« first day (3042 days earlier)      last day (1927 days later) » 

12:41 AM
canonical VTC "unclear what you're asking" question
-6
Q: dfgd fgdfdf gdfg d dfg dfgd fgdfsdf s df

ivandfgdf gdf gdf gd dfg dfg d fgdf gd fd gdf https://unix.stackexchange.com/questions/ask sdfsdfsdfs debian. sdf sdf sdf sdf sd. sdfsdfsdf fgjdlfkgsdk gshdfkj gksjdkfg ksdhflghslkdj fgsdfg. d sfgjskdjfkg hskdjfhkg hskdjfghskdjfgjhskdfgsdf gdsfg sd ds fjhgkjsdkf gkhskdfhgksdf? dfsdflsjldjk s,sdfsdf ...

 
 
1 hour later…
1:47 AM
Actually, I just did a VTD for unix.stackexchange.com/q/498929/4671. I'm not sure why it was a VTC.
Did someone really call his or her project SLURM? Yikes.
That word belongs in a Roald Dahl novel.
 
 
7 hours later…
9:13 AM
@JeffSchaller Typical Qwerty user...
 
How much of a security risk is configuring my machine as a Bluetooth sender/receiver?
> BlueZ is the official Linux Bluetooth protocol stack.
Hmm, that's sort of encouraging. Never heard of it before now, though.
And part of the kernel too, apparently.
 
9:57 AM
@FaheemMitha 7, on some scale. That's my professional assessment. ;-)
 
@Kusalananda What scale is that, exactly?
Also, what profession?
 
On a scale from OpenBSD to Linux, I'd say "not OpenBSD" (bluetooth is not available at all).
 
BREAKING — famous Internet pundit Kusalananda assesses the security risk of configuring a PC as a Bluetooth sender/receiver
@Kusalananda so currently, enabling Bluetooth on OpenBSD is perfectly safe
and OpenBSD maintains its reputation as a safe operating system
 
They looked at it and decided it wasn't for them.
 
Bluetooth requires pairing. And apparently some kind of token is usually sent/offered.
So any real vulnerability would require a fairly clever MITM attack. Speaking as a layperson.
@Kusalananda Bluetooth is not available on OpenBSD? For reals?
@StephenKitt Would you like to offer your opinion, as an Internet pundit?
 
10:03 AM
@FaheemMitha I wouldn’t leave a device with Bluetooth open in a discoverable manner
 
@FaheemMitha For reals. They had some bluetooth support for a while but later removed it due to security concerns.
 
@StephenKitt Discoverable is bad. But it seems standard that a phone times out discoverability in a minute or two. At least my phones seem to be doing that.
@Kusalananda I see. Interesting.
 
@FaheemMitha yes
I tend to favour defence in depth, so I only enable Bluetooth when I want to use it.
(And I pay the price for that in annoyance when I want to listen to something in my car and realise I forgot to enable Bluetooth — on my phone and my car — before setting off.)
 
@StephenKitt Sounds like a good idea.
Since I was thinking about it, I turned bluetooth off on all my phones. It was on on all of them.
Does bluez have sane defaults regarding discoverability?
 
@FaheemMitha I don’t know off-hand, I use the GNOME integration, I don’t configure BlueZ directly. (I did in the past but I’ve forgotten the details.)
 
10:13 AM
@StephenKitt So you do use Bluetooth on Linux, then.
Weird, the reviews for this are full of references to Kali - amazon.in/Panda-Wireless-PAU06-300Mbps-Adapter/dp/B00JDVRCI0
 
10:26 AM
@FaheemMitha yes, I’ve been using it on Linux for quite a while
 
10:37 AM
@StephenKitt Ok.
@StephenKitt Let's hope it has sane defaults. It's hard to imagine that it would not.
@StephenKitt Do you use a USB Bluetooth adapter?
 
PRY
What does pkill -0 do, I looked into the manual but didn't get it.
 
@FaheemMitha I used to, now I only use Bluetooth on my laptop which has a built-in radio.
 
@StephenKitt Oh. For security reasons, or just happenstance?
I mean, that you don't use it on a desktop?
 
@FaheemMitha just happenstance, I used to use it on my desktop but not any more.
I still have my USB adapter somewhere...
 
