I can see why this question was voted unclear but I think it's ok on that score. Just wondering if I should vote to leave open or if it's better closed as a bug report. The answer is valuable but I don't think it's solved the problem
I'm currently going through some posts with one-character links as returned by data.stackexchange.com/askubuntu/query/1130245/… and editing them to have proper link texts, or removing if necessary. If anyone feels bored wants to help, that would be nice :)
This edit massively expands a very highly voted post with material that is potentially useful but not necessarily the best approach. I recommend rejecting it as "clearly conflicts with author's intent." I've rejected it with a custom reason advising the editor to post a new answer, and I've unprotected the question so they can do that.
no longer needed - the OP is still editing the question to improve it so presumably doesn't want to delete it; also, I'm not sure the OP could delete it, considering the score on its existing answer
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, link at end of body, pattern-matching product name in body, +3 more (592): Weight Loss Pills 2019 by justathompson1 on askubuntu.com
Is this question really no repro? Removing all but one Java implementation seems like it continues to be answerable and useful. It's also possible to tell LibreOffice which Java implementation to use, to satisfy the package manager that one is installed manually to satisfy dependencies, and even to install and use most of LibreOffice without Java at all.
(Most of it has not been written in Java for many years. Word processing, which is probably its most common use, doesn't require Java.) The question might be too broad, or in theory even unclear. I suspect, and even hope, that it is a duplicate of something. But it seems to me that the no repro closure is a mistake.
I'm mostly agreeing with this comment and leaving it closed. If anything, it should probably be asked again instead, about a modern system and more specific.
Although this 5+ year old unanswered question has interesting historical value such as @ThomasWard used to be called @ LordOfTime it addresses Ubuntu 12.04 or earlier which are EOL. OP hasn't signed on in 4+ years and is unlikely to respond to new comments. The Java of that time has been updated many times so VTC as not reproducible. — WinEunuuchs2UnixJul 23 '17 at 19:59
I've recently hooked up a second monitor to my laptop, and for some reason, Ubuntu isn't very snappy any more. There's a very slight lag between me doing an action and it showing up. For example, after typing a letter, there's a tiny delay between me pressing the button and it showing up on the s...
Many Wine apps and also some Java apps like Evernote or JDownloader appears in Unity tray with an annoying white background.
Is there a way to solve this issue?
Here is a screenshot about it:
Yeah, I don't know if those two were rightly closed or not, but it seems like feedback may be needed for them and also it's not obvious if the problems they're describing continue to apply to anyone.
@EliahKagan I wondered if I could ask you about askubuntu.com/review/suggested-edits/990549 — this is literally my first time using chat on a Stack Exchange, even though I've been active otherwise for almost ten years, so if I'm in the wrong place or this interaction is otherwise incorrect, please excuse me
These messages can be moved if it turns out there is a better place for them. I suspect I will end up moving them to the Downboat. But there is no need to wait for that to continue this conversation. Please do ask!
Would you be willing to post an answer with the information you had proposed in that edit? I think that would be helpful, and I unprotected the question so that you can do that, if you so choose. Of course you are under no obligation to do so!
@EliahKagan Well, from my research, the answer as acccepted no longer works on Ubuntu 18 (maybe 16) and later except in a case where --fix-missing isn't required—which are uncommon for users who do not manage their own package repositories like corporate users. So, back in 2015 this was the right answer, but no more. Oh, there's one other wrinkle to mention:
The question as asked here has logs that indicate it is one of those cases where --fix-missing won't be required. But many other questions have been closed as duplicates pointing here that will not work without --fix-missing. Hence the many comments "doesn't work for me"
I can answer it as an unaccepted answer, but it seems odd to me to have the accepted answer be one that will almost certainly not work.
If you think this is a special situation where it seems like the accepted answer to a very popular question doesn't work anymore, it might be reasonable to edit it, but I recommend asking on Ask Ubuntu Meta because I'm not sure how people will want to proceed.
There are some alternatives, including adding an answer which someone bounties, adding an answer index to the question which clarifies which answers are likely to work when, editing a short note into the existing accepted answer, making the whole question a duplicate of something else, and some others.
With that said, I used the technique shown in that answer last month on 18.04 LTS and it worked fine for me. The situation I was in may have differed from the most common ones, I'm not sure.
@EliahKagan Ahh... If I follow you, that explains something–on Ask Different, I edited an accepted answer that used a program that was no longer included in macOS to use a different method with programs still in macOS, an when that was rejected as not "reflecting the author's intent", I made a different edit showing how to install that program from third-party sources, and it was rejected also.
So, basically, answers shouldn't be seen as evolving things—if they worked when they were accepted, and may in some cases still work, they should be left alone and one should request protection be lifted so new answers should be added instead?
I wouldn't say that. Answers should be seen as evolving. One should just be careful when one radically changes answers. Yet sometimes that should be done. This may turn out to be one of those cases. It is also possible that I have missed something; for example, one of the reviewers of your edit, @Kulfy, had voted to approve it.
If you do post a new answer to the question, that need not preclude other actions. For example, one approach would be to post the new answer and then post on meta. (You don't have to post the answer to post on meta, of course.)
I'm going to go ahead and move these messages to the Downboat because they're on-topic there. (And secondarily for continuity; I had briefly mentioned the review item for this edit there.) Also, @Kulfy is active there and in the room right now. I hope you don't mind the inconvenience of the messages disappearing here and appearing there (and the automatic invite to the room from the system).
In that case where you do new answer, Posting on meta would be for the purpose of...? (sorry, I've never written on meta and have only read it for policy-oriented posts, not particular-post-related ones)
Whether or not you choose to post a new answer, in this case I suggest asking on meta about what to do about the situation where the currently highly upvoted accepted answer is not longer satisfactory. As mentioned, there are various possible approaches. On meta people can post answers with their ideas of what should be done, and consensus can likely be found based on votes there.
Although posting on meta is what I recommend, I should mention two other options: (1) there's a bigger group of people in the general room and you can ask people there what they think, and (2) you can submit the edit again, with a more detailed explanation of why it is important or necessary as the edit summary (perhaps including a link to your findings about why you believe it hardly ever works these days).
However, 16.04 is still supported and used by a number of people, and I have personally used the method in that answer successfully on 18.04 recently.
@EliahKagan IIRC once terdon told that edits must be constructive to be accepted. While reviewing I felt that @Trey's edit makes sense and it ain't deviating the post's intent and is neither an attempt to reply. And I approved that.
@EliahKagan In that case then I will add a separate answer with an "if that doesn't work" preface. Thanks for explaining.
But to widen the scope a bit, a couple times on different SE sites, I've asked a question, and in my question specified that it's not a duplicate of X, (where X is a question I've seen other questions like mine be closed as a dupe of) because the answer in X no longer works, and I've still had the question closed as a dupe. I've usually just sighed and maybe tweeted and somebody I know who works at Stack Exchange tells me what to do...
which I realize isn't a scalable solution. ;-) Is that a case where you should go to Meta?
I have an accepted answer on one site ("stack"? what are they called?) that I've seen misdirected as a dup this way. I've wondered if I should delete my answer. I haven't because it could still be useful to someone running old hardware.
I no longer have the hardware necessary to write an updated edit.
Oh, even by the author? Didn't know that. (Since as I said, I haven't yet tried it.)
I clicked "delete" experimentally to find out and it gave me the "do you really want to delete this?" confirmation so I said no. If I'd said yes, it would have then told me I couldn't, I guess?
How can a post be deleted?
By a user:
The author can typically delete their own posts at will; for exceptions, see When can't I delete my own post? below. To delete a post, just use the delete link below it, on the left (only available from a browser, not the SE/SO app).
Moderators can delete ...
You can't delete your own answer if it has been accepted.
@Trey What you're describing is a real issue, which I've observed and sometimes even unintentionally participated in. With that said, often the problem of wrong dupe closures can be solved by making more explicitly clear the specific reasons your question is not a duplicate. That can be done in edits to the question.
@EliahKagan Just saying that the answer doesn't work or no longer works is not usually sufficient--you should say specifically why it does not, usually by stating explicitly what you did and what happened instead of it working (including full terminal output or whatever is relevant in the context of the question).
@EliahKagan Questions that say they're not dupes but don't give enough information to make clear specifically why and to facilitate new answers based on that information are likely to be closed as duplicates anyway, or as unclear.
The situation is further complicated by how in most circumstances the system displays the close reason that the majority of voters voted for, so even after an edit that says your question isn't a duplicate, your question may get more close votes as unclear but still show that it was closed as a duplicate even if that wasn't the intend of the later voters.
But if you're talking about a situation where the question already contains that information (as best it can), then asking on meta is definitely a good next step. I've posted numerous times on Ask Ubuntu Meta about questions I thought should be reopened, especially before there was a review queue for reopening questions, but even after that. It's definitely one of the things the meta sites are for.
I may contribute to this issue in a way—if I am answering a question I fear is an XY question, I often explain the XY, subhead, answer the literal way, and then subhead, answer the functional way. I guess that sort of answer complicates the "what is and isn't a dupe?" issue.
Like, for a lot of questions, the questioner asks "how do I do X with [$shell, $language, $tool]" where the shell/language/tool they specify is clearly not the right one. But there will both be people who need the literal answer and the "better way" answer.
Because they're forced into using the suboptimal tool for some situation-specific reason, or they don't care what tool to use and are pleased to discover some other tool solves the issue easily.
Okay, I submitted a new answer with a note about what makes it different. askubuntu.com/a/1185766/333672 — is there any reason in this case to post to meta, @EliahKagan and @Kulfy ? Appreciate the assistance, and sorry for wasting your time!
Thanks very much! I'm googling the meta sites to look for this, but I'm not seeing it (probably my own bad search-fu): is there any general thought about the issue where someone posts an issue or error message that sounds very specific but turns out to have a broad range of possible answers? These seem to cause many problems down the line.
Ah, nm, I have found a couple questions on meta about this—all closed with "This question pertains only to a specific site in the Stack Exchange Network". So I guess this is a per-site question. I feel like the issue is larger than that, but, maybe not, I guess it can depend on subject matter, especially for the newer sites.
I would say that if you think further action should be done, such as having the accepted answer edited in some way, then it is definitely still appropriate for you to post on Ask Ubuntu Meta. But if you think the situation is okay as-is, then you need not do so. It's totally up to you.
@Trey Yeah, to be embraced on Meta Stack Exchange, I suspect a question about this would probably have to bring together a number of cases from different SE sites and reference existing child-meta discussions about them. Alternatively, feature requests related to the way the system works in connection with this issue are on-topic on Meta Stack Exchange... whether or not they would be popular, I don't know.
But this is definitely on-topic for the child meta of the site where one or more questions where it applies exist.
So, for example, in connection with questions on Ask Ubuntu, you could post on our child meta, Ask Ubuntu Meta.
(Btw, it looked like you may have been trying to reply to your own message to clarify threading, as I often do. To do this, start the message with a colon -- that is, with :, not # -- followed by the message number.)
@EliahKagan Understood. The example I've used to describe this issue is how, if my grandmother had asked "why is the food in my refrigerator rotting so often?" the only right answer for her would be, "you live right under the train tracks—you need to secure the fridge door so it doesn't let in warm air whenever a train passes". But that's not going to be a common answer to what may be a very common problem!
As someone newly trying to understand policy by reading meta posts as I've been doing for the past two hours, it's really frustrating how much is covered by "for example, [LINK] is ok, but [LINK] is not" where the links are to things I can't see.
At least some sort of one-sentence summary would be helpful in these cases
And when so many things link back to FAQs where it's clear that because it's an FAQ, everyone in the know knows what it is all about, but if you don't know and go read the 8 year old post itself, the things it links to are all deleted so you have to kind of figure it out through inference
Many Wine apps and also some Java apps like Evernote or JDownloader appears in Unity tray with an annoying white background.
Is there a way to solve this issue?
Here is a screenshot about it:
