search of the:   for: 
code logs -> 2009 -> Sat, 03 Jan 2009< code.20090102.log - code.20090104.log >
--- Log opened Sat Jan 03 00:00:57 2009
00:34 Derakon [~Derakon@Nightstar-4879.hsd1.ca.comcast.net] has joined #code
00:34 mode/#code [+o Derakon] by ChanServ
00:37 * Derakon snerks at the title.
00:37
<@Derakon>
s/title/subject/
01:01 You're now known as TheWatcher[T-2]
01:05 You're now known as TheWatcher[zZzZ]
02:34 Derakon is now known as Derakon[AFK]
02:48 Derakon[AFK] is now known as Derakon
04:20 Vornicus [~vorn@Admin.Nightstar.Net] has quit [Ping Timeout]
04:24 Vornicus [~vorn@Admin.Nightstar.Net] has joined #code
06:20 Vornotron [~vorn@Admin.Nightstar.Net] has joined #code
06:31 Vornotron [~vorn@Admin.Nightstar.Net] has quit [Connection reset by peer]
06:32 Vornotron [~vorn@Admin.Nightstar.Net] has joined #code
06:35 Derakon is now known as Derakon[AFK]
06:35 Derakon[AFK] [~Derakon@Nightstar-4879.hsd1.ca.comcast.net] has quit [Quit: Leaving]
06:36 Vornotron [~vorn@Admin.Nightstar.Net] has quit [Quit: This computer has gone to sleep]
08:39 KBot [AnnoDomini@Nightstar-28974.neoplus.adsl.tpnet.pl] has joined #Code
08:41 KarmaBot [AnnoDomini@Nightstar-29010.neoplus.adsl.tpnet.pl] has quit [Ping Timeout]
08:42 KBot is now known as KarmaBot
09:02 You're now known as TheWatcher
10:01 Vornicus [~vorn@Admin.Nightstar.Net] has quit [Quit: ]
11:29 AnnoDomini [~farkoff@Nightstar-28974.neoplus.adsl.tpnet.pl] has joined #Code
11:29 mode/#code [+o AnnoDomini] by ChanServ
13:48 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has quit [Ping Timeout]
14:59 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has joined #code
14:59 mode/#code [+o Consul] by ChanServ
15:01
<@Consul>
Okay, can trying to convert video as non-root in user-space somehow, someway, cause the kernel to lock up? It's happened twice now, same file.
15:01
<@Consul>
I've tried three times, the second time, it was just Devede that crashed.
15:02
<@Consul>
Running into a kernel bug somehow, maybe?
16:07
<@ToxicFrog>
IO problems can cause this.
16:08
<@ToxicFrog>
It goes to read/write the disk, hits a bad sector, gets wedged trying to recover from it...
--- Log closed Sat Jan 03 16:11:34 2009
--- Log opened Sat Jan 03 16:13:01 2009
16:13 TheWatcher [~chris@Nightstar-29731.dsl.in-addr.zen.co.uk] has joined #code
16:13 Irssi: #code: Total of 17 nicks [11 ops, 0 halfops, 0 voices, 6 normal]
16:13 mode/#code [+o TheWatcher] by ChanServ
16:13 Irssi: Join to #code was synced in 57 secs
16:24
<@Consul>
But it's happening at the same point during the conversion of one specific file.
16:24
<@Consul>
Well, one thing I haven't done is actually to try just watching the file. I'll do that tonight.
16:26
<@ToxicFrog>
Try just md5summing it or dding it to /dev/null
16:26
<@ToxicFrog>
That'll read the whole file
16:26
<@ToxicFrog>
Does it always crash at the same point in the file?
16:26
<@ToxicFrog>
Aah, yes it does.
16:26
<@Consul>
Okay, I'll give that a shot. Right this second, though, I have to go install Fedora 10 on a lawyer's computer. :-)
16:37
< simontwo>
hello there
16:37 * simontwo is new here
16:39
<@jerith>
Hey there, simontwo.
16:39
<@jerith>
Is there a simonone?
16:46
< simontwo>
jerith, on many networks there is already a 'simon', yes.
16:46
<@jerith>
Ah, right.
16:46
< simontwo>
jerith, on undernet there is also often a simon_, a simon__, a simon___, a simon____, a simon`, a simon`` and so on. but there are never any simontwos. :)
16:47
<@jerith>
Yay uniquificcation!
16:47
<@jerith>
-c
16:48 * jerith considers his ircbot a bit more.
16:49
< simontwo>
are you building one?
16:49
<@jerith>
Myself and two other people.
16:50
< simontwo>
what language doesn't yet have a functioning IRC protocol interface?
16:50
<@jerith>
Dunno.
16:50
< simontwo>
or are you rather building an intelligent set of behaviors?
16:50
<@jerith>
I started writing a bot in Erlang a while back, but got bored.
