« first day (1453 days earlier)      last day (3509 days later) » 

12:48 AM
Yay! I now have >10K on 4.5 (I have access to the 10K tools on bio beta) sites. Huh? What do you mean "get a life"? Who said that?
 
1:48 AM
@Gilles in SU they decided just to get rid of them all
 
slm
@Gilles yeah vendor tags are generally useless.
 
I'd argue for , and though. All three have specific Linux issues/associated tools and experts.
 
slm
I just got rid of the 7 Q's that were using the lenovo tag
the problem w/ thinkpad is that there are multiple models
ideapad, thinkpad, they're all the same hardware, better to just mention the model info in the Q and not get into tagging them
nvidia and ati are ok, I agree w/ you there
@terdon - what's 4.5?
 
@slm I don't have a very strong opinion on this. It's just that between thinkpad wiki and the various (at least 3 I can think of) thinkpad specific packages and tools, there is such a thing as a Linux on thinkpads expert.
@slm On beta sites, the 10k equivalent threshold is lower and I've passed that on Biology.
 
@terdon mmm... as thinkpad hardware?
I know I created alienware-backlight in AU, because the backlight of the Alienware are custom
 
1:58 AM
@Braiam No, I guess. Well, the track-clit thing might be.
But there are at least 7 specific packages in my Debian's repos for thinkpads. Plus some others that come bundled with other packages:
 
I've never used a thinkpad, so I don't know if there's anything custom made... for example, I would use hp-scanner or hp-printer for HP's scanner or printers
 
$ apt-cache search thinkpad
acpitool - command line ACPI client
acpitool-dbg - command line ACPI client (debug)
gkrellm-thinkbat - ThinkPad laptops battery status indicator for GKrellM
hdapsd - HDAPS daemon for IBM/Lenovo ThinkPads and Apple iBooks/PowerBooks
pidgin-blinklight - Blinks your ThinkPad's ThinkLight upon new messages
thinkfan - simple and lightweight fan control program
tp-smapi-dkms - ThinkPad hardware/firmware access modules source - dkms version
tp-smapi-source - ThinkPad hardware/firmware access modules source
That would suggest at least some specific hardware.
 
@terdon "IBM ThinkPad(tm) special keys" should be about thinkpad-keyboard, no?
 
tpb?
 
yeah
what I want to prevent is "why X company is ok, yet Y is not, MEANIEE!"
 
2:05 AM
True, that can be a problem.
 
slm
I have 1/2 dozen thinkpads
 
I have one. Sigh. And I don't like the new ones. They changed the keyboard!
 
slm
me neither, I was looking for a TP keyboard for my desktop
they have them
I love the keyboards, my samsung is an awful keyboard IMO
 
I'm working myself up to buying a CODE. The idea of spending that much money on a keyboard throws me off a but but that sounds like one hell of a good keyboard.
 
If Jeff touched it it should be good.
 
2:08 AM
It should indeed.
 
slm
$170 wow
 
Yeah.
That's about the going rate at that level apparently.
I've never bought a keyboard for more than, dunno, $40 or so. Usually a good deal less.
None of them felt like the thinkpad though which tought me the joys os a good keyboard.
 
I think I know someone in RA that has one (a CODE, I mean)... maybe allquixotic
 
I've never used anything other than a stock HP for anything like an extended period of time. I rather like my Apple'ish flat laptop one. What am I missing out on?
 
slm
yeah I was desperate when I was in FL for training and went one night to officemax to buy anything. I bought a logitech k520 which is ok, but sucks compared to a TP kb
 
2:14 AM
@Seth Can't really describe it. Solidity and responsiveness. And anything I say sounds like "yeah, whatever, it's a keyboard" and then you try a good one and understand.
 
slm
The pointing device and the back/forward along with the spacing are what I miss most
the samsung and macbook keyboards are very similar so I get why ppl that want/like a mac like them. I think the TP's were trying to cater to that crowd w/ the new ones
 
@terdon mmkay. adds to bucket list: Try mechanical keyboard
 
@Seth Haven't tried a mechanical one either. The only good one I know is the thinkpad's which isn't mechanical.
 
