^^ Can more useful advice be given there? I'm inclined to suggest posting a new question, but I can't tell if the issue they're having is the same as the one asked about already or not.
@Zanna A lot of the posts the query returns seem to just have trailing whitespace after opening or closing triple-backticks. Although I think this is a sign that the post may have formatting problems, it doesn't itself cause a problem. I didn't especially want to include those, I just didn't know of a way to avoid them in LIKE expressions. Although many of these posts can benefit from editing, I think the ones with truly broken formatting, where a line is hidden, are especially important.
So I don't think the current query--specifically, the part I generated with my Python script--is really working out. We might get through all the posts in it in a couple months, assuming a modest amount of edits per day.
With a more focused query, we might find the truly broken posts and fix them faster. One benefit of doing that, I think, is that it would give a sense of how widespread the problem is.
A number of meta posts offering formatting advice should be edited sometime soon to account for how the Markdown dialect is changing to CommonMark. With information about how common this particular mistake is, it may become clear how prominent cautions against it should be (and perhaps also whether or not it would make sense to make feature requests for SE).
@EliahKagan I saw your notes - thanks for explaining. Since I didn't understand (or more accurately, did not try to read) your full set of LIKE expressions I didn't understand why those posts were getting caught
@EliahKagan I'm fine with working with the query as it is, but if you want to adjust it and run it again, that is also cool. Since a significant amount of editing has already been done, if you want to change the query and run it again, I think it would be better to do that and re-start the etherpad after the data dump (used to be Sunday UTC morning I think, but they may well have changed the schedule) (assuming you actually want to continue using etherpad or similar).
@EliahKagan I wonder if this is why code fences weren't supported previously XD
@Zanna That's quite all right, considering that I did not try to write them. :)
I wrote a script that generated them. The script is thrown together in a slapdash manner, but still much more readable than the full set of LIKE expressions. (I mean, I assume it's more readable than them. I haven't tried reading my LIKE expressions either, though of course I inspected a few.)
@Zanna I'm not sure if I should learn to write a more focused query for this, or make the query far broader and process the output locally (as jokerdino suggested doing).
This is very ugly and it doesn't recognize ~~~ fences (though the existing query doesn't find them anyway), but this basic idea seems to work okay:
csvgrep -c Text -r '(?:^|\n)(`{3,})(?!`|[^\r\n]*\1)[a-z]*(?:[^a-z\s]|[ \t]+\S)' fences.csv
This returns rows whose Text column contains, either at the beginning or just after a newline, a group of three or more backticks not followed by a backtick and not followed by itself anywhere on the same line (i.e., not closed on the same line, since SE would treat that as a code span), which later on the line contains either (a) a character that is neither a lowercase letter nor whitespace, or (b) a space or tab followed by a character that is not whitespace.
Should this post be edited to make clearer that it is an answer? Technically it does contain the solution, but it's phrased as a thanks post. Also, it was intended as a thanks post. But there is no other answer presenting that solution.
@Natty tp
user435118
@EliahKagan It is mentioned in a comment under the question
Right, but that's not an answer. At least if the question is to remain open, the solution should probably be clearly stated in an answer.
@Natty tp
@Natty tp
@Natty tp
@Natty fp - as far as Natty is concerned. But examining the question, is this really answering what was asked? I think the problem is that hibernation didn't work at all, not that there was no convenient way to trigger it.
@Zanna I think that answer is correct. I recall we have a question for this error.
OTOH it usually shows the version of the message with "are you root?" when that happens.
I've commented and voted to close the question as unclear.
@pomsky Did the linked question actually solve that? I don't think the answers there cover what the OP did in the firmware ("BIOS") setup. Also, the OP seems to be saying that they didn't just enable UEFI mode, but also enabled Secure Boot, and that having Secure Boot enabled somehow contributed to fixing the problem.
@EliahKagan hmm... I edited it but I am not sure I have improved the situation at all
is OP's comment on this question a workaround, or are they just confirming that the problem is related to locking or something? Should that question be treated as a bug report?
I don't know. Ordinarily I'd comment to say they should post that as an answer if they consider it an adequate solution to their problem and otherwise edit their question with information about it. But this was quite a while ago.
@Natty fp Should be expanded, but the first line is usable even without the link. I'll expand it, but I still think this is a better fit for fp than ne.
@EliahKagan I'm not quite sure. It seems OP's issue was addressed in the comment exchange below the question. I'm not sure whether it can be expanded to an answer or whether that would worth it. OP didn't visit the site after that. Perhaps closing as no-repro would be useful? Or any better target?
@Natty tp Borderline but I think its asking a question and not trying to give a partial workaround.
@pomsky I wouldn't close that as no repro. The solution given in that comment is somewhat unclear, but I think not enough to be useless. I agree with looking for some other dupe target to include in the dupe banner. But it would also be okay for an answer to be posted based on the comment. I think the important part may have been enabling UEFI mode, and that Secure Boot was on was mostly irrelevant. Also, it may be valuable for people to know that it wasn't necessary to disable Secure Boot.
@Zanna Thanks!
But I feel it would be better if simplified to the point where it is no longer impressive. :)
@pomsky I have reworded the question and deleted the NOTE. I think the question now makes sense about the confusion. I hope now it's clear. It still need 1 undelete vote. Also, the comments may also be considered as no longer needed.
No problem! I wouldn't have edited just to remove that trailing whitespace (it wasn't actually breaking anything), but that's what made the post turn up in the SEDE query and I thought there were a couple other things I could improve too, so I went ahead.
@Zanna This version is more understandable and also fixes a bug:
#!/bin/sh
# filter-fences - Uses csvgrep to find posts that may have broken code fences.
# shellcheck disable=SC2016 # The operand to csvgrep -r is meant literally.
csvgrep -c Text -r '(?mx) # Match ^ on each line. ("." does not match "\n".)
^[ \t]*(`{3,})(?!`)(?!.*\1) # A code fence, not all one one line.
[a-z]* # Optional lowercase.
(?:[^a-z\s] # Non-whitespace non-lowercase, OR
|[ \t]+\S) # blanks followed by non-whitespace.
@EliahKagan I think I should just flag their comments as no longer neeeded instead of pinging them. By pinging I may also be leaving some redundant and unhelpful comments.
Do we have something for this? The commenters should've posted answers instead, but I'm reluctant to suggest doing so since I strongly suspect we already have something that explains it.
@Kulfy It's against the promotions policy because it doesn't have disclosure. I don't recommend red-flagging it. I flagged it NAA, but I also !!/reported it in Charcoal. The author is likely unaware of the rules on that, and they have numerous non-spam contributions.
@Kulfy I don't know if it's that linked bug or not, but I don't think that's strong enough reason to close it as a bug. Also, I'm not sure where the current consensus (if any) is about when questions should be closed as bugs. This doesn't seem to be done as aggressively anymore.
@EliahKagan I've gotten a lot of pushback voting to close things as bugs, most of it does not seem to result from any misunderstanding of how the site works or what policies have been establishes. People who object to such closures very often have good points about how closing questions like that gets in the way of people being able to get help.
Not only that, but my attempt to close this question of mine, which I would've thought was a pretty clear case of a question that turned out to be a bug, was unsuccessful.
At this point, I basically only VTC questions as bugs when they are (a) clear attempts to actually file a bug report rather than ask a question, or (b) transient bugs with useful bug reports where they're unlikely to affect people in the future.
I'm not saying that these are the only cases where questions should be closed as bugs, and I'm certainly not saying we have an actual policy that they are (because we don't have that policy, as far as I can tell), but those are the only situations where I'm personally confident closing questions as bugs is actually consistent with community consensus.
I think this question is about Ubuntu (they're interested in installing Ubuntu, they tried installing Lubuntu, they came here for help, and the mention of Mint does not change that). Also, I really don't think it's OT EOL. Someone suggested installing an old release in comments, but the OP is not asking for that and the release they tried to install is supported.
@Kulfy The question and answer would benefit from being expanded by editing. But even if they're not, I'd say nothing has to be done there.
The recent answer would benefit from expansion, yes, especially considering that it does not even link directly to bugs.launchpad.net/ubuntu/+source/vino/+bug/1668309. It's possibly NAA, though it does say something about the connection between what the OP described and the indirectly linked bug, so I haven't flagged it as such.
