I was personally amused by this sentence in the blog post: "We saw no decrease in question quality." ... which means the goal was really "We saw a significant increase in the number of people who started and posted a question.". Combined with the increased fake internet points, this can only tie back to "how we can reduce friction between people"
not that I see any connection; just trying to see a thread through the post
anyway, hope everyone "take(s) the responsibility reverently"
@FaheemMitha vi ~/doc/notes/notename.txt. Or if you want to get fancy, write it in markdown. or latex. vi ~/doc/notes/notename.md (vim has nice syntax highlighting for .md and .tex files). You can use sub-directories to organise notes into categories. and/or by date. And you can write trivial wrapper scripts/functions to create new notes with title, date, etc nicely formatted at the top of the file before launching vi. Use git or hg or something for version control.
@cas Thank you for the suggestion. The advantage of a software system is that it could be more flexible. I.e. multiple tags. Which you can't get by organizing in directories.
@ilkkachu Apparently some people just have nothing better to do with their time.
@Kusalananda I did.
Maybe tomorrow they'll decide to decorate the entire network with pink and yellow stripes?
Handel can be very cheery at times. Consider, for example "For Unto Us a Child is Born".
I wonder if anyone has that set as their alarm?
Actually, it might be a bit startling to wake up to.
Damn, I was waiting for the far distant day when I would finally attain 25k on this site. SE robbed me of that.
This might sound like a weird question, but has anyone ever had any success getting a business (or businesses) to use PGP/GPG? Like a bank or broker, for example? And is GPG usable on Windows these days?
Key exchange is still a problem and in some ways PGP handles that well, except for the major unfixable flaws in the protocol that have made it unusable
I still like the concept but keyservers just aren't workable any more
@FaheemMitha you can add plain tags or whatever you want to a plain text file. just structure it kind of like an email message (or a debian Packages file entry) - a header section with fields, a blank line, and then free-form text. e.g "Title: title", "Tags: tag1,tag2,tag3" or "Date: 2019-11-14". the headers can be easily grepped, or parsed with perl or awk or whatever.
yeah, i use text files with that format for lots of things. it's simple and easy to work with (even json is overkill for most of the things i want to do).
@cas Ok. I actually use TeX heavily all over the place, all the time. But it's not necessary for everything. E.g. it doesn't really buy one much when writing TODO lists.
Of course, TeX is technically text too. Sometimes not very readable text, though.
I use plain text files (actually plain text with some markdown-like formatting for bold & italic) to generate latex files. useful when I have a list of things that I want to format consistently in latex. write the .txt files, then write a perl script to process them all and concatenate them into a .tex file (generating ToC and hyperref links at the same time), then create a PDF or whatever from that.
and one of the nice things about it is that because it's all plain text I can embed .tex into it wherever I like. e.g. for embedded tables or figures.
(the bold and italic markdown is just a minor convenience - it's easier to type ** and * than textbf{} and emph{}....and trivial for a perl script to convert from basic md to tex. I originally used pandoc for that, but it was overkill for my needs. and worse, pandoc's md -> latex conversion is crap: gold in, garbage out)
i thought luatex was for embedding lua code into .tex files?
anyway, while lua seems like an OK language, i don't have any burning need to learn it. also, perl is almost idea for working with text files, and I already have over 20 years of experience writing perl code.
Mods, what do you think of using “reject and edit” as a way to short-circuit the voting process on edits which introduce errors, e.g.here, here (minor, admittedly), or here (subsequently “accepted and edited” by another user)?
I tend to favour the community vote but I’m seeing lots of unfortunate approvals in edit voting :-(.
@EliahKagan yes, that particular user is a repeat offender, but we don’t have nice edit education tools aside from comments left under posts, or reject messages (but that only works if the edit is effectively rejected), or dedicated chatrooms (which often involves mod intervention).
@EliahKagan And another, though less severe. This one uses code formatting excessively, but perhaps I should've reviewed it as Edit rather than Reject.
@StephenKitt If some of the editor's bad edit suggestions were accepted, I think the editor should be pingable in comments on posts where they were. Whether or not it would be a appropriate to do so would probably depend on context, though. (And I would not suggest doing that on all of them, but maybe just one.)
@terdon a bunch of edits by one specific user, whom I’d noticed before. Many of them are good, but some of them go too far (copy editing code, or introducing grammar problems of their own).
I’ve always though it best to leave the voting systems as-is, for the reason ilkkachu gives just above.
But I probably need to recalibrate my “reject and edit” trigger...
And there are users who do that: they have significant levels of contribution to the site, which has earned them a lot of reputation points, but they give them all away in bonuses.
@StephenKitt The general mentality for SE sites is if you have a privilege, you are trusted to use it. And you already have the rep to edit posts without being reviewed. Therefore, there is no reason for you to think twice about it: if the edit is something you would have edited again to correct had you seen it in the wild (in a post instead of in the queue) then go ahead and use reject & edit.
Don't think of it as circumventing the voting system: you have full edit privileges, so you already "circumvent" that system every time you edit. This is no different.
My posts have been on the sites long, and experienced those days when the traffic per post was high. If those who have targeted me had similar experiences, I believe they would earned much more reputations
@JeffSchaller It's how we all learned. But Stephen is worried that because his choosing to "reject and edit" will immediately reject the edit with a single vote, that means he's circumventing the voting system.
@JeffSchaller weird choice of words there, "currency". It's not like voting costs anything in rep (well, except downvotes, but it's minor compared to gained upvotes)
what was the rep limit for being able to approve/reject edits? is it the same as being able to edit yourself?
because, if it is, then one could just always choose to unilaterally accept/reject the edit, in a sense. (that leaves the reviewers name visible as the last editor, but does anyone care about that)
No, but repeated edit suggestion rejections from multiple users will cause your edit suggestion rights to be suspended for (n) days, where n is currently 7.
Note that after a recent change, automatic rejections due to edit conflicts are ignored and won't affect the suggested edit ban mechanism. ...
@JeffSchaller It is not a bad idea. It would be better that someone else post it, and let whoever propose posts and whoever in guilty mood and in leisure time award the bounty
Regarding the motivation of reverting the reputation points of Qs back to the old days, I guess SE/SO the company senses that questions generate more traffic than answers, and that is what their business model needs
something I've meant to learn more about are chat "conversations" -- I wonder if they'd be useful here (if we tagged a "bounty" conversation that would have any future benefit)
Yeah, I was hoping (imagining? hallucinating?) that it would extract the comments & replies that were all related to that topic into something more useful.
instead, it seems to grab everything, and requires an end-point
@FaheemMitha yes, complete with MWE code, thanks to your hint! and now I need to work on question #2. Yes, about minipages -- trying to keep 3 elements together on a page somehow.
Matching fields from passwd with grep indeed seems hard. Five answers with regex solutions to that question and all had the same one or two errors (the other one's a bit marginal though)
There's a comment on a deleted answer from the OP that said "no, I can't use awk or perf (perl?). They might be better solutions, but I can only use grep or egrep with regular expressions."
"I need to print the whole line for this to work, but the 4th field value should be highlighted." -- Highlighting just that one field makes it even harder. Wonder if that's even possible with just egrep
I didn't look at the deleted answer, but that comment there seems much stricter than what the Q says, or what you wrote in comments under it :)
For (all your) information, I made the mistake of look at Meta just now and saw Why is SE removing links and community ads about legal issues? in which it says "Stack Exchange staff will actively remove links to a legal fund campaign from user profiles as well as community ads surrounding Monica's situation & reinstatement", just in case you're involved.
@JeffSchaller that's interesting! I'm at a comfortable 1.57 ratio. Looks like there are only two people with a higher rep and a higher ratio...and you're one of them. :)
@JeffSchaller: Not sure if you just wanted that user to add it to their answer or generally wondering but the grep [s]tring will not find the grep process itself in the ps list
@Jesse_b although as present in that answer, if there is a file called "program" in the directory the script is running in, then the grep process will be in the list. And then the kill -9 will hopefully fail, unless the PID has already been reused. :)
Why people use ps | grep | awk instead of just pgrep is really a puzzler.
The pkill/pgrep in OpenBSD was inherited from NetBSD. They wrote it in 2002. I believe that the version you'll find in Linux is derived from one of these, but an early version (there's usually no -q (quiet) option for pgrep on Linux).
Hmm... there's not even a -q in NetBSD's version. That must be an OpenBSD invention (it is, added in 2012). It's handy.
I think it's also that people learn one command to do one particular task, and then another command to do another task, without learning that the tools in question can do other things too (let alone knowing how to use them for that)
@Kusalananda Interesting. I recall when I wanted to run some command with a one-second timeout. My mentor (a sysadmin with 20+ years of experience) was just in the middle of saying that that particular command didn't have a timeout option...when I found (using apropos) the timeout command and used it. It was too new for him to have learned it. :)
everyone knows grep can find a matching line, and awk {print $2} is also a common-ish pattern (since cut -f2 doesn't work for columnated data). And they know how to connect the commands with a pipe, so no need to go further
@ilkkachu There is also a conflict between the plain statement of the UNIX philosophy and the realities of dealing with certain tools. For example, you have to use the sorting options built into ls if you want the output sorted. Piping to sort is crazy.
Hah, anyone else getting an error loading the main site?
> Description: An exception occurred while processing your request. Additionally, another exception occurred while executing the custom error page for the first exception. The request has been terminated.
@Kusalananda That's why in my own shell training, I stress how to find documentation and tools, more than just covering some particular pre-selected set of tools.
Server Error in '/' Application. Runtime Error Description: An exception occurred while processing your request. Additionally, another exception occurred while executing the custom error page for the first exception. The request has been terminated.
Why am I seeing my chat messages twice when I first post them?
@FaheemMitha that's an understatement. Here is the recent discussion (started, you might notice, by an ex-mod—and NOT by the people proposing the change). meta.stackexchange.com/q/337843/307622
Anyway, it seems like they want more questions. That's one way to try to get them. Though I don't see why people would ask more questions just because they are getting IIP. People ask questions when they want to know the answer. Or maybe they're just bored and don't want to watch TV. More likely the former.
@Wildcard Yes, I saw that already.
My point was just that there should be some sort of consensus about these changes. Not some people get an idea in their heads, and then implement it.
@Kusalananda I can't see those linked articles saying anything about the rep change? If they have UI improvements or such that make it easier to post better questions, then I can't see how anyone could oppose that. But that's different from how much rep the votes on Qs are worth
@FaheemMitha the consensus part is what many in meta.SE seem to wish, yes.
> We can look back on this decision with the benefit of hindsight. This decision may have been the right call then with the information we had at the time, but we have seen the effects it has had on our community. We reward people who give answers at a higher rate than people that ask questions.
The "effects it has had on our community" assumes that Stack Overflow would have been just as successful had this decision NOT been implemented, and that the only effect was to make the place less welcoming for question askers.
This reminds me a bit of how Airbnb behaves. In both cases, they behave like they own the shop. In actual fact, all they own are a few domain names, and web pages. Which would be mostly empty if it wasn't for the people who populate them.
In the case of Airbnb, it's the hosts. In the case of SE, it's the user communities.
@FaheemMitha there's a rather important distinction that SE has all the content which has been contributed, and Airbnb has absolutely nothing if people stop hosting new guests.
@Wildcard That's certainly a valid point. But content does need to be maintained. And people would have new questions. An inactive site is ultimately a dead site.
Anyway, it will probably come as no surprise to anyone here to learn that Airbnb hosts constantly complain about Airbnb management.
@FaheemMitha yeah, I think there is no way to have a smooth running operation when you have one management unit (no matter how large) directly managing an unlimited number of people.
The Soviet Union tried that too.
The key word being "directly." If you manage a small number of people directly who each are executives with others reporting to them, it can work.
The large group being comprised of smaller groups, rather than individuals.
@Wildcard Well, Airbnb probably isn't a smoothly running operation. But I guess there are enough people who have property to rent and who are desperate for cash that it's still running, for now. As someone leaves, others enter.