USB is convenient, I guess. But I don't particularly care for having things plugged into my machine. But I don't know if it is such a big deal in practice. Might be a pain if you had to move the machine frequently.
@StephenKitt Ok. Well, maybe I'll give it a try. It's convenient/low overhead. Unfortunately, that tends to have high correlation with high security risks.
 
10:50 AM
@FaheemMitha I plug things into my screen ;-).
 
@StephenKitt Your desktop monitors?
 
@FaheemMitha yes.
@FaheemMitha how likely is it that someone can find themselves within 10m or so of your system, without you noticing?
And if you don’t notice, is the scenario any different that that person having physical access to your system?
 
@StephenKitt Some of those things have a much longer range than that. Though I wonder if, in that case, the range is settable.
There are people in the next building who might not be much more than 10 meters away. Maybe less.
 
@FaheemMitha right, so that’s part of your risk assessment ;-).
 
Though the chance that they are up for a spot of Linux bluetooth hacking, is, on reflection, perhaps not that likely.
It's actually quite hard to find people in Bombay who know anything about Unix or Unix-type systems.
Does anyone know if those bluetooth adapters have an adjustable range?
Doing a search suggests no.
 
11:06 AM
Would it be fair to assume that anyone trying to break into a Debian system via Bluetooth would have to be fairly well versed in Linux security issues?
I.e. would a Kali Kiddie cut it?
 
11:29 AM
@PRY pkill -0 sends signal 0 (null) to the matched processes. You will find what it does in the manual page kill(1): it does nothing, but performs error checking — so I think it's used for e.g. checking if a process exists.
 
@PRY kill or pkill with -0 will exit successfully if the given PID (or process name with pkill) corresponds to a running process.
In that particulac question, I see no real reason to use pkill -0, as it's even more obfuscated.
(documented in kill(1))
(eh, but not on Linux it seems. Odd)
man.openbsd.org/kill and also POSIX: "An early proposal invented the name SIGNULL as a signal_name for signal 0 (used by the System Interfaces volume of POSIX.1-2017 to test for the existence of a process without sending it a signal). Since the signal_name 0 can be used in this case unambiguously, SIGNULL has been removed."
 
 
4 hours later…
3:08 PM
Huh. Can someone remind me what "Reviewed" means? It seems someone actually "Reviewed" that garbage post from the First-Posts queue, and I'm curious what they did to cause it to be "reviewed" -- perhaps just a downvote?
perhaps a flag shrug
 
3:26 PM
@JeffSchaller Aren't first posts automatically put in a queue to be reviewed?
 
@FaheemMitha yes
I just don't know (Stack Exchange trivia) what actions result in a "Reviewed". I guess I'm hoping it wasn't a "this looks ok" action, and was instead a downvote followed by "I'm done" - but the paper trail doesn't show any of that.
 
@JeffSchaller The review history suggests that anything other than “no action needed” ends up being “reviewed”...
 
@StephenKitt whew, good. I suppose votes are anonymous, so I understand why that's not shown.
I haven't been in first-posts for a while, so my memory is extra rusty.
 
@JeffSchaller you need to go back to the “first-post first reviews” queue ;-)
“first reviews of first posts” to make it a bit more understandable
 
@StephenKitt but first, I need to...
 
3:34 PM
being able to review reviews is something I miss on occasion
one can @ a reviewer in some cases but not all
 
dang, it's too bad ed doesn't have an image/icon to use for for the community promotion ads
 
@JeffSchaller you could always ask Rob Pike for a logo ;-)
 
@StephenKitt hmmm!
or maybe just a big question mark
 
haha yes
with a blinking cursor
 
now now, a large question mark is already pushing my graphic-design capabilities. blinking is beyond me
(considering also posting a Red Hat one, though RHEL seems to be a minority around here)
 
3:50 PM
@JeffSchaller right, and it seems a bit weird to have free community advertising for a commercial distribution
(which pays my wages but nevertheless)
 
@StephenKitt good point
 
This is odd:
2
Q: Bash can't find scripts in $PATH

Modus OperandiI'm setting up a completely new, fresh Ubuntu 16.04 Minimal system, and have run into a weird problem. Even after adding $HOME/bin to the $PATH, bash can't seem to find any of the scripts in it, at least for root. I have logged out several times and even restarted the machine once. The path is b...

