good example. Andrewa (talk) 20:52, 5 September 2017 (UTC)

I have taken the behavioural allegation to the user's talk page. [4] Andrewa (talk) 21:09, 5 September 2017 (UTC)

Archives

Thanks for the links to the archives. My immediate reaction is that it's all been said before, that some users object strongly to comments being edited in this way and proposed to ban everyone from doing it even by mutual consent, while others find it useful, and the proposed blanket ban was not supported.

Interested in other views of course.

How can we improve on these previous discussions? I suggested above and suggest again that a way of users flagging whether they object to the practice, either opt-in or opt-out or maybe even both, would be a good thing. Some actual examples of good and bad interleaving would also help, as would a clearer idea of exactly what we mean by interleaving. And, dare I say it, some clear guidelines on its use, covering both both how to use it and when it is permissible.

The devil will be in the detail of when it is permissible. Andrewa (talk) 22:16, 4 September 2017 (UTC)

Interleave defined

The word "interleave" does not appear in the WT:TPG archives until Archive 10 in September 2013.  WT:Talk page guidelines/Archive 11#Non talk page reviews and posting style shows that interspersed and interleaved are synonyms, and that the article Posting style discusses interleaving.  Unscintillating (talk) 21:54, 4 September 2017 (UTC)

This edit has been described as precisely the problem we are trying to avoid. [5] The edit summary of that last diff is also worth repeating I think: can anybody tell from the above thread that the paragraph beginning "In a long discussion, it is hard work to follow the threads at the best of times" was not written by Thnidu?. So there's clearly a problem to address.

I have no problem telling that the paragraph... was not written by Thnidu, less problem in fact than I have in understanding that edit summary! But with a little work even that edit summary is clearly decipherable. So perhaps with some better explanation of how interleaving works, other editors would have less trouble with it. That's not the whole solution, but I think it should be part of it.

I did always intend that edit to be an example, and have apologised to any I offended by it. I think it has proved to be a good example. It has helped me to better see what others are upset about. I think we need more concrete examples, both good and bad. Andrewa (talk) 00:49, 5 September 2017 (UTC)

I have never seen a good example of interleaved comments. Good examples probably exist because Wikipedia is big place and it is conceivable that an article talk page with low traffic might have someone post several questions in one comment, and a reply giving an interspersed answer for every question. If the discussion stops there, it may be ok. However, even in that ideal case there is no easy way for a third-party to see that the second comment did not alter some meaning of the original. Such checking would require careful examination of the details of the second comment. By contrast, if the second comment had been entirely underneath the original, it would be obvious that the original was untouched. The second comment could have the form: "Q1 Here is the answer. Q2 This is another answer. Q3 Etcetera." Often the second person would take half-a-dozen edits to get their comments fixed, and that complicates the integrity issue if others edit the section in the same period.

The real problem of interleaved comments concerns a page which is not low traffic. Consider a situation where person A posts five questions and person B interleaves five answers. At the same time, editor C tries to refute one of B's answers, while other editors (including A) want to add to the interleaved subthreads. The result is a total disaster, particularly in a contentious area where POV pushers and clueless editors are involved. Johnuniq (talk) 04:32, 5 September 2017 (UTC)

Johnuniq, please have a look here for a version of that scenario. How would you improve on that setting out of an already rather complex discussion? Show us! I hope I'm open to new ideas. But this one still seems counterproductive to me.

And everyone else, do you think I've correctly interpreted the convention? Open to other ideas on it. Andrewa (talk) 07:24, 5 September 2017 (UTC)

An artificial example could prove anything. I have never seen a good example of interleaved comments used in a discussion. Wikipedia resembles a massively multiplayer online role-playing game but everything is supposed to assist the development of the encyclopedia. Having fun on a guideline talk page is not useful, and sandboxes with meaningless examples are not helpful. The guideline should not encourage techniques which are known to irritate a significant number of editors. Per IAR, people can do anything that is genuinely useful for a particular situation, but the fact that no one has produced examples showing useful interleaved comments should be enough for anyone to see that the recent turmoil on this talk page is a gigantic waste of time. Johnuniq (talk) 07:44, 5 September 2017 (UTC)

It's your artificial example.