I normally don't like laptops kb at all... I keep felling that they are too low, or something
 
@terdon Oops. Close enough I guess :P
@Braiam I like mine, but I'm told I type wrong, and I know it's true.
 
2:26 AM
@Braiam That's 'cause normally they suck.
 
btw, @Braiam, you have some Ubuntu packaging experience, can you figure out from context what is missing here?
About page 6: http://packaging.ubuntu.com/html/packaging-new-software.html I used the guide for some packages, but made always one mistake because something non-obvious is missing. bzr dh-make creates the file source/format for you but doesnt add it to the bzr repo. Unfortunatally the simple packages (like the hello thing, and other simple ones) build perfectly. Especially if there are no patches needed (so tyou basically don't need quilt). Everything works fine: commiting, pushing, building, upload to PPA, linthian. No errors anywhere. Only the first reviewer notices that the file is missing.
Ubuntu Packaging Guide
Low / Triaged
I guess the instructions don't commit the source, but I'm not sure the proper way to correct that, and I don't really want to just start randomly throwing things hoping I hit the bull's eye.
 
 
2 hours later…
4:00 AM
@slm congrats on the illuminator badge. It was confusing that the three tiers of badges were assigned in reverse order.
 
slm
@Anthon - thanks. Nice to get something first for once 8-)
 
Yeah I know how that feels >:-)
 
@Seth "bzr dh-make creates the file source/format for you but doesnt add it to the bzr repo. " this? I never used bzr's dh-make, so I will need to experiment
 
 
3 hours later…
6:47 AM
@terdon is about a product line, that's fine. It's tags about a company that are generally meaningless
or about an organization, like
 
7:13 AM
Microsoft is keeping with the tradition of Windows version numbers not in sequence, seems that 7→8 was a fluke: 3, 95, 98 2000, XP, Vista, 7, 8, 10. It's often said that the odd-numbered versions are the good ones… is 10 odd or is that an announcement of the quality to expect?
 
7:28 AM
@Gilles: I think maybe they don't want number 9, like many country in the West don't like number 13
Interestingly that the start menu of Windows will come back :)
 
7:44 AM
@Gilles If we're talking about MS, the term "good" must have been used in a different sense than the usual.
@terdon You realise that Stack Exchange could disappear overnight, right? Or drastically change its form, if someone else bought it.
 
@Gilles I think they got tired of everyone complaining about every second release and decided to go for a consistent lack of quality rather than an inconsistent one
also, good morning
or at least morning
 
8:08 AM
Hey @JennyD
Does anyone have a better suggestion for seeing network traffic than iftop?
 
8:44 AM
@FaheemMitha ntop? tcptrack?
 
9:07 AM
@JennyD Are those better?
 
 
2 hours later…
11:21 AM
@FaheemMitha I think it depends partly on your personal preference and partly on what exactly you're looking for
 
slm
11:32 AM
@FaheemMitha - what does better mean? BTW I like nethog, shows traffic centered around the app that's using it.
 
 
2 hours later…
1:18 PM
@slm Hmm, i guess i don't understand iftops display so well. A more idiot-proof app would be handy.
@JennyD True.
 
1:37 PM
I should have said something like - do you prefer something else to iftop?
@JennyD Not a good morning?
 
@Gilles meh, at least they dropped that duality between apps for the new UI and the old one
 
slm
wow the Q's are sooooo horrible this last few weeks....
 
@slm September :(
 
slm
There even more awful then I remember last yr
 
1:52 PM
maybe is time to go back and review old unanswered question
 
slm
I have been
I got so sick of these new ones that I was doing that last night
Been working on this one
5
Q: Open signed PDFs in Linux

AgentTuxWe have some PDF files which are secured by a .pfx certificate. On Windows we can use Adobe Reader to read these PDF files (with previously imported certificate). I've googled for a long time but I can't find a PDF reader which can open this.

cannot find any such app for linux
so the answer is...there isn't one
still looking
 
Can someone push
1
Q: Select certain column of of each file, paste to a new file

Jun ChengI have 20 tab delimited files with the same number of rows. I want to select every 4th column of each file, pasted together to a new file. In the end, the new file will have 20 columns with each column come from 20 different files. How can I do this with Unix/Linux command(s)? Input, 20 of this...

