« first day (1390 days earlier)      last day (3566 days later) » 

slm
12:50 AM
@Braiam I created an A for this. Spotify worked for me on Fedora 20
 
1:39 AM
@mikeserv, you are in a roll nowadays :)
 
@Ramesh. Thanks.
 
@mikeserv, ywc. Where do you work?
 
I don't.
@Ramesh - will you upvote that password prompt question? I cleaned it up - and it's a pretty good question. I don't care too much about votes on my answer - but the question was good in principle.
 
@mikeserv, done.
 
Thanks, sir. What are you up to?
 
1:47 AM
I never downvoted in the first place.
I hardly down vote anyone. :)
 
Yeah, I know - your comment upvote is mine.
I followed the link - empty. That was good advice.
 
yeah. sometimes it is good to provide the right pointers to OP so that they can do their research properly.
 
agreed
 
So what do you do for living?
 
Not a lot... collect military disability in the main. I also fly a kite now and then - but there's no pay in that.
 
1:55 AM
dude, are you not interested in serious jobs? I mean, I always see postings like senior linux admin all the times. I would definitely say you are a fit for it.
Sorry, by serious jobs I meant the jobs you like. :)
 
Yeah, I should look. But I don't really have any qualifications.
 
hmm, if I were to hire someone, I would just see his stack exchange profile irrespective of which stack site he uses.
 
You think?
Hmmm... I never really considered that,
Maybe the points are worth something then?
 
Not exactly the points. But I would say if you are stuck at some place, you know the right place to ask.
 
this is true.
 
2:01 AM
I have even mentioned a thank you note to unix.stackexchange.com in my thesis work :)
Thanks to the members of the Stack Exchange community
sites, especially unix.stackexchange.com for answering many questions in connection
with my thesis work.
2
 
Yeah? That's pretty warm and fuzzy.
 
 
6 hours later…
8:29 AM
@mikeserv What is?
 
6 hours ago, by Ramesh
Thanks to the members of the Stack Exchange community
sites, especially unix.stackexchange.com for answering many questions in connection
with my thesis work.
 
8:55 AM
@mikeserv Hardly what I'd call warm and fuzzy.
 
What would you call it?
 
@mikeserv Do these jobs require qualifications? They might want prior experience, but you could do some volunteer work to gain experience, perhaps.
@mikeserv Polite.
I have similar things in my acknowledgements.
Warm and fuzzy might be something like - I love you guys, I really do! You are the best!
 
Well, I get warm and fuzzy about polite. Although, I really hate that polite connotes political.
 
@mikeserv "polite connotes political." ?
Not doing the reply thingy today?
 
Come on - it was immediately below your own statement! Does context count for nothing?
 
9:00 AM
@mikeserv Sure it does. But I'm a big fan of the reply thingy. I miss it in IRC. And I believe if a feature like that is available, one should use it.
BTW, for Debian users, check out debian-handbook.info/browse/stable/index.html
If you haven't already seen it.
 
 
4 hours later…
1:06 PM
Who here uses Arch Linux? Does the X server have some kind of odd name, or is there some reason why it would not show up in the list of active processes? I'm interested because of this discussion:
Hm, ps -fC X outputs nothing; ps X and ps -fC startx do. The UID I get is my own just as I expected. But may question may have been imprecise: I'd like to know if X is run rootless or not. I will elaborate on that in my question (trying to be brief it seems I was too brief). — lord.garbage 1 hour ago
^^ Okay, never mind, it looks like it's Xorg.bin.
Or there's an Xorg and a Xorg.bin, presumably the latter is for running non-root.
 
1:29 PM
@goldilocks @strugee uses arch.
Questions don't get tweeted automatically, right? Someone has to do it? And is that person a script or a person?
 
1:56 PM
I wonder this also. I've only seen questions tweeted (I don't look often) but the other day the aviation.se twitterbot tweeted one of my answers
 
2:22 PM
@FaheemMitha I got the thank you note from you :)
 
2:44 PM
@casey Yes, I asked because a question I asked on dba.sx was tweeted.
@Ramesh Yes, I remember discussing this.
 
3:13 PM
@FaheemMitha Some people don't like the reply thingy, because it always causes a ding. Especially when its obvious.
 