I agree with all the platitudes above. And agree that we should avoid techniques which are known to irritate a significant number of editors as far as possible. I suspect that avoiding interleaving will irritate editors too. We will see.

The problem with finding real examples is that they will mostly be from complex, problematical discussions. How are we to assess how they could have been helped by avoiding interleaving? I'll try to find some.

But we don't have examples of a total disaster produced by interleaving either, although it seems assumed by some that they exist. Surely they should be more easily found if so? Andrewa (talk) 07:58, 5 September 2017 (UTC)

Here is a total disaster due to interleaving. I thought you were an advocate of interleaving and would be able to produce examples of their use from your reasonably recent contributions.

Regarding the merits of breaking-up other people's comments, if I present an argument above my signature, I intend the whole comment to stand as-is. If someone breaks-up my comment they are damaging my case, making it much harder for others to follow. If they copy my signature to the fragments of my comment, they violate WP:SIGFORGE by making it appear that I presented the fragment as a comment, when in fact it is taken out of context. Do not do that. Johnuniq (talk) 09:52, 5 September 2017 (UTC)

Conversely, if others are not allowed to intersperse comments, it encourages wall of text arguments, does it not?

Interleaving is rare in my experience. But in replying to excessively long posts and ones that mix in valid points among others, it is sometimes very useful. These do tend to be discussions which are already problematic for other reasons, so I'm trying to find ones that aren't a mess for these other reasons. And it's difficult to find them, I admit that. Andrewa (talk) 20:32, 5 September 2017 (UTC)

As a screen reader user, sometimes I'm OK with interleaving, but in the example above it got ridiculous ... I couldn't tell that Thnidu hadn't written the comment. Graham87 14:36, 5 September 2017 (UTC)

Thanks for that... just the sort of input we need. How could the interleaving have been achieved in order to make it clear? Andrewa (talk) 20:32, 5 September 2017 (UTC)

I have no idea, to be honest. Graham87 04:59, 6 September 2017 (UTC)

A concrete example

From above: Here is a total disaster due to interleaving. [6]

Very interesting, and I think I begin to understand.

This section and an adjacent section are appalling!

I'm trying to find exactly where interleaving was used, I can't see it on first reading, so a diff would be great. I'm sure I'll find it eventually, and will post it here when I do, but if you have a diff handy, Johnuniq, it might save me and perhaps others a lot of time looking.

But watch this space, there's lots going on in that example, and thanks for it. Andrewa (talk) 21:21, 5 September 2017 (UTC)

LOL, that's the whole point. Say you want to find out why a decision was made around that time. Someone gives you the above link and says that the decision came from that discussion. Now it's your job to figure out. Johnuniq (talk) 22:10, 5 September 2017 (UTC)

There's no doubt as I said above that this is an appalling example of a discussion off the rails. The question I'm asking is, was interleaving really to blame for that?

Or even before that, we could ask was it used at all? I'm assuming that you found a place where it was, and that you could save a lot of time by indicating more precisely where that is. Andrewa (talk) 22:26, 5 September 2017 (UTC)

That discussion was archived to Talk:Shakespeare authorship question/Archive 20#Wadsworth Reference Should Be Deleted For Everything But The History Section. Johnuniq, is there any reason we need to look at the past revision, rather than the archive page? Andrewa (talk) 22:51, 5 September 2017 (UTC)

I can provide links but I cannot do your thinking for you. The history of the archive you just found is this. That shows five edits by MiszaBot. Not very helpful if someone wants to find out who said what as well when and why. The real history is here.

The guideline currently does not encourage interleaving. It is not up to me to find examples of bad interleaving to prove the guideline should not be changed. Anyone wanting to encourage interleaving needs to demonstrate that it is useful. Arguments that twenty years ago people used interleaving in a completely setting are irrelevant. Johnuniq (talk) 23:06, 5 September 2017 (UTC)

Johnuniq is on record in Archive 8 that what is now being called interleaving is not in basic form an issue needing comment, but that adding sub-section signatures is entirely concerning.  I think he elocutes the words of the concern better here.  Unscintillating (talk) 01:37, 6 September 2017 (UTC)

Thanks for the link to the talk page history. I was already moving in that direction, see here, but knowing that the interleaving you found takes place within this 100 edits is very helpful. Can you narrow it down any more than that?

