Help talk:Edit summary

Add description on Wikipedia App doesn't allow edit summaries

Hello! I notice that when I add descriptions to articles using the Android Wikipedia app, no edit summary is possible. Anybody know who to talk to about getting this fixed? "Short description added" or something equivalent would be all that's needed. DoctorMatt (talk) 04:40, 18 February 2024 (UTC)

@Doctormatt I don't, but I too wish it did. If I add a description via the Suggested Edits feature a default summary gets added, but if I use Add description on an article page, the summary is left blank. I try to remember to add it by editing the introduction instead, but frequently forget. Musiconeologist (talk) 02:10, 3 April 2024 (UTC)

Thanks for your reply. Oh, I think a request can be made here [1]. I'll have to try that. Cheers! DoctorMatt (talk) 04:25, 3 April 2024 (UTC)

Inadvertent scolding

I've caught myself writing summaries in the form of infinitival clauses like Avoid saying x or Put titles in italics. I mean simply that I've reworded an ambiguous sentence so it can't say x, or that I'm italicising some titles. But it can easily be read instead as an imperative addressed to a previous editor, so it comes across as "Don't say x, you idiot!" or "For crying out loud, won't you ever learn that book titles are italicised?"

It's convenient, but maybe a bad habit. So I wonder whether a very brief mention would be in order, maybe as part of the point about civility. (Something about awareness of how the summary might be read. I'm happy to come up with something.)

When something you add is edited very soon afterwards, it's easy to feel as though the next editor is 'marking your work' (sometimes without being at all qualified to), when really all they're doing is integrating it into the article or noticing a typo that you didn't. Musiconeologist (talk) 17:51, 13 April 2024 (UTC)

"Wikipedia:→" listed at Redirects for discussion

The redirect Wikipedia:→ has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2024 April 21 § Wikipedia:→ until a consensus is reached. Jay 💬 13:50, 21 April 2024 (UTC)

Editsummaries are not required

The wording on this page led people to believe that editsummaries are required, and worse still, that they could revert any edit without an editsummary.

And to be fair, it was very misleading. I have tried to improve the wording to accurately reflect the consensus that editsummaries are not required, but considered "best practice".

The section header: Always provide an edit summary

And these sentences:

According to the consensus policy, in general, edits should be explained.

According to the consensus policy, all edits should be explained (unless the reason for them is obvious)—either by clear edit summaries, or by discussion on the associated talk page.

are rather misleading.

Polygnotus (talk) 06:58, 17 September 2024 (UTC)

• I agree that the H:FIES heading is misleading, but the text of that section seems to be a fairly accurate description of WP:EDITCON, which also uses the word "should." Since EDITCON is the policy, I'm thinking you should focus your efforts to clarify the meaning of "should" there first.
  Meanwhile, what would you suggest as a better wording for the H:FIES heading? - Butwhatdoiknow (talk) 16:43, 17 September 2024 (UTC)
• @Polygnotus: The word "should" is weaker than "must" but stronger than "may". See rfc:2119. --Redrose64 🌹 (talk) 21:03, 17 September 2024 (UTC)

  Polygnotus' concern, as I understand it, is that some editors - unaware of rfc:2119 - read "should" as "must." - Butwhatdoiknow (talk) 04:47, 18 September 2024 (UTC)

  @Polygnotus left a note for me about this discussion (thanks!).

  Yes, I think that the must/should confusion is part of the problem, but I also think this help page is out of step with both the policy and with the community's practice. For example:

  • "Always provide an edit summary" – not "always". Perhaps "almost always in the mainspace", but the community doesn't care whether you use edit summaries when editing your sandbox or on talk pages.
  • "Except for the automatic summary when creating a redirect, which usually says all that needs to be said, these [automatic edit summaries] are not a substitute for a proper edit summary – you should always leave a meaningful summary, even in the above cases." – Nope. See Wikipedia:Revert, block, ignore.
  • "If you are copying text within Wikipedia, you must at least put a link to the source page in the edit summary at the destination page." – Best practice. Very extremely strongly preferred practice. But also not a fatal mistake.

  Overall, this feels like a page explaining what one person/group believes is the ideal. It doesn't feel like a help page. It might be helpful to compare this page to Wikipedia:Talk page guidelines and see how much duplicates TPG or could be merged over there.

  As for the WP:SHOULD problem, it's possible that we could address that in Wikipedia:Policies and guidelines#­Content. There's already a statement in there that the word must is not banned from guidelines, so we could add something like "The word should does not mean the same as must; see rfc:2119 for an explanation of how you should use the words in policy and guideline pages". WhatamIdoing (talk) 17:16, 18 September 2024 (UTC)

  I've made a few changes. Perhaps that will help. We can try more later if we need to. WhatamIdoing (talk) 19:59, 27 September 2024 (UTC)

Thank you! Polygnotus (talk) 21:23, 27 September 2024 (UTC)

• It may not be mandatory but if someone makes a big bunch of changes without any explanation, then I for one would have no guilt about reverting on sight (with the edit summary reverted, too many unexplained changes, try again with an edit summary). Choosing to make such an edit without explanation is contemptuous of fellow editors and deserves a response in kind. So yes, for minor changes it is a should but for significant changes it is a must. IMO of course. --𝕁𝕄𝔽 (talk) 19:40, 18 September 2024 (UTC)

  In such cases, do you look at the diff and attempt to figure out what happened?

  I wonder whether the problem is the complexity of the diff, or the size of the change. For example, if someone takes a neglected three-sentence stub and turns it into a 1,500-word-long article, how much of an explanation do you really need? WhatamIdoing (talk) 19:47, 18 September 2024 (UTC)

  Seems pretty clear to me that JMF is talking about complexity ("a big bunch of changes"). - Butwhatdoiknow (talk) 16:58, 19 September 2024 (UTC)

So what exactly is the consensus?

• Butwhatdoiknow summarises it thus: It is a best practice to provide an explanation for all edits.

This to me does not adequately reflect the sense of "you had better have a very convincing excuse for consistently failing do so", so I ventured

• Consequently, editors are expected to summarise their edits proportionately.

especially when we have a section entitled Edits should be explained, but that got reverted.

It seems to me that we are pussy-footing around here, we either mean it or we don't. De minimis non curat lex: nobody is going to get hauled up before ANI over the occasional oversight or trivial omission (not unless the complainant wants to be WP:BOOMERANGed for their pains). The purpose of the protocol is have a basis to tell serial abusers in words of one syllable to shape up or ship out. --𝕁𝕄𝔽 (talk) 22:53, 28 September 2024 (UTC)

you had better have a very convincing excuse for consistently failing do so is not the consensus.

The consensus is that editsummaries are not required.

The idea that, what you call serial abusers and what I would describe of good-faith users who do not see the point of editsummaries (like myself), need to shape up or ship out is bizarre and there is no consensus for that. Who really cares if someone does not use editsummaries as long as they are making good edits. Editsummaries are, in the vast majority of edits, pointless.

Just looking at some of your editsummaries, I sometimes don't even know what you mean, for example [2] and [3]. The editsummary is very often meaningless when not looking at the actual edit in its context, and I can almost always judge the actual edit without ever looking at the editsummary. In rare cases where I can't judge the edit without an editsummary, that is where using an editsummary is useful.

Polygnotus (talk) 22:56, 28 September 2024 (UTC)

@JMF: sorry I forgot to ping. Polygnotus (talk) 22:56, 28 September 2024 (UTC)

I've removed the line, since no version has been accepted. I'm not convinced that this line is necessary, especially in the lead, but here are the four recent versions, in case anyone wants to keep workshopping them:

1. According to the consensus policy, in general, edits should be explained.
2. It is a best practice to provide an explanation for all edits, or at least all edits that aren't completely and easily self-explanatory.
3. Consequently, editors are expected to summarise their edits proportionately.
4. It is a best practice to provide an explanation for all edits.

I think the objections to overcome are:

1. Do we need to invoke The Policy™? Also, that policy says "All edits should be explained (unless the reason for them is obvious)", and leaving off the exception feels like a POVish oversimplification.
2. This was objected to on the grounds that it provides (though in different words) approximately the same exception that is in Wikipedia:Consensus.
3. This seems unclear (what counts as "proportionate"?)
4. This is not the best practice. Wikipedia:Revert, block, ignore is a best practice that involves deliberately not providing an explanation.

Overall, I think that omitting this sentence might be best. WhatamIdoing (talk) 23:27, 28 September 2024 (UTC)

I don't think referring to The Policy™ makes much sense. Also I think the part of WP:EDITCON that says All edits should be explained does not accurately reflect consensus. If someone makes an obvious improvement, what is the point of explaining it when no one even asked for an explanation? It is only when people disagree with an edit that it could be useful to provide an explanation. But when the consensus is 25 for and 1 against, it does not matter if an edit has been explained or not. And even when someone disagrees with an edit and it is 1 for and 1 against an explanation is not always required see WP:RBI. Another exception is when things have already been explained, but not to everyone's satisfaction. See WP:IDHT and the article about Sealioning. Polygnotus talk) 23:38, 28 September 2024 (UTC)

That sounds like a very practical understanding of edit summaries to me.

Also, I have noticed a gap between "explained" and "explained through the exact method of adding an edit summary", and it's primarily the latter that is relevant for this Help: page. WhatamIdoing (talk) 23:54, 28 September 2024 (UTC)

I also think that a lot of people read All edits should be explained as All edits should be explained in advance. But in reality an explanation later upon request is also fine. And if someone makes a good edit but then does not provide an explanation later upon request that is also fine (maybe they are busy irl or their internet went out due to a hurricane or they are sick or whatever I dunno). Their lack of an explanation is not enough to retroactively make the edit bad. We don't punish people for writing bad editsummaries that are incomprehensible and of no use to anyone. And we can't demand that people are here 24/7 to explain any edits they made. They are not our employees, demanding they do something is a bad idea.

The idea that editsummaries are somehow useful for vandalfighters is obviously false. Vandalfighters often see vandals using editsummaries that do no accurately reflect the content of the edit. Even good faith users often use editsummaries that do not accurately reflect the content of their edit. Thanks to stuff like WP:ORES vandalfighters never even see the edits of longterm goodfaith users.

It is interesting to see that often people find it difficult to explain why they want others to use editsummaries, especially when you give some examples of editsummaries they used and ask how these were useful. I think it has a lot to do with a sunk cost fallacy, they spent all this time and effort and now it is hard to admit that a large portion of that was pointless.

It is easy to write a (java)script that can add a editsummary that is >95% accurate based on which namespace you edit and how many bytes the edit is (e.g. add the editsummary "My 2¢" on every edit over x bytes in the "talk" namespaces) but how would that benefit anyone? It would not. Note that the overwhelming majority of my edits are typofixes, YMMV depending on the kinda stuff you do here. For example, if you are an admin it is more important that you are able to explain your edits upon request per WP:ADMINACCT. Polygnotus (talk) 00:03, 29 September 2024 (UTC)

Again, Wikipedia is not a court of law. The purpose of this page is to give as "the long answer" to editors who seem to believe that the the edit summary box doesn't apply to them. And the only time that anyone will feel sufficiently irritated to draw their attention to this page will be in response to a series of seemingly pointless edits. Or indeed if they have given a summary but it is singularly unhelpful. So if you need this page to lend weight to your friendly advice, it needs to say clearly what it means. Of course there are exceptional cases but they don't need top billing in the lead.

If you care enough about a page on your watchlist, you will check all edits to it, summary or no summary. 𝕁𝕄𝔽 (talk) 07:25, 29 September 2024 (UTC)

"you will check all edits to it, summary or no summary" You give me too much credit. - Butwhatdoiknow (talk) 16:42, 29 September 2024 (UTC)

"Of course there are exceptional cases but they don't need top billing in the lead." Support. Butwhatdoiknow (talk) 16:50, 29 September 2024 (UTC)

The problem is not a lack of edit summaries, it's edit summaries that are completely unrelated to what was actually done, such as these two. --Redrose64 🌹 (talk) 10:26, 29 September 2024 (UTC)

True but this article will never reach that type of editor, so I suggest that we just set them aside. IMO, the purpose of this article is (a) to help new editors who want to understand how do it right or (b) those very few editors who challenge the basis for a {{uw-editsummary}} and so need pointing at an article that gives the long answer. [Another reason not to get sidetracked by the exceptions: anybody who already knows about those exceptions doesn't need teaching how to suck eggs.]] 𝕁𝕄𝔽 (talk) 15:52, 29 September 2024 (UTC)

I disagree with both of your comments on the main purpose of this page. This is in the Help: namespace. It's supposed to tell people how to do things, not to convince people that they should do it. It sounds like you'd like to write a persuasive page at WP:Always use good edit summaries instead. WhatamIdoing (talk) 20:30, 29 September 2024 (UTC)

Per WP:HOWTOPAGES (bold added), help pages "are intended to supplement or clarify Wikipedia guidelines, policies, or other Wikipedia processes and practices that are communal norms." That seems pretty close to "convince people that they should do" something. - Butwhatdoiknow (talk) 00:58, 30 September 2024 (UTC)

This help page is used to fool people into believing there is consensus to require editsummaries, when in reality there is not. Clarifying communal norms means explaining that editsummaries are not required, but can be useful in some cases, it does not mean that we should misinform people. Polygnotus (talk) 01:05, 30 September 2024 (UTC)

There have been attempts to do this elsewhere, e.g., at WT:CONS last year, where ButwhatdoIknow argued for more and better edit summaries.

We should also write this page so that it stays consistent with the information Wikipedia:Perennial proposals#Automatically prompt for missing edit summary – namely, edit summaries aren't required (also, reviewers shouldn't trust them). WhatamIdoing (talk) 01:27, 30 September 2024 (UTC)

And you take the position that we should have fewer and worse edit summaries?

Turning to the substance of your post, we all agree that edit summaries aren't requited. That doesn't change the fact that policy says edit explanations "should" be provided, with certain exceptions that are noted herein. That seems like stating a best practice to me. And HOWTOPAGES says informing editors of "common norms" is appropriate. - Butwhatdoiknow (talk) 15:53, 30 September 2024 (UTC)

The fact that bad wording on the Consensus policy is misleading is a problem that should be fixed, not a reason to not improve this helppage. Polygnotus (talk) 16:02, 30 September 2024 (UTC)

ButwhatdoIknow, I don't take the position that we should have fewer and worse edit summaries. I take the position that the help pages should tell people how to do things, and that our documentation in general should be accurate.

That means, for example, that if edit summaries are encouraged but not required, then this page should not tell editors (as it did, until a few days ago) to "Always use edit summaries". WhatamIdoing (talk) 22:33, 30 September 2024 (UTC)

The policy says edits "should" be explained (except in certain cases). It does not say they are required and, with the exception of a heading which has been changed, this page does not say otherwise. I'll post soon with a suggestion to move this discussion forward which, perhaps will reduce the chances that it is misused to "fool" people. - Butwhatdoiknow (talk) 15:42, 30 September 2024 (UTC)

Do you think that sounds like to give as "the long answer" to editors who seem to believe that the the edit summary box doesn't apply to them or to argue with those very few editors who challenge the basis for a {{uw-editsummary}}? I don't. I think this page should tell people what an edit summary is, how to post make one, how to write a decent one, and maybe even how to find other people's, but I don't think it's purpose is to tell editors they're wrong if they don't live up to someone else's standards. WhatamIdoing (talk) 01:03, 30 September 2024 (UTC)

Not really, {{uw-editsummary}} is to tell rude and inconsiderate editors that they are being rude and inconsiderate. This help page (IMO) should explain why anyone might consider their behaviour that way. 𝕁𝕄𝔽 (talk) 10:41, 30 September 2024 (UTC)

@JMF: You keep making rather bold claims without explaining why you believe those things, which does not help me understand your POV. Can you explain why you think that this editsummary was useful to anyone, and why you would've been rude and inconsiderate if you didn't type those 4 mysterious letters in the editsummary box? Or this one? You are aware that those editsummaries are completely meaningless without looking at the diff right? Also, you are aware that your views are held by a minority right? Do you recognize that the consensus is against you? Polygnotus (talk) 14:21, 30 September 2024 (UTC)

In your first case, a change that I made to my sandbox, the edit summaries for are for my benefit, nobody else's. In the history, I can see what I was doing when, especially as I may have deleted the item since then. I fail to see its relevance to this debate.

In your second example (which again is not in main space, it is your an editor's talk page) the case you cite is one of a pair of changes which, taken together, identifies what is going on. So if someone looks at the history of your talk page, they can see that (a) I welcomed youthe editor and (b) maybe there was a problem with that welcome. That gives them a clue as to whether it is likely to be useful to visit the page or, deeper still, look at the diffs. Without a summary, you remove that choice. That would be a discourtesy if it were a main-space article.

But these are (very) edge cases. Coming back to substantive edits to main-space articles, frustration happens when somebody makes a series of non-obvious changes without any explanation for doing so. In my book, that means that the editor concerned has no interest in consensus or even acquiescence: they are treating a main-space article like their personal sandbox. That is what I mean by rude and inconsiderate. Trivial changes don't fall into that category. To put it in perspective, I have very rarely given someone a {{uw-editsummary}}, it really has to be egregious behaviour. But tbh, I am more likely to revert the lot with the summary rv unexplained deletion and I am far from unusual in doing that. But I do WP:AGF and make at least some effort to understand the intent of the edits before giving up in irritation.

If the change is a just trivial few letters tweak, a summary is redundant. But otherwise, otherwise. The onus should not be on me to figure out what lies behind the changes your are making. True, you don't have to leave a summary but it is a courtesy to your fellow editors if you do. 𝕁𝕄𝔽 (talk) 15:09, 30 September 2024 (UTC)

Thank you, that is easier for me to understand. So, if I understand you correctly (please correct me if I am wrong), the only times it is useful to use editsummaries is when:

1. You are trying to convey meta-information not suitable to put on the page itself, and this meta-information is useful for others who try to understand the context of the edit (example)
2. You are using it as a bookmark/reminder for yourself or others (e.g. like I sometimes do when editing my javascripts) (example)
3. You are RFA-ing and worried that people will call you out if the percentage is less than 100.

Do you agree that editsummaries that fall outside of the categories I described above are not useful to anyone? Polygnotus (talk) 15:49, 30 September 2024 (UTC)

"including, but not limited to:"

1. yes, indeed failure to do so in this case is begging to be hit with trout UW template
2. yes
3. no

Another criterion is when you want peripatetic page watchers to recognise that these are changes that could use a second opinion (as in this case. It worked.). So its converse is a summary that says in effect "nothing to see here, have you no work to do?". This is what I mean by courtesy to fellow editors. 𝕁𝕄𝔽 (talk) 18:44, 30 September 2024 (UTC)

I was trying to make a list of all valid reasons to use an edit summary. Even if we add when you want peripatetic page watchers to recognise that these are changes that could use a second opinion to that list we still end of with a small minority of edits that are actually improved by an editsummary. And a large majority of edits where adding an editsummary is pointless. I can't think of any other valid reasons to use an editsummary, can you? Polygnotus (talk) 20:32, 30 September 2024 (UTC)

Perhaps convenience for other editors, or even your future self. I fairly often need to find a particular edit in a policy or guideline. An edit summary that exactly which key word is being added is rarely present but usually helpful. WhatamIdoing (talk) 21:11, 30 September 2024 (UTC)

Thank you, I added that to #2. Polygnotus (talk) 21:24, 30 September 2024 (UTC)

Both are problems. The first is laziness. The second is lying. - Butwhatdoiknow (talk) 16:40, 29 September 2024 (UTC)

What to say about FIES in lede

Okay, I think we all agree on what the policy says (as opposed to what we think it should say). The question is how to present that in the lede. Here is a proposal for what to say there -

Explanations of edits are strongly encouraged and edit edit summaries are one way of providing explanations.

No doubt this can be improved. - Butwhatdoiknow (talk) 16:04, 30 September 2024 (UTC)

Well you are using the word "edit" twice in a row. And there is no consensus to strongly encourage edit summary use. The fact that the policy is worded badly is not steady ground to build on; au contraire. The policy should be improved so that it follows consensus, and so should this help page. I understand that that sucks for those who disagree with the consensus. Polygnotus (talk) 16:07, 30 September 2024 (UTC)

I think the duplicate word is just a typo. WhatamIdoing (talk) 18:59, 30 September 2024 (UTC)

What the helppage should do is inform people that the consensus is that editsummaries are not required. It can then go on to say that editsummaries could sometimes be useful if they fall in the 3 categories I mentioned above. Polygnotus (talk) 16:19, 30 September 2024 (UTC)

I suggest you are over-interpreting the consensus. "Not required" does not mean "discouraged", which is how you appear to be positioning it. {{Uw-editsummary}} has not failed (or even proposed?) at TFD. Afaics, the debate is only how firm the encouragement should be: it seems obvious from the comments above that most editors believe that summaries should certainly be encouraged and the lead should reflect that. (Although the current version – with nothing at all, leave it to the body – actually works rather well IMO. Do we need to change it?) --𝕁𝕄𝔽 (talk) 18:54, 30 September 2024 (UTC)

"discouraged", which is how you appear to be positioning it no, I do not think it is discouraged. I just pointed out that there is no consensus to strongly encourage edit summary use. Polygnotus (talk) 20:08, 30 September 2024 (UTC)

I think that both of these things are true:

• Edit summaries are not required.
• Edit summaries are encouraged.

I think I'd start with "encouraged" rather than "strongly encouraged". I'm also not sure that this needs to be mentioned in the lead. WhatamIdoing (talk) 19:00, 30 September 2024 (UTC)

If we need to add anything, it is that there is no consensus to require editsummaries. Polygnotus (talk) 20:11, 30 September 2024 (UTC)

I've seen the situation before where people who do not agree with the consensus watchlist a policy- (or help-)page and derail any attempt to implement the consensus. The solution is simple, either they allow improvements that are in line with consensus (even though they personally disagree), or we need a wider discussion (something like WP:3O or WP:RFC or whatever) which ensures that a minority viewpoint can't block the implementation of a wide consensus. So it is probably wise to add some stuff to make the page more in line with the consensus, and if they get reverted move to the less-preferable option. Polygnotus (talk) 20:36, 30 September 2024 (UTC)

Or perhaps the correct approach in this case is to make no edits, and have the lead say nothing about the should/must/always/required-ness of edit summaries. WhatamIdoing (talk) 22:36, 30 September 2024 (UTC)

Hmm, maybe, gotta think about that. Polygnotus (talk) 22:57, 30 September 2024 (UTC)