16:50
<@jerith>
We're using Python and twisted.
16:50
< simontwo>
ah, fun
16:51 * simontwo was going to look into twisted two nights ago
16:51
<@jerith>
So all the protocols we need are implemented.
16:51
< simontwo>
how far have you gotten?
16:53 jibid [~jibid@Nightstar-18491.telkomadsl.co.za] has joined #code
16:53
<@jerith>
jibid: Hi
16:53
< jibid>
Hi jerith
16:53
< simontwo>
jerith, I thought about writing a curses IRC client with twisted myself.
16:53
<@jerith>
I'm working on some kind of RPC interface at the moment.
16:55
<@jerith>
If I were to write a curses client for anything I'd try put together a decent IM client.
17:03
<@ToxicFrog>
I'd like that. All the curses IM clients currently available are pretty bad.
17:03
<@ToxicFrog>
(well, actually, I wouldn't. But the me-of-three-years-ago would have~)
17:04
<@jerith>
ToxicFrog: My reasoning also.
17:04
<@jerith>
I actually almost started hacking on the irssi codebase, but then sanity prevailed.
17:05
<@jerith>
(I was going to use libpurple.)
17:05
<@ToxicFrog>
Someday, I should read up on libpurple and write an IM module for xchat using it.
17:12
< simontwo>
ToxicFrog, I like both irssi and weechat, but I would really like something that is python scriptable.
17:16
<@ToxicFrog>
Like xchat? ??
17:21 * jerith likes running irc on a remote server.
17:21
<@jerith>
On the other hand, I prefer IM to imply that I'm around.
17:24
<@ToxicFrog>
As it happens, I am running IRC on a remote server. That's what NX is for.
17:25 jibid [~jibid@Nightstar-18491.telkomadsl.co.za] has quit [Client exited]
17:30
<@jerith>
Can't afford the bandwidth for that.
17:34
<@ToxicFrog>
Nod. It's usable all the way down to dialup speeds, but when you're on sharply metered bandwidth...
17:34
<@jerith>
I must be past my Ballmer Peak.
17:35
<@jerith>
Which is sad, considering I had one glass of red wine with supper.
17:36
<@jerith>
I'm making *stupid* mistakes.
17:36
<@AnnoDomini>
Are there smart mistakes?
17:36
<@jerith>
Like forgetting self. on a method call and forgetting self in a method definition.
17:36
<@jerith>
AnnoDomini: There are.
17:36
<@jerith>
Like the one I just made now.
17:37
<@jerith>
Which is not having imported a class I used in some copy/paste code.
17:38
< simontwo>
there definitely are smart mistakes, either because they were made to deliberately avoid having to do stuff, or by their unpredicted consequence. :)
17:39
<@jerith>
The stupid mistakes are mostly because I've been writing far too much Ruby recently.
--- Log closed Sat Jan 03 20:21:21 2009
--- Log opened Sat Jan 03 20:21:26 2009
20:21 TheWatcher [~chris@Nightstar-29731.dsl.in-addr.zen.co.uk] has joined #code
20:21 Irssi: #code: Total of 17 nicks [11 ops, 0 halfops, 0 voices, 6 normal]
20:21 mode/#code [+o TheWatcher] by ChanServ
20:22 Irssi: Join to #code was synced in 57 secs
21:30
<@Consul>
ToxicFrog: Well, running md5sum on the video file in question didn't cause any kind of lock-up or crash.
21:31
<@Consul>
I'm gonna try making that DVD again, in the hopes that I get simply another crash instead of a lockup, to see if I can figure out the actual error somehow.
21:35
<@Consul>
So if I drop off the network suddenly, that's why. :-)
22:06 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has quit [Operation timed out]
22:13 Vornicus [~vorn@Admin.Nightstar.Net] has joined #code
22:13 mode/#code [+o Vornicus] by ChanServ
22:28 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has joined #code
22:28 mode/#code [+o Consul] by ChanServ
22:28
<@Consul>
Okay, I'm seriously pissed off now.
22:29
<@Consul>
I cannot, no matter what I do, mount my other hard disk (formatted ext3, mind you) anything other than read-only, even from root.
22:30
<@Consul>
My system simply will NOT let my regular user write to the other hard disk.
22:30
<@ToxicFrog>
Check dmesg.
22:30
<@ToxicFrog>
If it refuses to mount a filesystem readwrite, that generally means that the filesystem is damaged in some way and it's disabling writes to prevent further damage.
22:31
<@Consul>
I don't see anything unusual in dmesg
22:32
<@Consul>
In fact, I don't see anything at all about sdb1 (the disk in question).
22:33
<@Consul>
Yeah, it's says absolutely nothing about sdb1.
22:34
<@Consul>
There's nothing on sdb1 I care about. I could blow away that partition and recreate it.