out of the reopen review queue? It only needs one more vote to be reopened.
 
slm
2:08 PM
@Anthon done
 
2:20 PM
@slm thanks. Strange that your reopen vote doesn't show up in the queue. Or did you directly go to the question (that way you'll never get the last steward badge).
 
slm
went to the Q and reopened
yeah I'll do that next time
 
@Anthon I know dammit! I do most of my review work on the questions directly :(
 
@terdon, @slm have you seen that another tenacious badge has been awarded? I'm considering upvoting some answers just to prevent the first Unsung Hero on the site.from happening :-)
 
Should this edit be approved?
 
@ramesh no that should be a comment
 
2:30 PM
@Anthon Thanks. I rejected it.
 
slm
too radical
 
@ramesh so did I
 
can we close this one as a dup of this one? slm's answer has more tools.
 
@Ramesh mm... nope
 
slm
2:48 PM
@Anthon yeah unsung hero, that's one way of putting it.
 
Is there any equivalent like wajig in fedora based systems?
Fedora has whatprovides.
 
slm
wajig?
 
@slm I had to google it en.wikipedia.org/wiki/Wajig
 
slm
me 2
smart had features similar to that. I used to use it but haven't in ~2yrs now
dnf.baseurl.org never heard of it
 
3:06 PM
@Anthon Heh, we don't have one? Cool!
So, why is my php file not executed by my freshly updated apache server?
A test file with phpinfo() works as expected, the static HTML parts of the page do also, yet not the php bits.
Anyone have any idea what I should check and where?
Hmm, pointing the browser to the actual php file in the cgi-bin directory works. However, including it within index.php doesn't. I'm confused.
 
what happens when you do php filename.php in your cli?
 
@Ramesh The html contents are displayed. The php is shown as code and does not seem to have run.
 
@terdon does this answer help your situation?
 
3:22 PM
@Ramesh YES! Thanks, I was using <? instead of <?php.
 
@terdon cool. you are welcome.
However, am surprised that I do not have the application type php in my web server. But still it works.
 
Hello, everyone. Back at work for today and tomorrow... then, another con!
 
@derobert how about Comicon?
 
3:37 PM
@FaheemMitha Never been to Comicon. Any of them...
Otakon was already too big. Can't imagine how bad Comicon must be.
 
Now, why is my include ./cgi-bin/foo.php line running foo.php instead of just making the contained functions available?
 
@terdon In php? That's what include does, I believe.
Include is like copy & paste.
oooh, vim is going to start soon
 
vars.php
<?php

$color = 'green';
$fruit = 'apple';

?>

test.php
<?php

echo "A $color $fruit"; // A

include 'vars.php';

echo "A $color $fruit"; // A green apple

?>
@terdon ^^^^^^^^^^
 
Nah, it was the same issue. I was using <? instead of <?php which is why the functions were being called instead of just defined.
grumble grumble, changing formats, grumble, grumble, php sucks.
3
 
4:04 PM
@derobert ok
how can i list available/registered swap?
 
free?
 
@FaheemMitha cat /proc/swaps
 
grep SwapTotal /proc/meminfo
 
@FaheemMitha swapon -s
 
4:10 PM
@derobert Thanks, that's good.
@Braiam thanks, that's really easy to remember.
You guys should add something to
9
Q: How do I find out if I have a swap partition on my hard drive?

dswhite85I used the 12.04 live cd to install Ubuntu over my Windows 7 partition and deleted everything so I just have Ubuntu on my laptop. But since during the installer I chose the simple "erase entire disk" option, did the installer create a swap partition or is that something I should've done with the ...

these answers suck, predictably.
free -m is the only reasonable answer, and has 1 vote. I added 1 more.
 
@FaheemMitha what's wrong with izx's one?
 
@derobert oh, i see he mentioned swapon -s. I only saw the graphical thingy.
 
its Ubuntu, the graphical thingy is a perfectly reasonable answer
 
slm
yeah what's wrong with the graphical A?
 
@slm is not universal...?
 
4:29 PM
Sigh. @derobert @slm sorry, i'm not going to go into that.
 
If it were here, I'd say there is a problem with that answer. But it's not. It's on AU.
@FaheemMitha well, the obvious answer is add more swap (sorry, to your deleted message)
 
Any ideas about how to handle an process that the OOM killer thinks needs to die? I'm running a memory intensive process.
The most recent invocation is:
12:27 < faheem> Oct 1 21:55:39 orwell kernel: [1054941.260318] Out of memory: Kill process 18045 (clang) score 315 or sacrifice child
12:27 < faheem> Oct 1 21:55:39 orwell kernel: [1054941.260322] Killed process 18045 (clang) total-vm:6501284kB, anon-rss:6376296kB, file-rss:0kB
@derobert Yes, someone else suggested that. I have 16gb of ram, and 4 gb swap.
 
Add more swap. Or set its oom score adjust and then maybe the OOM killer will kill something else instead.
 
seems like it should be enough.
@derobert can i adjust per process?
 
yes
 
4:32 PM
hmm, i wonder if that is safe. how much additional swap do you suggest?
 
6
Q: How to set OOM killer adjustments for daemons permanently?

gertvdijkRunning some Linux servers with single or just a few vital system service daemons, I would like to adjust the OOM killer for those daemonized processes in case something odd happens. For example, today some Ubuntu server running MySQL got a killed MySQL daemon because tons of apt-checker processe...

@FaheemMitha Keep adding swap until you stop having OOM.
that question mentions one of the files you can change in proc
 
@derobert ok, thanks
 
@FaheemMitha kernel.org/doc/Documentation/filesystems/proc.txt section 3.1 documents it
 
@derobert Thanks. Looking now.
 
hmm, is oom_score calculated based on some priority of the running process?
 
4:36 PM
I don't want to crash my machine, though. So, proceeding cautiously...
 
This says oom_score is calculated based on some heuristics.
 
Well, the cautious approach is to add more swap. Or to figure out why clang is using so much memory; maybe you should compile that file with less optimizations?
Or quit some other program that has a swap+mem usage
 
One other tweak which is applied is to add the value stored in each process's oom_score_adj variable, which can be adjusted via /proc. This knob allows the adjustment of each process's attractiveness to the OOM killer in user space; setting it to -1000 will disable OOM kills entirely.
 
All good suggestions. I'll consider adding more swap.
That is the brainless way to do things, and has much appeal.
@Ramesh True, but I wonder if that is safe, though.
even disabling for a specific process, assuming i could do that, might be unsafe.
 
yeah particularly if you have memory leaks in the user program, then definitely it would be a risk to disable it.
 
4:39 PM
@FaheemMitha Well, the system is OOM. So the OOM killer is going to kill something. If you disable it for clang, it'll kill something else.
Eventually, it'll OOM-kill everything else on the box
And if it runs out of things to kill, it probably panics
 
@derobert will it kill init as well?
 
@Ramesh I suspect init is special-cased not to be killed, as that'd instantly panic the box.
 
@derobert I thought the OOM killer acts if the memory usage of a process is rising too fast. Defcon 4 or similar.
So, before the machine actually runs out of memory.
 
@FaheemMitha It runs before the machine actually runs out of memory (because actually running out = panic) but not before its almost out.
If you watch free/top as right before its killed, the machine should basically be out of memory... I forget what the threshold is, it's a fairly small amount left.
 
5:03 PM
@FaheemMitha just curious, is a compiler or debugger?
 
@Braiam is what a compiler or a debugger?
@derobert I see. Sounds scary.
Maybe I should exit chrome. or kill more windows/tabs. I've got an obscene number of tabs open.
 
@FaheemMitha the problematic proccess
 
@Braiam Oh, it is a CL implementation build.
 
@FaheemMitha That would probably be fairly effective at freeing memory. Web browsers sure seem effective at using tons of it.
 
@derobert They do, unfortunately.
Written, improbably, by a Chemistry professor in Philadelphia. Temple.
Currently, we are sitting in #clasp in freenode.
 
6:04 PM
I'm getting permission denied errors (errno: 13) after updating my Linux box and attempting to start the mysql service. The files it complains about are owned by the mysql user. It occurred to me that perhaps it's running with a different user name. How can I check what the mysql user's username is?
 