@derobert Yes, I gathered that.
I suggesed a reply thingy version without the ding, but nobody seems to be interested.
 
unix.stackexchange.com/questions/147450/… .... hmmm, that seems off-topic.
 
@derobert that's unix system programming, isn't it?
 
Nah, that's "go check your C textbook".
 
@derobert Ok.
Yes, I see, it's really a C programming question, and only incidentally a system programming question.
Does U&L currently have a user migration path to SO?
Apparently yes. Voted to send it to SO, though I don't know how it will fare there. They're not the gentlest people.
 
3:22 PM
I threw on an answer, and also another close/migrate vote. At least OP will have something before being thrown to the wolves.
 
Wow, three answers almost simultaneously. Migrating seems a bit redundant now.
 
Maybe it can get a few more answers over on SO, before being closed for some reason or another...
The IP tag wiki currently explicitly says "and IP addresses". I'll remove that. Should ip be v4 only, or should it include v6? — derobert 3 mins ago
We have a separate ipv6, so I'm going to guess v4 only. Though never mind, we have ipv4 as well... — derobert 2 mins ago
 
@derobert Have the current answers completely answered the question? If so, I doubt the poster cares.
 
The poster might realize that SO is a better place for his future questions about C.
 
@derobert True.
 
3:37 PM
user image
3
LOL
 
@derobert That is hilarious.
 
Wow, this mike-m chap is really staying on top of his question.
 
Indeed.
@mike-m It sounds like you've hit an NFS bug, probably in the NFS server. Either that or filesystem corruption on the server. I doubt you can really do anything else other than wait for the NFS server admin(s) to deal with it. — derobert 24 secs ago
@FaheemMitha I'm pretty sure he's out of things to try.
 
@derobert Yes, he seems to have covered the ground quite throughly.
Network filesystems are a pain, in my experience. Also, isn't NFS supposed to be obsolete? I thought AFS was recommended instead.
 
3:53 PM
autofs right?
 
There are probably other alternatives, but this is something I know almost nothing about.
@Ramesh ?
 
by AFS do you mean autofs?
 
@Ramesh No, the Andrew File System. It's a network filesystem
The Andrew File System (AFS) is a distributed file system which uses a set of trusted servers to present a homogeneous, location-transparent file name space to all the client workstations. It was developed by Carnegie Mellon University as part of the Andrew Project. It is named after Andrew Carnegie and Andrew Mellon. Its primary use is in distributed computing. == Features == AFS has several benefits over traditional networked file systems, particularly in the areas of security and scalability. It is not uncommon for enterprise AFS deployments to exceed 25,000 clients. AFS uses Kerberos ...
They used to use it at UNC, probably still do. That is where I encountered it. But like I said, I don't like network filesystems. You're much better off using distributed version control and pushing/pulling stuff around.
One problem with something like AFS is that it is really, really, complicated.
Back in the 90s I spent a fair amount of time fooling around with AFS on Linux, and I never did really understand how it worked, except that it was really complicated, and didn't cope well with slow network speeds.
 
@FaheemMitha Well, if you have a workload that allows that. But a lot of workloads don't. E.g., large files.
 
@derobert One can use things like git-annex. I don't think the hg equivalents are comparable yet though.
 
3:58 PM
git-annex is still a fair bit more complicated, and relies on fully pulling files
 
AFS isn't magic, it still has to transfer data over the network if you want to use it. and I prefer not having this done by magic in the background.
@derobert more complicated than...?
And yes, it does require fully pulling files.
 
You have to run manual git-annex commands to pull a file local. As opposed to NFS (etc.) which do it in the background as needed.
 
Disclaimer: never used git-annex (since I don't use git) and have only read about it.
 
Haven't used it either, but that's because I almost always have a fast Internet connection. And I just run NFS...
 
@derobert True, but I guess I prefer to do things manually, and know what is going on, than rely on behind the scenes magic. If I am modifying a large file via a network filesystem, it's got to transfer all that data back to the server and make the changes correctly, right?
 
4:01 PM
E.g., I have videos on my fileserver at home. I stream them on the TV PC. The old TV PC ran off 4gb of flash... so I had videos that were larger than the total storage on the machine playing them.
 