It helps your case to provide an example, and you have done so, a total disaster due to interleaving you said. But if it turns out that you were mistaken in thinking that there was interleaving there, it will have done great damage to your case IMO. If it turns out that there was interleaving but it's not clear whether it was the cause of the disaster, it will be less damaging. It is agreed I think that the case was a disaster.

The history of the archive is of course useless, but the archive has some advantages over an old version of the page for searching.

I'm not necessarily encouraging interleaving. But I do want it permitted in cases where it is helpful. Andrewa (talk) 02:43, 6 September 2017 (UTC)

The relevant part of archive 8 mentioned above is here.

Re narrowing down, a clue is that I am not indefinitely blocked, but that condition applies to the editor below me on the history link I gave. I included my posts in the history because they show ("please put new posts at the bottom of each section") my response at the time. The editor whose name I did not mention is shown on the history page as making five consecutive posts (the five diffs after my two at the top). Try clicking them. Of course it is hard to see what has happened, and that is the whole point of why interleaving is undesirable. That, and the fact that the original comment is broken up and made unreadable.

Re wanting interleaving permitted when it is helpful: anything is permitted when it is helpful per pillar WP:5P5 which refers to policy WP:IAR. What is not helpful is putting wording in a guideline that would encourage interleaving by people who are not good at judging when one of the extremely rare occasions arises where it would be helpful. If WP:TPG said that interleaving was sometimes ok, the person responsible for the total disaster mentioned above would have used this guideline to justify their style of commenting. Johnuniq (talk) 04:54, 6 September 2017 (UTC)

It's an undecipherable mess with tragic consequences, a multibarnstarred editor indeffed (and not the first, unfortunately). And I can understand how, having been involved in it, you'd like to keep the format of talk pages as simple as possible.

The five edits in question seem to be [7] [8] [9] [10] and [11]. Agree that all are problematic, and that this form of interleaving is confusing and should not be permitted. But to argue for a ban on all interleaving from that seems rather a stretch. Andrewa (talk) 03:31, 7 September 2017 (UTC)

Regarding this and this, I definitely think we need to include something about this in the guideline. Interleaving is problematic for reasons noted by others above. And "do not" is stronger wording that I support in this case. Flyer22 Reborn (talk) 23:45, 6 September 2017 (UTC)

Well, I'd certainly support the Do not wording we had [12] if someone proposes it. But remember to ping me since I'm not watching this page anymore. EEng 00:16, 7 September 2017 (UTC)

@EEng: Perhaps: "On pages in the Talk: namespaces, do not... However on some project pages such as WP:FAC where it is accepted practice..." I tend to agree that interleaving is just plain bad but it may be tough justify a "do not" statement without widespread consensus. @Flyer22 Reborn: I am happy with "Generally" and perhaps giving the specific FAC exception. I would otherwise consider starting an RfC (possibly on a fresh subpage here to keep things clean) and advertise it around but I am not the one to do it since as a novice editor I don't have the standing or experience to make claims about what proper talk page practices should be. —DIYeditor (talk) 04:22, 7 September 2017 (UTC)

I still suspect that a ban on interleaving (which is what this is) will cause more problems than it solves. But I'm a lone voice, so I guess we'll need to try it and see. I'll abide by whatever guideline is current... as I've always said... and I'll try not to be pointy in pulling up others who violate it. Andrewa (talk) 07:17, 7 September 2017 (UTC)

Thanks, guys. I just want something on the page stating that it's problematic. By the way, DIYeditor, regarding the pings, pings only work with a new signature. Flyer22 Reborn (talk) 02:43, 8 September 2017 (UTC)

In this example it wasn't the interleaving that was problematic. It was the way the interleaving was done, which is agreed I think to have been unhelpful, combined with other even more problematic edits. Andrewa (talk) 03:58, 9 September 2017 (UTC)

Establishing better expectations in WP:TPG to protect proper inserts

My premise remains that the biggest problem in interleaving comes not in the unusual occasions where it is appropriate, but when editors violate WP:TPO to restore their non-existent "ownership rights" and disrupt discussions as a result of not having their expectations sufficiently set in WP:TPG.  Unscintillating (talk) 01:37, 6 September 2017 (UTC)

Separating portions of a comment from its original context and signature is the real violation of TPO it seems to me, in that it potentially changes the meaning. You believe an editor would be in the wrong to restore their comment to its original form? Now I did just remove three levels of indentation in your post in a new subsection which does not change its meaning. But if I partitioned it out that would - who is to say two distinct sentences or paragraphs weren't meant to provide context for each other? What guidance would you suggest for TPG on this matter? —DIYeditor (talk) 02:34, 6 September 2017 (UTC)

My belief remains that the biggest problem is that people do not understand and follow the indenting convention. And this is not just a problem when interleaving might otherwise be helpful. It creates awful messes even without any attempt at interleaving.

I haven't yet found the interleaving referred to in #A concrete example above, but the links so far provided do lead to some awful ancient trainwrecks. See this diff and don't waste too much time trying to work out what is going on, it's a mess. But note that in the edit being restored, two colons are being used to set out an inline quotation.

If this is normal on that talk page at that time, then those accustomed to it are going to hate interleaving, and find it confusing and even misrepresenting the flow of discussion. The two conventions are not even remotely compatible. So it's starting to make sense to me.

But I say two conventions with some reservations. It seems to me possible that those involved in this particular discussion were making the rules up as they went along. Andrewa (talk) 03:55, 6 September 2017 (UTC)

@Andrewa: Another tangential problem I see with the indenting convention is when, as you did correctly above per the Wikipedia:Indentation essay, a reply is indented only one more level than what is being responded to even if there is an intervening reply from another editor. To avoid confusion in this circumstance I do indent only one more than what I am replying to regardless of who else has responded, but I put a {{Replyto}} template at the start of my reply, because it definitely looks as if multiple comments are combined into one unless you make careful note of signatures (which naturally comes after reading a good portion of text). If I didn't want to ping a user for some reason I might just do "@User:" manually. —DIYeditor (talk) 04:22, 6 September 2017 (UTC)

I don't find that remotely confusing, and think that it's reasonable to expect editors to learn some basic conventions of which this is one. The help page at WP:THREAD has only one very simple example, but links to the essay at Wikipedia:Indentation which has four examples of increasing complexity, and their example #2 describes this situation precisely. That essay has over 1000 incoming links, so it's fairly widely read and quoted and in the absence of anything better should be regarded as the standard IMO.

So there is no ambiguity. But there is obviously confusion.

The conventions need to be made clearer and/or more readily available.

And possibly also modified to allow for mobile computing and page readers, neither of which were considered when I started editing. There has been little change since then, and even what change there has been has not been effectively communicated to some of the participants in this discussion. Andrewa (talk) 01:25, 7 September 2017 (UTC)

And it seems that even those involved in this discussion can't follow the most basic example, so there is much room for improvement. And before when I have pointed out that indenting here was in error, the only response was a personal attack.

I say again, it seems reasonable to me to ask that people learn a few basic conventions. Andrewa (talk) 01:58, 7 September 2017 (UTC)

It doesn't matter if you know one knows the convention or not, I find it hard to believe that you do not initially parse multiple comments at the same indentation level as coming from one person. You're saying the first thing you do when you look at a new indentation level is scan for signatures before you look at the beginning of the sentences to get an idea of what each paragraph is about? Really hard to believe. Especially since 95% or more of the time comments at the same indentation level are from the same person. That flies in the face of the premise that indentation without a preceding signature in the case of interleaving should make it obvious that the indented portion is from someone else - what exactly is the cue if not indentation? And you don't see the problem when that cue is lacking? Believe me, I agree with the logic behind replies to a comment being only one more indentation level over but it trips me up every time there are responses from different people and I can't imagine what process you are using to read comments where it doesn't happen to you. —DIYeditor (talk) 02:09, 7 September 2017 (UTC)

Interesting... No. I have no such problems. But I'm not saying the first thing you do when you look at a new indentation level is scan for signatures before you look at the beginning of the sentences to get an idea of what each paragraph is about at all. As I read through a post, if I come to a sig, I know it's the end of the post, and until I do I assume there is more. It seems to work most of the time! It only fails when correct indenting is not followed... such as in the appalling trainwreck claimed as due to interleaving but in which I'm increasingly doubting that there was any interleaving at all. Maybe I'll still find it. Another pair of eyes would be welcome, see #A concrete example.

I find it equally hard to believe that, reading a post, you would not notice that there was a sig at the end of it, but would continue reading on the assumption that the text following the sig was by the same person just because it's at the same indentation level. Is that what you're saying? It makes no sense at all to me. What am I missing? Andrewa (talk) 02:36, 7 September 2017 (UTC)

Speed reading. I don't think most people read word by word, sentence by sentence, left to right, top to bottom. Not initially or exclusively. Having to look for signatures at the same indentation level disrupts the normal process of skimming and scanning. Especially since almost all the time the same indentation level is the same post. I would say we are getting pretty far off topic except that the only defense of interleaving I've seen is that indention is an expected cue - almost always for a different author. —DIYeditor (talk) 03:09, 7 September 2017 (UTC)

Aha... my initial reaction is, that is your problem... you are using a completely inappropriate technique and it's no wonder you have problems! Speed reading is great in its place but making sense of a complex and/or controversial talk page discussion is definitely not one of them.

But if a significant number of editors do this then yes, I guess we will need to accommodate them, unless we can talk some sense into them. Andrewa (talk) 07:01, 7 September 2017 (UTC)

It now appears that the interleaving in question was five edits [13] [14] [15] [16] and [17] by a now indeffed user, which are indeed interleaving but do not follow the normal conventions, and I agree that they are problematical... using outdent in an interleaved comment is a bit bizarre, and I have never seen it done before that I can remember. Andrewa (talk) 07:28, 7 September 2017 (UTC)

Comment I hate to add to the confusion here, but I'm going to. I understand completely when a discussion is on content, but a discussion on wording of a passage following a GOCE copy-edit is slightly different. I often leave detailed comments after I complete a copy-edit on the talk page of the requester. I always number them and try to include the precise location in the article of the sentence in question, including a link to the section. I often copy a sentence, group of sentences, or short paragraph for easy reference. We then go back and forth, working on that sentence, group of sentences, or paragraph, sometimes proposing slight changes in wording or sentence construction. Each version, and the accompanying remarks, really need to be one after the other to see the differences between them. When we resolve it, we go on to the next item. The discussion is almost always only between myself and the editor who requested the copy-edit. We usually use the colon for progressive indentation, but sometimes I suggest alternatives using either the bullet or (a), (b), and (c). I have copy-edited hundreds of articles, and I haven't had any problem with this method. Whenever someone replies to all my concerns at the bottom of the page, I have to go back to the top to see what the original comment or passage was, then back to the bottom to add a new comment or version. It would take a lot of time and space to repeatedly "quote" previous versions in order to add them to the bottom of the page. It makes things more difficult. I really think that if you implement this, it should be a recommendation, not a requirement.  – Corinne (talk) 17:16, 7 September 2017 (UTC)

Sure, and doing what is best is the whole point of WP:IAR and the fact that this page (WP:TPG) is a guideline and not a bright-line policy. If there are situations where interleaving is useful, of course people should do that. However, interleaving is not useful in most discussions, particularly those on a noticeboard or on a contentious topic where multiple people may want to offer an opiniion. Another example of desirable interleaving is at feature-article discussions such as this long example. Johnuniq (talk) 23:27, 7 September 2017 (UTC)

List of editors who would like to be notified when this discussion is closed and archived so they can add the page back to their watchlists, having removed it for the time being so as not to be annoyed over and over by this pointless waste of time

• EEng 23:33, 5 September 2017 (UTC)

Proper use of the colon

Improper use of the colon.

–EEng

I said above I was beginning to understand. Here's part of that.

In one of the edits I stumbled upon while investigating the claim that #A concrete example had been a total disaster... (which it was) ...caused by interleaving (which it was not), two colons are used to cite the text of a policy. This rang a warning bell.

This to me is misuse of the colon. If we allow it, then the interleaving convention I have followed does become ambiguous and confusing. One or the other does need to be banned.

But there are far better ways to cite someone else's text. Andrewa (talk) 04:20, 9 September 2017 (UTC)

Template:Reflist-talk

I was going to boldly add a paragraph to the page, but because of the warning at the top—

You are editing a page that documents an English Wikipedia guideline. While you may be bold in making minor changes to this page, consider discussing any substantive changes first on the page's talk page.

— I'm posting it here first for discussion:

References on talk pages

If your comment includes references that will create footnotes, use Template:Reflist-talk at the end of your comment section. This will force your references to appear in a box at the end of the section, rather than at the foot of the page as they would in an article. Like this:

References

Comments, please! {{Ping}} me.--Thnidu (talk) 18:57, 27 August 2017 (UTC)

• Honestly, this page is complicated enough as it is. Refs on talk pages are fairly rare (usually they're there by accident – copied in with some other text, and of no importance at all) and it's not the end of the world if they get rendered at the end of the page. Some other more experienced editor might come along and add {talkref}, or not, and either way it's not a big deal. I'd skip it, and let it be something people learn by example. EEng 19:09, 27 August 2017 (UTC)
• @Thnidu: I agree with the proposal. I see this about once a month and while not necessary, I'd like to see the practice formalized. Chris Troutman (talk) 21:26, 27 August 2017 (UTC)
• EEng (Ye gods, I just wasted at least half an hour in your "museums". Fun but dangerous!) Um... As I was saying... Yes, I still think it's a good idea. It can save a lot of scrolling (→ time → spoons), and make it a lot easier to compare the references with the text. I'mma put it in. --Thnidu (talk) 21:53, 27 August 2017 (UTC)

  I'm not saying it's not useful to have it where it applies. I'm saying that I wonder if it's worth adding to the crushing weight of detail on this page, which is one of the first we recommend newbies absorb. Anyone can come behind an initial post and add {reftalk} when they recognize that it would help so I'm saying let a more experienced comes-along-later editor do it -- no harm done by the delay -- instead of adding one more thing a newbie thinks he has to try to remember. EEng 22:02, 27 August 2017 (UTC)

  EEng is correct. It would be nice if every editor could be given a pill that allowed them to manage talk pages, but that's not going to happen, and they certainly will not read this guideline before dumping refs in their comments. Learn-by-example is best for {{reflist-talk}} and the guideline should focus on basics which are much more important. Too much detail makes it impossible to see essentials. Johnuniq (talk) 22:59, 27 August 2017 (UTC)

  You know, I think there's a place for an essay WP:LEARNBYEXAMPLE on erring on the side of relying on learn-by-example instead of stuffing every detail into long intro pages no one can possibly read. The Help: space is a trainwreck because its builders (who, I gather, simply dropped dead of exhaustion one day) couldn't decide whether to make it a set of for-dummies quick-start pages, or a full regurgitation of every consideration and feature, drawn out with numbing examples for each and every point. Favorite examples: WP:How_to_make_dashes, Help:Footnotes, and (my all-time favorite) WP:Picture tutorial. That word tutorial in there was someone's idea of a joke. (Wikipedia:Extended_image_syntax is even more indigestible, but at least it doesn't advertise itself as a tutorial or help page.) EEng 23:23, 27 August 2017 (UTC) P.S. Sorry, I momentarily blocked on the absolutely, positively, worst help page every: Help:Table.