@terdon if you cannot find it by glancing at the names in /etc/password, it is probably in something like /etc/default/mysql or in /etc/init.d/mysql
 
@Anthon Thanks, Never mind, the problem was elsewhere, I figured it out.
 
@terdon, you can probably try a chown -R mysql:mysql /var/lib/mysql
@terdon related to SELinux?
 
@Ramesh I'd done that. It's just that the update overwrote my.cnf so I had to set my default paths again.
 
cool. Somtimes, with selinux it won't allow to change the data directory path.
 
6:08 PM
That might be the cause of my next problem...
141001 20:07:49 [Note] Server socket created on IP: '10.1.3.30'.
141001 20:07:49 [ERROR] Can't start server : Bind on unix socket: Permission denied
141001 20:07:49 [ERROR] Do you already have another mysqld server running on socket: /var/run/mysqld/mysqld.sock ?
141001 20:07:49 [ERROR] Aborting
And no, at least as far as I can tell, there is no other mysql instance running.
 
I believe you can specify some bind-addr in your my.cnf.
That might fix this issue.
port=3306
bind-address=0.0.0.0
Do you have them in your my.cnf file?
 
port = 3306
bind-address = 10.1.3.30
Trying 0.0 now
Nope.
 
Not sure if this might help.
 
141001 20:15:57 [Note] Server hostname (bind-address): '0.0.0.0'; port: 3306
141001 20:15:57 [Note]   - '0.0.0.0' resolves to '0.0.0.0';
141001 20:15:57 [Note] Server socket created on IP: '0.0.0.0'.
141001 20:15:57 [ERROR] Can't start server : Bind on unix socket: Permission denied
141001 20:15:57 [ERROR] Do you already have another mysqld server running on socket: /var/run/mysqld/mysqld.sock ?
141001 20:15:57 [ERROR] Aborting
That can't be right. Shouldn't it resolve to localhost or 127.0.0.0 or something?
@Ramesh Yeah, thanks, I tried that but no.
 
Did you try rebooting the machine? It worked for me once :)
 
6:18 PM
Trying now. This had better not work...
Nope. OK, time to post a question I think.
 
@terdon, dba chat is friendly like us. You could try there as well.
 
@Ramesh I asked a quick question a second ago but got only one reply by a guy telling me he's not a mysql person. I might ask there again if I get no help from here. I just posted a question.
 
@terdon updates should not overwrite config without asking. Is this Debian?
 
6:37 PM
@FaheemMitha No, the bleedin thing is a bleedin Ubuntu.
Not my choice.
And I know, that might have been my fault. I was doing many things at the same time and I do remember getting a message about doing something or other after the upgrade which I forgot to do. That might be my problem.
 
is there some apparmor thingy?
I see it related to ubuntu. Not sure if that might be a case.
 
@terdon The original file may still be there. In debian at least, if you choose to replace, it is saved as like dpkg.orig. have you checked?
 
@FaheemMitha Yes, and that one still fails with the same error.
 
@terdon huh
 
Yeah...
 
6:46 PM
[mysqld]
datadir=/var/lib/mysql
port=3306
bind-address=0.0.0.0
socket=/var/lib/mysql/mysql.sock
user=mysql
# Disabling symbolic-links is recommended to prevent assorted security risks
symbolic-links=0

[mysqld_safe]
log-error=/var/log/mysqld.log
pid-file=/var/run/mysqld/mysqld.pid
 
What's that?
 
@terdon, this is a working mysql configuration file in one of my servers.
I see datadir is listed as /data/database in your configuration file.
 
@Ramesh I tried it, no difference.
 
Have you changed permissions for that datadir?
 
@Ramesh Yes, 'cause that's where my databases are.
@Ramesh Yes.
Right, I'm off. 9 hours in the lab is enough. See you all later.
 
 
4 hours later…
10:51 PM
A Microsoft developer explains in a Reddit thread why Windows 10 is called Windows 10. http://t.co/qJFi8yJEJg
@Gilles apparently was a pragmatic solution ^
the decision of using 10 instead of 9
 

« first day (1453 days earlier)      last day (3509 days later) »