« first day (2017 days earlier)      last day (3231 days later) » 

00:00
@Gilles Thanks.
Though in this case, the output is not from a file.
@Gilles weird -- why did it work when I did: array2=( $(printf "%s\n" 1 2) )
@JeffSchaller because the only whitespace there is newlines
try: with_spaces=($(printf '%s\n' 'foo bar'))
and with_wildcard=($(echo '*'))
true; I assumed Faheem's output was representative
(no spaces or wildcards)
@JeffSchaller Well, I'm solely interested in my usecase. Not in the most general solution possible.
Ahhh, I see -- I misinterpreted your statement. I agree 100%
reworded clarification might be "that only works when the output is newline separated and has no whitespace"
(...or wildcards)
00:05
It's certainly better to have a solution that works with whitespace.
@Gilles I know you're much busier now that Unix.SE exists, but if you'd like to chuckle at (and provide any critiques on) unix.stackexchange.com/questions/276607/… I would appreciate it :)
@FaheemMitha mapfile -t lines < <(printf "%s\n" "1 2" 3) # looks safer, then
substitute borg... for printf... of course
-t removes the trailing newline from each line, if that's what you want
Hi
Is unix stackexchange right place to ask about gtk3?
I have problem where windows are rendered afrer they are fully drawn, I want to enable window redrawing before full render is done
it makes gtk3 very unfriendly when using over NX
gtk2 is ok
@AndrewSmith it's not off-topic, but judging from the tag's usage history (45 questions from ~90,000 in total) you may not get many bites
ehh
by the way, QT5 is also very slow over NX, but for example, LibreOffice is OK as it seems using GTK2 widgets
@JeffSchaller sorry, a randomized progress bar simulation fails to excite me
@JeffSchaller by the way, unix.stackexchange.com/review/suggested-edits/136856 is plagiarized, please check when reviewing tag wikis, it's a very common problem
(I think I let through another by the same author earlier, sorry)
00:18
@Gilles auuugh, good catch; thank you. Straight from Wikipedia, I assume?
@Gilles thanks for the feedback! It's not innovative, but I wondered if it could be a generic "answer" for people who insist on "progress bars" for apps that don't have them
@JeffSchaller no, some other random site (which may be an earlier version of the Wikipedia article or the project description, I didn't check)
@JeffSchaller there's pv if the data comes from a pipe, and I think trickle can show progress too
in general, you can't do anything useful, because only the application knows when it's going to stop
true enough; I'd just seen enough half-baked ones that I wanted to whip up something posix-ly usable. I guess emulating a useless Windows progress bar isn't a high bar.
alright, I've caused enough trouble for today! g'night!
 
5 hours later…
06:04
@JeffSchaller Yes. Thank you.
 
6 hours later…
11:36
So, I'm looking at:
13
A: Store output of command into the array

glenn jackmanIf you just want the numbers at the end of each line: numbers=( $(pdc ... | grep -oP 'okay.+?\K\d+$') ) If you want to store each line into the array mapfile -t lines < <(pdc ...) To retrieve the data from the arrays: for (( i=0; i<${#numbers[@]}; i++ )); do echo ${numbers[i]}; done echo p...

Is lines an array? It looks like it.
12:31
@FaheemMitha yes
@JeffSchaller Ok. Thanks.
12:43
So, I've got the following:
mapfile -t mailarchives < <(borg list --short '/media/faheem/My Passport/backup-Mail')
for i in "${mailarchives[@]}"
do
    borg extract -n .::$i
done
This will be way too expensive to do for all the values in mailarchives in general. I think I'll just do the last one.
13:10
@FaheemMitha you could compare $i to a date string to just get certain years/ etc
@JeffSchaller That's true, I could.
Or do random spot checks.
But this is all intrinsically unnecessary with a deduplicating backup. You just need to check the chunks.
13:29
@FaheemMitha yay
sadly, only for xenial
checking for BABL... yes
checking for GEGL... no
configure: error: Package requirements (gegl-0.3 >= 0.3.0) were not met:

Package 'json-glib-1.0', required by 'gegl-0.3', not found

Consider adjusting the PKG_CONFIG_PATH environment variable if you
installed software in a non-standard prefix.
oh joy, gimp apparently FTBFS.
anthony@Watt:/tmp/gimp-2.8.16$ dpkg-checkbuilddeps ; echo $?
0
grumble
obsolete build-depends
trying again after installing libjson-glib-dev, guess I should report a bug if that works...
@derobert Where? That should be impossible on stable.
On backports, that's not unusual.
@FaheemMitha testing/unstable
13:37
@derobert Ah. Then join the club. Happens all the time.
I just mentioned it a few days ago. In the context of Okular, I think.
Are you trying to backport?
@FaheemMitha Not that often, actually. You were trying to build a unstable package on stable... I'm not backporting it.
@derobert Oh, you're building on unstable?
Sorry, I just assumed you were using stable.
@derobert wait, how the heck that dependency goes? gimp asks for gegl 0.3.0 or is json-glib?
And yes, I don't run unstable.
@FaheemMitha testing/unstable mix, but mostly testing
13:39
@Braiam How goes it? Haven't seen you around much recently.
@derobert Ok.
Unstable just seems like more unstability than I want. I'm not sure about testing.
I've run testing close to releases before.
@Braiam it asks for gegl, but apparently that requires libjson-glib-dev... I think pkg-config has its own dependency thing.
@derobert then that's weird... gegl configure.ac files makes no mention of libjson-glib
ok, I was looking for libjson, turns out it's called just json :/
$ pkg-config --print-requires gegl-0.3
gobject-2.0
gmodule-2.0
gio-2.0
json-glib-1.0
gio-unix-2.0
babl
14:40
In case anyone is looking for a backup solution to try, Borg Backup is looking good.
It's a fork of attic. It was forked last May. But since then it's been noticeably gaining traction.
15:10
So, on #borgbackup:
> 20:39 <Ape> faheem: mdadm has sync_action check that will read all the data and verify it
Does anyone do that? And if so, what's a good way to do this. And will it kill my disks?
15:45
Hmm, looks like Debian, at least, already does this.
16:39
@FaheemMitha Everyone sane does that routinely
@derobert Glad to hear it. Where "everyone" means distributions, I assume.
@FaheemMitha Well, if your distro doesn't, you have to... because otherwise bad blocks go silently uncorrected. Which becomes a problem if a disk fails.
@derobert Yes, so the Borg folks said.
@derobert Are you ready to be assimilated yet? Resistance is futile.
it looks to not do a lot of critical stuff Bacula does. At least for work. Like Windows.
@derobert Such as?
You could open issues. If not already open.
16:48
Windows being the first thing I notice.
Is Bacula deduplicating?
Nor does it do tape, I expect.
No, bacula is not deduplicating
@derobert You mean Windows support?
Yes. We have Windows machines I have to back up
Well, they're aware of those limitations, I expect. Though I don't know if they care about tape.
16:49
And it would appear the Borg backup has a model where the client machine is trusted, the backup server less so. That's the opposite of at work.
@derobert yes, that seems to be a frequently requested feature.
I think they'll implement that eventually.
The problem with free software projects is that they die.
If they stay alive, stuff gets done.
Indeed. If the client gets compromised, it's essential that it not be able to delete the backups from the backup server...
As I said earlier, they seem to be getting decent pickup. Which is by no means a given.
@derobert Quite. I don't know if there is an issue for that.
You're talking about a pull from the server model, right?
There was a thread on reddit where more than one person mentioned that.
@FaheemMitha yeah, or it seems that backup works by the client ssh'ing to the server. At least from a very quick read.
@derobert Yes, it does.
 
4 hours later…
20:43
Hi, Is there any part of stack that is just intended to server infrastructure?
Thanks @Fabby
the second one it's the correct
@c1b3r yeah, sorry for the first one.
How does one check whether module-stream-restore is loaded?
21:00
modprobe --showconfig | grep stream | grep restore
(maybe: I don't seem to have it running on Ubuntu 14.04 LTS.)
@Fabby I don't think that's a kernel module.
It's some Pulseaudio thingy.
21:35
The list of DPL candidates in no particular order: debian.org/vote/2016/platforms
This is a very elementary question, but it looks like cron is using an old version of a shell script even after I changed it.
This makes no sense, of course.
@FaheemMitha pacmd or pactl or whatever it's called can list loaded modules. (No time to look it up, got pinged as I'm preparing to leave...)
21:54
@derobert Ok. Thanks.
 
1 hour later…
cas
cas
23:17
anyone know how i can flag a tag for deletion? someone has created /memchashed and /memchached, neither of which should exist. /memcached is the correct tag.
@cas remove it from the post
unused tags are cleaned automatically
cas
cas
ok. i have removed it from the post. didn't realise they'd be auto-cleaned. wanted them deleted so they didn't hang around forever.
67
Q: How can we get rid of misspelled and unused (or "zombie") tags?

MottiDuring the re-tagging of questions, tags sometimes become orphaned from existing questions. Are these zombie tags ever removed from the tags list? What if a tag is misspelled and needs to be removed? How do we get rid of it? Return to FAQ Index

They will be gone in a few hours.
@derobert If you want to take a look,
1
Q: Mint 17.3 Not Playing Sound

Hunter PetersMy Toshiba Satellite A105 is running Linux Mint 17.3. It will not play any sound through the onboard hardware, and when I look into the sound menu, it only displays a dummy output. I have tried all the other solutions I found and spent nigh on 5 hours scouring Google, but nothing has worked thus ...

It seems he doesn't have the ~/.pulse directory, but does have PA running.
Unfortunately the poster's total rep is under 20, so I can't get him into chat, which would make things easier.

« first day (2017 days earlier)      last day (3231 days later) »