 
@terdon true -- 499079 ends in 9
 
argh!
:)
 
4:33 PM
@JeffSchaller You should talk to the TeX people. They will tell you that the answer is always TikZ.
 
5:08 PM
@StephenKitt running no risk of being hired as a graphical designer, but:
 
@JeffSchaller Hrm, you made me re-read that question several times now before I got it ;-)
 
@Kusalananda the distracting oddness or the question itself?
 
@JeffSchaller The oddness of the number and how it related to the question.
I'm a bit slow today it seems.
 
@Kusalananda must be that keyboard layout ;)
 
@JeffSchaller Yeah, makes reading harder.
 
5:10 PM
@Kusalananda I can't read -or pronounce- "dvorak" quickly, that's for sure
qwerty rolls right off the lips
I should get some food before more delirium escapes
 
The only word I can type quickly on a qwerty layout nowadays is "us.dvorak" :-) I need it when I upgrade my system.
@terdon Possibly a noexec mount. I left a comment asking about it.
 
@Kusalananda Wouldn't that give an error? And wouldn't that cause both /root/bin/newuser and cd /root/bin && ./newuser to fail as well?
 
@terdon Ah, I somehow miss those parts of the text.
 
@terdon not sure about the error, but yes, it would cause other invocations to fail
 
Hashing? If the old newuser command was moved?
... if there was one.
 
5:18 PM
@Kusalananda ah, or if there’s another, invalid newuser earlier on the PATH
 
@Kusalananda reboots and re-logins already, though
 
Trying to set up a test for it now.
 
joke's on us, the newuser script contains the line printf '%s\n' '-bash: newuser: command not found'
2
 
@JeffSchaller That really would be brilliant :)
 
@Kusalananda doesn't your error mis-match the OP's "command not found" versus "No such file or directory"?
 
5:25 PM
And:
> I have logged out several times and even restarted the machine once.
 
Oh, you are silly and troublesome ;-)
Yes, you are both correct.
Boo
 
:)
 
Is /etc/profile.d even used for root?
 
Should be, I don't think it's user-dependent.
But either way:
root@Pioneer ~ # echo $PATH
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games:/snap/bin:/root/bin
So the dir is in the PATH.
 
So, "stop moving administrative commands out of /sbin".
But anyway. I'm out of ideas on that one.
I still think it's a hashing issue, but I don't know why that would be. And the error message is wrong.
 
5:40 PM
I had a hare-brained idea that the she-bang line was messed up, but I can't make that work, so I'm out of ideas, too
oooph, a downvote? harsh
 
@JeffSchaller Not really. If you're referring to my deleted answer, it didn't explain the reason adequately.
 
@Kusalananda I guess I have the benefit of hindsight, knowing that you already deleted it
 
I even looked at the Bash sources to see whether they made some sort of exception for paths under /root, but couldn't find anything.
I suppose that he's not copying the script into /root/bin from somewhere else every thime he starts a root shell. That wouldbe silly.
"thime"?
Fat fingers.
 
dang, so close with a re-exec idea
 
$ newuser () { command newuser "$@"; }
$ newuser
bash: newuser: command not found
If they have something like that, maybe?
But I don't know, since the PATH seems correct.
Unless they use command -p
... which sets a new PATH.
 
5:58 PM
not close at all, nevermind
 
6:16 PM
I might very well be wrong still. They say they can't run any of those scripts.
 
6:39 PM
Great, another mystery question: unix.stackexchange.com/questions/499118
 
You can get permission denied even as root: try to write to a read-only filesystem, try to write to a special filesystem (e.g., /proc or /sys) in a not-permitted way, try to write to a network filesystem (server does permission checks, too), etc... You haven't really given us enough information to figure out which you're seeing. What is the current working directory of startscript.sh? — derobert 1 min ago
I get to work around this insanity today: gitlab.com/gitlab-org/gitlab-runner/issues/1736
 
7:19 PM
@ilkkachu the highighting should work now. Let me know if you see any other issues like that.
 
@terdon, yes, thank you!
 
Weird bug though.
 
 
5 hours later…
11:56 PM
I just had the same question-upvoting experience as Kusalananda's been seeing
 

« first day (3042 days earlier)      last day (1927 days later) »