  @EEng: I just saw that you'd deleted my paragraph from the page. I wish you had at least mentioned that action in this discussion. --Thnidu (talk) 18:55, 30 August 2017 (UTC)

I would have if there was anything that needed saying beyond what's in my edit summary. You're expected to keep pages you care about on your watchlist. EEng 20:04, 30 August 2017 (UTC)

• I wonder if there might be a technical solution to this. It is an annoyance, especially when trying to manually archive, collapse or remove something, and you can't find where those refs at the bottom belong. On a crowded talk page in need of archiving, it can be quite difficult to find which section to stick the template in after the fact, so it would be nice if it could be either automatically generated in the first place or added by bot soon after. Beyond the scope of this talk page, I guess, but I thought I'd see if anyone thinks it's feasible before finding a place to request it. Rivertorch^FIRE_WATER 05:19, 28 August 2017 (UTC)

  I see no hope of some automated solution. But I really feel this is a search for a solution which, when found, will then be in search of a problem. Sure, it's cleaner if each talk thread ends with its own refs, but if they refs end up at the bottom of the page, so what? They're still there, and when a thread is archived the refs move properly with the thread itself to the archive page, appearing at the bottom there. Sometimes the refs are in the thread accidentally anyway e.g. got copied in as part of some text under discussion, and no one cares where they appear or indeed realizes they're even there. If it really matters that the refs be in the thread proper, someone will have the sense to add {talkref}. Otherwise, no big deal. EEng 05:29, 28 August 2017 (UTC)

  @Rivertorch: It's very easy to work out which sections to add the {{reflist-talk}} to. Assuming that you are starting with all the refs displayed automatically at the bottom of the page:

  1. Have a look at the first of those refs; it will have one or more backlinks close to the start of the line (if there is one backlink, it will be a caret "^"; if there are two or more, they will be lowercase letters).
  2. Click the first of those backlinks, this will take you to some point further up the page, almost certainly in one section or another.
  3. Edit that section, add {{reflist-talk}} to the bottom, and save.
  4. Return to the bottom of the page, check to see if there are any remaining automatically-displayed refs; if there are, return to step 1.

  and you're done. --Redrose64 🌹 (talk) 11:25, 28 August 2017 (UTC)

Yeah, that's quick and simple! Another approach would be to just not worry about on exactly what part of the page the refs display. EEng 12:02, 28 August 2017 (UTC)

Thanks, Redrose64. That's more or less what I already do, and it almost always works, but a while back I encountered a talk page where all bets seemed to be off. I don't remember exactly what I eventually found the problem was (hatted sections? an improperly closed tag, maybe?), but it just stuck in my mind and when I saw this thread it occurred to me that it might be feasible to address the problem through automated means. @EEng, I appreciate that you consider it no big deal. I certainly don't think it's a big problem, but I also think it's often worthwhile to at least consider addressing minor issues that make the interface more confusing than it needs to be, especially for new users. I'm no perfectionist, but I also dislike the "good enough" approach when something might be easily improved. Rivertorch^FIRE_WATER 18:11, 28 August 2017 (UTC)

I agree about considering, and that's what we're doing now, but to me the answer is that it's not an improvement to add these instructions to this page. A big problem in editor retention is the learning curve, and by adding this we've made that curve a little steeper in order to get a slight cosmetic adjustment to 1 in 1000 talk pages – maybe (i.e. if this new instruction is remembered and heeded by newbies). And in many cases someone else will make the slight adjustment anyway. EEng 18:27, 28 August 2017 (UTC)

(At ten levels of indentation the text is virtually unreadable on a smartphone.)

@Chris troutman, Johnuniq, Redrose64, Rivertorch, and EEng:
Clearly I left out a crucial point in my proposal: Using {{Reflist-talk}} on one's own comment requires making sure that all previous comments with references have it as well. And that does complicate it, so it would be important to note that this is an option that you can use, but you don't have to.--Thnidu (talk) 19:28, 30 August 2017 (UTC)

Great – yet more complication to an instruction that solves a tiny use case in the first place. Until five years ago {reflist-talk} didn't even exist, and we got along fine. It was invented for a the very few times where, for some special reason, it really clarified things to emit the refs accumulated to a certain point (usually how-to pages, MOS pages, etc., on which the mechanics of refs are being themselves explained). You mean well, but this whole thing is a bad use of novice editors' very limited ability to absorb our already complicated rules and guidelines. I suggest we remove it completely. EEng 20:04, 30 August 2017 (UTC)

Honest question: would you say this guideline should be primarily for newbs? Rivertorch^FIRE_WATER 23:27, 30 August 2017 (UTC)