22:36
<@Consul>
My main concern is GParted is underreporting the size of sdb1 by about 30 gigs.
22:37
<@ToxicFrog>
What does the rest of the system say? parted, /proc/partitions, etc?
22:39
<@Consul>
Actually, it's by 20 gigs, but still.
22:39
<@Consul>
parted reports the right size.
22:39
<@ToxicFrog>
And df -h? What's the filesystem size like?
22:40
<@Consul>
That one doesn't show sdb
22:40
<@Consul>
I already deleted the partition I had on sdb
22:41
<@ToxicFrog>
...
22:41
<@ToxicFrog>
If you deleted it, what are you looking at?
22:41
<@Consul>
The unallocated space is underreported.
22:41
<@ToxicFrog>
Then why didn't you say that instead?
22:41
<@ToxicFrog>
How large is the drive?
22:42
<@Consul>
Well, when I first started mentioning it, that partition was still there, and then I got ahead of myself.
22:42
<@Consul>
320GB
22:42
<@Consul>
GParted reports 298GB
22:42
<@Consul>
parted reported 320GB
22:42
<@Consul>
dinner, BRB
22:42
<@ToxicFrog>
...
22:43
<@ToxicFrog>
GS>320 10 9 exp mul 2 30 exp div ==
22:43
<@ToxicFrog>
298.023224
22:44
<@ToxicFrog>
parted is using drivemaker's gigabytes, 1e9 bytes to the gigabyte. This is a longstanding bug.
22:44 * AnnoDomini scratches head. Is that some reverse notation?
22:44
<@ToxicFrog>
gparted is using actual gigabytes, 2e30 bytes to the gigabyte, like everything on earth except parted and hard drive manufacturers.
22:45
<@ToxicFrog>
In short, your shiny new drive that says "320GB" on the case is, in fact, 298 (and a bit) gigabytes in size, and gparted is reporting the size correctly.
22:45
<@ToxicFrog>
AnnoDomini: postscript. Uses a stack machine. RPL.
22:46
<@ToxicFrog>
Similarly, a "500GB" drive is actually around 466GB, and a "1TB" drive, you will find, actually holds 931GB of data.
22:47
<@AnnoDomini>
Uh, yeah. I thought it was so because the bastards used X instead of Xi for magnitude, to sell stuff more expensively.
22:47
<@Consul>
ToxicFrog: I had a feeling that was the issue.
22:47
<@ToxicFrog>
"X instead of Xi"?
22:47
<@AnnoDomini>
K is 1000, Ki is 1024.
22:47
<@Consul>
Oh well, as long as it's not a real drive problem, I'm okay with that.
22:47
<@AnnoDomini>
IIRC.
22:48
<@ToxicFrog>
AnnoDomini: only if you're an HDD manufacturer~
22:48
<@AnnoDomini>
Wikipedia uses this, AFAIR.
22:48
<@ToxicFrog>
(in fairness, "kilo" is meant to be 10^3, and similarly for "mega" and "giga", regardless of what units you work in.)
22:49
<@ToxicFrog>
(However, the use of 2^10 when working with bytes is well established, such that the use of strict SI prefixes on hard drives amounts to deceptive marketing.)
22:49
<@AnnoDomini>
Yeah, yeah, but that's not how bases of 2 work. :P
22:49
<@AnnoDomini>
Ninja'd. :/
22:49
<@ToxicFrog>
(and "kibi" is a fucking stupid prefix.)
22:49
<@AnnoDomini>
Kirbi. :D
22:50
<@Vornicus>
* Kirby swallows the SI unit system.
22:51
<@AnnoDomini>
* Kirby becomes what he eats.
22:51
<@ToxicFrog>
Consul: if you suspected that, why didn't you do the math and find out for yourself?
22:52
<@Consul>
Because I was busy eating?
22:52
<@Consul>
And I thought of it right as I was getting food.
22:52
<@Consul>
Besides, reparting and reformatting didn't solve the issue. My regular user still can't get write access to the drive.
22:53
<@ToxicFrog>
"your regular user". But root can?
22:53
<@Consul>
From the command line, yes.
22:53
<@ToxicFrog>
This contradicts what you said earlier.
22:54
<@Consul>
Okay, so it's a slight improvement, but it still doesn't solve my problem.
22:54
<@ToxicFrog>
And in light of this new information, it doesn't sound anything like a filesystem problem.
22:54
<@ToxicFrog>
It sounds like you're just mounting the drive incorrectly or setting the permissions on it wrong.
22:55
<@Consul>
Well, once upon a time, the automount whatever-it-is got it right.
22:56
<@ToxicFrog>
Why not just keep using that, then?
22:56
<@Consul>
Because it stopped working?
22:56
<@Consul>
It stopped getting it right?
22:57
<@Consul>
I could open up Nautilus, click on the drive I wanted, and BANG! It worked. It stopped working.