@derobert Haven't use git-annex?
 
Yeah, I haven't used git-annex
 
@derobert ok
 
The new machine has more storage, but still there is no reason to pull the local; its on a gigabit LAN. Playing over network starts instantly; pulling 8GB would take a bit. Even over gigabit.
 
Since git-annex was written by JoeyH, one assumes it is well done. Though I wish he didn't insist on Haskell...
@derobert Well, that's a read-only situation. I was referring to when you want to write locally.
I guess I'd be concerned about file corruption issues. I never used network filesystems enough to notice anything like that, but I'd be surprised if it didn't happen sometimes.
 
4:05 PM
@FaheemMitha Isn't he also the one that lives mostly off-grid? So you'd assume its also really good at that.
 
@derobert he is, yes. I think he lives in TN, in the Blue Ridge Mountains or something.
I read his blog sometimes, via Debian Planet.
And runs everything off solar power.
Also, as you are probably aware, the author of debconf, debhelper, alien, and one of the main people behind debian-installer (was the project lead for awhile).
Recently seems inclined to hermitude, for some reason.
 
@FaheemMitha I write stuff to NFS as well. Sure, you can have corruption if you try to simultaneously write from multiple places (though NFS has locking). But you'll get a nice merge conflict if you do that with git.
 
@derobert I prefer merge conflicts to corruption. :-) The great thing about dvcs vs network filesystems is that you can do stuff locally, as opposed to doing things across the network. I hate doing stuff across the network.
 
@FaheemMitha Doing stuff across the network is fine, as long as the network is fast. When it isn't, I fix the !#@(*# network :-)
 
My point was more general, all software has bugs, and you are relying on the network filesystem software to sort stuff out for you.
@derobert Well, fast networks aren't always an option. Or are expensive if they are.
 
4:10 PM
@FaheemMitha Well, NFS is pretty well tested software. I'd be more worried if I were running AFS, Coda, etc., all those newfangled complicated ones.
And things like corruption from editing files in multiple places simultaneously don't come up much, because I haven't yet mastered splitting myself in two.
 
is autofs a feature of NFS?
 
@derobert AFS has been around for ages. At least 25 years, I'd say. They were using it in production at UNC in the 1990s. And Russ Allbery has been using it at Stanford for a similar length of time.
@derobert :-) Well, other people could be trying to edit your files.
 
True. It is probably quite stable by now.
@FaheemMitha danged crackers! Must fix the security hole, then.
 
@derobert ?!
 
Almost all my stuff on the fileservers is private stuff... No one else should be trying to edit them.
 
4:13 PM
@derobert Ok, right. I was just speaking generally.
Russ Allbery is the Debian AFS maintainer, btw. He works for Stanford.
Also has a fairly extensive blog.
Also on the Debian Tech Committtee. But you probably already know who he is.
 
At work we have file servers that multiple people use, and even the same documents. Locking (it even works between NFS and Samba, amazingly!) has proved much easier for everyone else to figure out than even Subversion.
 
And does lintian/policy work.
@derobert It works reliably, then?
 
I think it'd take several lifetimes to introduce our Windows folk to git.
 
@derobert Hmm.
 
@FaheemMitha Yes. Reliably as in it tells everyone else who opens the document 'someone else opened this first, you can't edit it.'
 
4:15 PM
But don't your guys need to have stuff version controlled? or does everyone just hack away at once?
 
Source code is version controlled.
 
@derobert Hmm, that's a crap model. That's where dvcs shines.
@derobert With subversion? How many people in your org?
 
Yeah, unfortunately with svn. It was set up before Git existed. Only three devs here, though, so its not too bad.
 
@derobert Oh, that's really small. Have you thought about transitioning to git/mercurial?
mercurial is generally less horrifying for svn refugees.
 
Yes. But getting the other two to use svn was hard enough... Switching to git would be, ummm....
But yeah, I randomly miss git features all the time.
 
4:18 PM
@derobert I've got to wonder what kind of devs they are. But maybe you are too tactful to comment. :-)
 
In the words of No. 2, that would be telling. :-P
 
@derobert No. 2?
 
The Prisoner is a 17-episode British television series first broadcast in the United Kingdom from 29 September 1967 to 1 February 1968. The premiere was 29 September 1967 on ATV Midlands and the last episode first aired on 1 February 1968 on Scottish Television. The world broadcast premiere was on the CTV Television Network in Canada on 5 September 1967. Starring and co-created by Patrick McGoohan, it combined spy fiction with elements of science fiction, allegory, and psychological drama. The series follows a British former secret agent who is held prisoner in a mysterious coastal village resort...
 
@derobert oh
 
@FaheemMitha OTOH, I do archive video files to optical media (I started with CDs, then moved to DVDs, then Bluray)... I wonder how well git-annex would handle all that...
From the examples, appears it'd need git repositories on each disc. Could be arranged for new discs, but not so much for old ones. Wonder how well it works with hundreds of repositories.
 
4:30 PM
@derobert git-annex probably has a mailing list or something. if not, there are probably other places you could talk to people.
 
I would, if it were a real problem I had to solve. I have plain-text directory listings of the discs, that works well enough.
 
Apparently not. Joey isn't really big on mailing lists. I guess he doesn't want the hassle of adminstration.
@derobert ok
@derobert ever had any contact with joeyh? He unreasonably rejected a bug report of mine for etckeeper, but i was overawed and intimidated, and didn't argue. :-)
I still think I was right, but it can be hard to get people to listen.
 
I think I've emailed him before.
git-annex.branchable.com/forum/… ... would appear to be something similar to what I was wondering
 
@derobert Yes.
Sounds like a reasonable application.
Sounds like a lot of work, though.
 
Appears he's setting up git-annex on the new disks before burning them. Which is something I could add to the Perl script which makes the disks... but it'd not handle old ones at all.
 
4:38 PM
@derobert Right. Is this something you would consider? You already use git, right?
Wonder what the story with JoeyH is. His reclusive existence, that is. I don't really know much about him.
 
He used to have a page about his reclusive existence... but I don't see that on the site anymore. Must be buried somewhere.
@FaheemMitha Seems like a lot more work than the gain is worth. grep already will tell me where a file is.
 
@derobert You mean, a "why I am doing this"?
 
@FaheemMitha I think there was some of that. Mostly what and how, though.
 
@derobert Sure. You already have this info in plain files, so yes.
@derobert Ok. I don't recall reading anything like that. However, much discussion about his batteries and so forth. Living with solar power seems quite complicated.
 
4:56 PM
BTW: That C question got migrated to SO over 30m ago, and so far has survived.
 
@derobert Ok
 
@goldilocks I do, and its just that way because the person writing that has installed a startx script. ps isn't very good at tracking process name changes after a process execs another.
it has something to do w/ process inbreeding anyway.
 
6:12 PM
hoi
 
6:56 PM
ahoy
@derobert - there's a bug in that sq(). Or rather, it doesn't handle a bug in some shells.
zsh's emulate sh and yash both seem to do field splitting out of order. So the ${1%%\'*} glob gets chewed up.
They're supposed to expand parameters first then split on $IFS but instead they generate an en extra one. I fixed it.
qchk() { $1 -s -- "$@" ; } <<-\SCRIPT
        sq() ( [ -n "${ZSH_VERSION:+?}" ] &&
                   emulate sh
            spl() { ${1:+:} return 1
                IFS=\' ; set -f ; set -- $*
                printf \'%s "$1" ; shift
                printf "'\\\\''%s" "$@"
                printf "' "
            }
        for q do spl "$q" ||
            printf "'' "
        done ; echo
        )
shift ; sq "$@"
SCRIPT
printf '<%s>\n\n' *
for sh in mk z ya da ba ''
do  echo ${sh}sh
    qchk ${sh}sh *
That's how I tested it.
<
horribly misguided filename''''''""""\\\and so on>
mksh
'
horribly misguided filename'\'''\'''\'''\'''\'''\''""""\\\and so on'
hooray!
zsh
'
horribly misguided filename'\'''\'''\'''\'''\'''\''""""\\\and so on'
hooray!
yash
'
horribly misguided filename'\'''\'''\'''\'''\'''\''""""\\\and so on'
hooray!
dash
'
horribly misguided filename'\'''\'''\'''\'''\'''\''""""\\\and so on'
hooray!
bash
'
horribly misguided filename'\'''\'''\'''\'''\'''\''""""\\\and so on'
hooray!
sh
'
horribly misguided filename'\'''\'''\'''\'''\'''\''""""\\\and so on'
So those are all the shells I know about - and they all do the same thing.
Anyway, most of that is the test stuff. The function was just replaced with:
sq() ( [ -n "${ZSH_VERSION:+?}" ] &&
           emulate sh
    spl() { ${1:+:} return 1
        IFS=\' ; set -f ; set -- $*
        printf \'%s "$1" ; shift
        printf "'\\\\''%s" "$@"
        printf "' "
    }
for q do spl "$q" ||
    printf "'' "
done ; echo
)
With that bit.
Argh.
Ok, so that testing was missing a key character - the actual filename argument.
That kinda fucks the whole thing. Anyway: the results are the same with this:
 eval [ -e "$(qchk ${sh}sh *)" ] &&
        echo "hooray!"
So it was just a typo.
And you don't get any hooray with eval [ -e notafile ]
I think I'll write another little helper function, so it can prepend like var{1...$#}= to each if you want. That should make for fairly portable arrays, maybe.
 
7:32 PM
Anyone here knows to to regenerate the local key for tripwire?
Accidentally delted it
 
@Nick Use recuva
 
@DavidFreitag found a command that regenerates the key
 
7:49 PM
15
Q: How to delete this undeletable directory?

mike-mI untarred a corrupt tar file, and managed to end up with some directory that I can not delete, If I try to delete it, it seems like it can not be found, but ls shows it's present, both with bash and with python I get similar behaviour, except right after I try to delete it with rm -rf, ls comp...

what the ^????
 
@Braiam somebody messed up baad :P
 
8:23 PM
@mike-m It sounds like you've hit an NFS bug, probably in the NFS server. Either that or filesystem corruption on the server. I doubt you can really do anything else other than wait for the NFS server admin(s) to deal with it. — derobert 5 hours ago
 
Do they really have NFS server admins?
 
Apparently they do. Of course, I suspect that admin has other responsibilities as well.
 
 
1 hour later…
9:51 PM
Is this
0
Q: Relation between the filename and inode of a file

TimFor a file, it is said that its "filename" points to its inode. Does it mean that its "file name" is a data structure that has or is a pointer pointing to its inode? It is also said that its "filename" is a string. Is this string stored somewhere (perhaps also in a file)? Are the two sayings in...

dupe of this one?
17
Q: Where is filename stored on a filesystem?

yegleWhere is filename stored on a filesystem? It's not in inode or with the actual file content since we have hard link that two filenames can point to the same inode.

 
 
1 hour later…
11:18 PM
dv-plz. This answer has a squillion upvotes, but it's wrong, it can't possibly help in this situation
 
11:35 PM
@Gilles 11, to be precise.
 
slm
@FaheemMitha - it's down to 8
 
@slm I meant, 11 upvotes, 3 downvotes. :-)
 
slm
ah
 
Wouldn't such posts be OT?
0
Q: Are there posts about good references?

TimI wonder if there have been some posts about good references (in particularly books, either sold in stores or available online) for learning OS, in general, with Linux and Unix (and Windows) in consideration Linux, for understanding how it works, similar to Andrew Taneubam's Modern Operating Sy...

 
4
A: How can I send reads and writes on a single file descriptor to different places?

Stéphane ChazelasWith socat: socat 'system:cat input.txt & cat > output.txt,commtype=socketpair' \ 'system:foo,nofork' Or even better: socat 'CREATE:output.txt%OPEN:input.txt' 'system:foo,commtype=socketpair'

nice
 
11:45 PM
I just found out Elizabeth Jane Howard died this year. Anyone read her novels?
@Gilles I'm unclear o the point of this question. can anyone elaborate? Still, nice to see a sensible question once in a while. First, why (and how) would one write output to input (fd0)?
 
@Gilles already done
 
@FaheemMitha write(0, "foo", 3)
or from a shell script echo hello >&0
 
@Gilles OK, and more importantly, why?
 
@FaheemMitha Bad design of that software? It happens.
 
@Gilles Hmm
 

« first day (1390 days earlier)      last day (3566 days later) »