First and foremost it's a reference for OK and not-OK behavior, for when arguments flair up. To the extent possible, it should present that in a way calculated to allow newbies to absorb it readily. That's a hard balance to strike, and way down on the list is technically complicated oh-and-in-this-rare-case-also-do-this minutiae. I'll say it again – leave this for learn-by-example. EEng 00:12, 31 August 2017 (UTC)

The practice of putting citations in a talk page comment doesn't happen all that often on a single talk page, so I understand EEng's argument. I often insert this template and I feel that including it in the guideline moves my practice of adding the template beyond being my personal preference in not having citations ride the bottom of the page, but makes it a general norm which is why I support the inclusion. Chris Troutman (talk) 12:36, 1 September 2017 (UTC)

I have never seen an argument about whether reflist-talk should be included in a talk page section—it's part of referencing and does not need a "this is a good idea" guideline. The text is out of place here because WP:TPG is a behavioral guideline (don't use a talk page to express personal opinions on a subject or editor). If a reflist how-to belongs anywhere, it is at Help:Using talk pages, which is WP:TP. Further, adding how-to information anywhere will not help the problem of editors pasting refs into talk pages because the editors (often newbies) will not see it. They will only learn how to fix the issue when they see someone else add the correct reflist. Johnuniq (talk) 23:54, 1 September 2017 (UTC)

Great point about behavioral guideline vs. tutorial/help page. Talking about {reflist-talk} here makes it sound like you could get dragged to ANI for not doing it. I think the added text [18] needs to be removed as failing to have gained consensus. EEng 00:33, 2 September 2017 (UTC)

• This notion that this is not a common problem is misguided. I think a help page updates would help....but agree no need for this How-to info in guidelines .....add disputed text to [[[Help:Using talk pages]] ....Wikipedia:Talk page layout.....Help:Introduction to talk pages/3.--Moxy (talk) 15:26, 2 September 2017 (UTC)

Well, I've removed it here -- and look carefully [19] -- there was already text on the page about {reflist-talk}! EEng 17:26, 2 September 2017 (UTC)

Deleting bot notices

After chasing one link to another to find an answer, the question can only be asked here:

Q. Can bot notices (such as the InternetArchiveBot) be deleted from Talk pages? If they cannot be deleted ... can they be collapsed to use less page space?

Q. If there is a guideline somewhere about deleting bot notices from talk pages ... where is it because I didn't see it in WP:GTD, WP:NOTIFS, and WP:BOTPOL.

Thank you. Pyxis Solitary talk 00:00, 6 September 2017 (UTC)

Is there an example of a talk page which is suffering due to bot notices? I'm not sure what the problem is. I do not know of any bot notices which are unhelpful, so they should be visible for easy review. The problem with deleting them is that other editors will see that someone has removed text, and they may feel obligated to check that the removal was desirable. It's much less hassle for everyone if there are no deletions unless necessary, such as with spam. That has the benefit of not lighting up watchlists with unnecessary edits. Johnuniq (talk) 04:00, 6 September 2017 (UTC)

Talk pages don't "suffer". But they can become cluttered. What I understand by your answer is that when IABot edits have been checked, the notice needs to remain on the Talk page. If that's the case, fine.

Now ... say there are two or three IABot notices, one after the other: can they be folded within a collapse box? Pyxis Solitary talk 06:29, 6 September 2017 (UTC)

No authoritative answer is possible I'm afraid because as you say, there is no guideline. My answer would be that if someone is monitoring a particular topic and making substantive edits to the articles concerned, they might feel that some bot notices were intrusive and collapse or even delete them. However, editors who merely notice such bot posts should leave them alone mainly because an edit on a talk page can cause a dozen editors to feel they check what happened. That particularly applies for contentious topics where people might feel they need to inspect the talk history and check the diff of the removal to see that no other change occurred, and whether they agree with the removal. In short, I leave IABot notices but I do whatever it says about closing its request. Keeping the notice might also be a helpful record for anyone in the future who wonders why a certain edit occurred. Johnuniq (talk) 07:34, 6 September 2017 (UTC)