22:57
<@Consul>
And here I am.
22:59
<@Consul>
And now I get to hunt for that obscure command line option in mount that says to allow permissions for all users, or something like that, that I have no hope of working.
23:00
<@ToxicFrog>
Stopped working how?"
23:01
<@ToxicFrog>
"It doesn't work" is completely worthless as either an error report or a cry for help.
23:01
<@Consul>
It's not a cry for help, it's just venting.
23:02
<@ToxicFrog>
Anyways, odds are that just mounting it normally and then using chmod to change the permissions on the mount point will work.
23:04
<@Consul>
Nope...
23:04
<@ToxicFrog>
What did you change the permissions to?
23:04
<@ToxicFrog>
What happens when you try to write to it as a normal use?
23:04
<@ToxicFrog>
*user?
23:04
<@Consul>
Oh, chmod, not chown...
23:04
<@Consul>
Well, I just chowned it to my user.
23:05
<@Consul>
Umm, it had the normal directory permissions.
23:05
<@ToxicFrog>
I don't know what "normal" is on your system.
23:05
<@Consul>
drwxr-xr-x
23:05
<@Consul>
I hate that notation. I wish it would give me the octal version.
23:06
<@Vornicus>
d755
23:06
<@ToxicFrog>
And you did that after mounting it, and it's now owned by the user?
23:06
<@Consul>
Yes
23:06
<@Consul>
Well, I chowned it after mounting.
23:07
<@Consul>
What the hell?!
23:07
<@Consul>
Goddamn thing's working now.
23:07
<@ToxicFrog>
...this is surprising
23:07
<@ToxicFrog>
?
23:07
<@Consul>
But only from the command line. Nautilus still refuses to think it has write permissions.
23:08
<@Consul>
Oh wait, there it goes now.
23:08
<@Consul>
I did not one damn thing since I chowned it.
23:08
<@Consul>
Suddenly it works.
23:10
<@Consul>
I hate computers.
23:11
<@Consul>
Each Linux distro in existence does exactly one thing right. The trick is to find one that does the one thing you care most about, then try to fix the damn thing the rest of the way.
23:37
<@Consul>
FUCK! I'm a moron!
23:38
<@Consul>
When I made the new directory to mount the drive to, I was still logged in as root.
23:38 * Consul gives up his geek creds.
23:43 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has quit [Connection reset by peer]
23:44 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has joined #code
23:44 mode/#code [+o Consul] by ChanServ
23:44
<@Consul>
Computer locked up again.
23:44
<@Consul>
Different file this time.
23:44
<@Consul>
Clearly, it's not a hard drive issue.
23:45
<@Consul>
Can I fsck a partition I have mounted at the moment?
23:47
<@Consul>
Apparently not...
23:49
<@Consul>
*sigh*
23:49
<@Consul>
And stupid me can't remember how to force a check on bootup.
23:50
<@ToxicFrog>
tune2fs, if it's ext2/3
23:51
<@ToxicFrog>
something like, say: sudo tune2fs -C 999 /dev/whatever
23:51
<@Consul>
What does that accomplish?
23:51
<@ToxicFrog>
-C sets the number of times it's been mounted since last boot.
23:51
<@ToxicFrog>
Er, since last check.
23:52
<@ToxicFrog>
By default, ext2/3 filesystems are checked periodically, typically every 20 mounts or whatnot.
23:52
<@Consul>
Do I need to be concerned that the man page says nothing about ext3?
23:53
<@ToxicFrog>
ext3 filesystems are just ext2 filesystems with a journal.
23:53
<@ToxicFrog>
Anything that works on ext2 will also work on ext3, it just won't make use of the journaling features.
23:53
<@ToxicFrog>
Anyways, use: tune2fs -l /dev/whatever to get the current settings
23:53
<@ToxicFrog>
-c n to set the max mounts between checks to n
23:53
<@ToxicFrog>
And -C n to set the number of mounts since last check
23:54
<@ToxicFrog>
If mounts since last > mounts between, it'll be checked on next boot.
23:54
<@ToxicFrog>
Also, just because it's a different file doesn't mean it's "clearly not a hard drive issue"
23:55
<@Consul>
Except I moved everything to a new drive...
23:55
<@Consul>
Hence the entire adventure with the mounting.
23:55
<@ToxicFrog>
Aah.
23:56
<@Consul>
Why is root's max mount count set to -1?
23:57
<@Consul>
Does that turn the periodic fsck off?
23:59
<@Consul>
Okay, reboot time, to let fsck run its course.
23:59 Consul [~consul@Nightstar-2535.dsl.sfldmi.ameritech.net] has quit [Quit: Leaving]
--- Log closed Sun Jan 04 00:00:23 2009
code logs -> 2009 -> Sat, 03 Jan 2009< code.20090102.log - code.20090104.log >
 search of the:   for: