[00:00] <jaka> k
[00:02] <arand> cjwatson: thank you, happy that it's being looked at, happy to provide more info if needed.
[00:03] <JanC> jaka: the official Gtk/GNOME/etc. channels are on irc.gimp.net, for the Ubuntu GNOME desktop, #ubuntu-desktop on this network is the right channel though  ;)
[00:03] <Keybuk> pitti: around?
[00:05] <jaka> ok
[00:05] <jaka> cool
[00:05] <jaka> thanks
[00:16] <Keybuk> eep
[00:16] <Keybuk> with sreadahead, 90s
[00:16] <Keybuk> with sreadahead modified to stay in the foreground, 73s
[00:16] <Keybuk> *without* sreadahead, 71s
[00:16] <Keybuk> !
[00:17] <ogra>  bah
[00:17] <ogra> throw it away
[00:18] <lifeless> bam!
[00:18] <lifeless> ssd ?
[00:18] <Keybuk> lifeless: slow laptop hard drive
[00:18] <lifeless> interesting
[00:18] <lifeless> btw
[00:18] <lifeless> have you considered getting blktrace traces and analysing for seeks ?
[00:18] <Keybuk> (ureadahead, for those paying attention, 57s :p)
[00:18] <ogra> whats *u*readahead ?
[00:19] <Keybuk> ogra: my little project
[00:19] <ogra> ah :)
[00:19] <ogra> cool
[00:19] <Keybuk> sreadahead -> super readahead
[00:19] <ogra> he
[00:19] <Keybuk> ureadahead -> über readahead

[00:19] <ogra> :)
[00:19] <ogra> like über splash ?
[00:19]  * ogra ducks 
[00:20] <Keybuk> I thought that one was "userspace" :)
[00:20] <Keybuk> still not really maxing out the disk on throughput though :-(
[00:21] <wgrant> Maybe we need Unicode support in package names...
[00:21] <cjwatson> "ultra readahead" to avoid that question. :-)
[00:22] <ogra> heh
[00:22] <JanC> for reference: I booted CRUX with XFce on a Pentium MMX @ 166MHz and with 64 MiB RAM from a 3200rpm disk in < 1 min four/five years ago...
[00:22] <Keybuk> JanC: that's utterly irrelevant, but thank you
[00:23] <JanC> it's relevant somewhat (I know Ubuntu provides a lot more convenience after that time)  ;)
[00:23] <Keybuk> why aren't you still using CRUX with XFce? :)
[00:24] <Keybuk> because it has roughly the same feature set as my fridge?
[00:28] <JanC> it lacked CUPS & auto-configuration of PnP-devices (USB, etc.), otherwise it wasn't that far away from what Ubuntu did back then...
[00:28] <Keybuk> *did back then* being the whole point here
[00:29] <Keybuk> more features -> bigger footprint -> more to load off disk
[00:31] <JanC> well, it should only read more if more is needed, which is still an issue
[00:32] <Keybuk> JanC: feel free to fix it :-)
[00:32] <JanC> hehe, I wish I could
[00:32] <Keybuk> if you don't know how to fix it, how do you know it can be fixed?
[00:32] <ogra> patches appreciated :)
[00:32] <Keybuk> in fact, how do you even know you know what the problem is?
[00:41] <JanC> Keybuk: I can see a lot of "useless" stuff being loaded by default, I understand some things in bootcharts that show where booting is sub-optimal, and I can see where the things you work on help with all that  ;)
[00:42] <JanC> of course, booting an optimized kernel saved me a lot of time on that system, which isn't really possible on Ubuntu (currently)
[00:45] <ogra> our only problem is that we dont still use linux 2.2 monolithic kernels
[00:45] <JanC> that was a 2.6 kernel  ;)
[00:45] <ogra> 2.2 would be even faster ;)
[00:45] <ogra> and 2.0 even more
[00:45] <JanC> but, mostly monolithic
[00:46] <ogra> and would your granny have been able to set up the system you refer to ?
[00:46] <JanC> one of the problems is that sometimes the software seems to need more time than we seem to gain from the hardawre improvements  ;)
[00:47] <ogra> with say 10min intro from you
[00:49] <JanC> ogra: obviously not, there is a reason why *I* (and probably you too) use Ubuntu
[00:49] <ogra> :)
[00:51] <JanC> but, I still consider it a reference, as that hardware was already outdated (4-5yo?) when I tried it 4-5 years ago
[00:54] <JanC> if we get the same boot times now with auto-configuration as hardware from 8-10 years ago with pre-configuration, there is a lot of space for improvements
[00:55] <JanC> the only problem is to find how  ;)
[00:55] <ogra> you really cant compare the two
[00:56] <JanC> yes & no
[00:57] <JanC> you can compare the extra features & what extra hardware resources it needs
[00:57] <ogra> and you definately cant make up something like "space for improvement" from comparing two unrealted things
[00:58] <ogra> they are completely different by design
[00:58] <cjwatson> to be honest, with that much difference in the software stack, you might as well just start optimising from scratch
[00:58] <ogra> right
[00:58] <cjwatson> you won't really get much out of a comparison
[01:13] <JanC> ogra / cjwatson: I think that "magnitude improvements" by using a different software stack indicate that the current software stack is not optimal yet, even if it has improved a lot over the last 4 years  ;)
[01:14] <ogra> it hasnt "improved", its largely completely new or rewritten
[01:16] <Moonraker12> Can someone fix the gtk filechooser so that its possible to select a directory as well as a file in the same dialog ? / Its kinda unusable as is...
[01:16] <ogra> Moonraker12, whats the bugnumber ?
[01:17] <Moonraker12> ogra: Im asking the people of this channel. Not sure if theres a bug on it yet... ?
[01:17] <ogra> so you didnt file one ?
[01:17] <Moonraker12> To that regard.
[01:17] <ogra> do that first
[01:17] <Moonraker12> .
[01:21] <Moonraker12> One more thing. I need a backup operator or network tech that wants to test some new additions to an application. One computer and another computer ith ssh is required to test it.
[01:21] <Moonraker12> Its a backup / Restore thing.
[01:22] <Moonraker12> ith/with
[01:22] <Moonraker12> (Puny screen:)
[01:37] <ScottK> ogra: Did you see our Kubuntu Netbook on freescale pic: http://www.flickr.com/photos/freeflying/4036438792/in/set-72157622515083587/
[01:39] <slangasek> apw: right - thanks for the dove upload
[01:42] <ogra> ScottK, nah, didnt see it yet, you didnt tell me the oem team tests for you ;)
[01:43] <ScottK> ogra: Just on his free time.
[01:43] <ScottK> He's been a Kubuntu user for a long time.
[01:44]  * ogra didnt know
[01:46] <Samus_Aran> does anyone know what in the world Ubuntu's partitioner actually *DOES* when it sits there for minutes "scanning the partition table" ?  there is no scanning needed, "fdisk -l" takes 1 second for 10 disks to gather the same information as Ubuntu's magical partitioner which takes 4+ minutes for those same 10 disks
[01:51] <LaserJock> I'm pretty sure it probes for other OSes
[01:55] <Samus_Aran> what does that even mean ?  that partition table specifies the type already
[01:55] <Samus_Aran> NTFS, FAT32, Linux, etc.
[01:56] <Samus_Aran> and even if it uses something to check the start of each partition for known filesystems, that wouldn't take more than at most 2 seconds
[01:56] <Samus_Aran> 4 minutes is enough time to read gigabytes of data
[01:56] <LaserJock> right
[01:56] <LaserJock> I'm just saying that it does more than fdisk -l
[01:57] <Samus_Aran> and it repeats this ridiculously long scan when you do almost anything
[01:57] <Samus_Aran> nothing more than an fdisk -l comes up on the screen, so I don't see what use any other scanning is
[01:57] <Moonraker12> Samus_Aran: Why dont you debug it and see if anything is actually wrong or if its acting securely ?
[01:58] <Samus_Aran> Moonraker12: who said anything about security ?  I said it is mind-bogglingly slow
[01:58] <Moonraker12> Samus_Aran: Do you code much Assembly and C btw ?
[01:58] <Samus_Aran> Moonraker12: why do you ask ?
[01:58] <Moonraker12> Samus_Aran: "Securely" To see if it can handle the partition or disk in the intended way.
[01:58] <Moonraker12> Dike
[01:59] <LaserJock> Samus_Aran: do you have this problem on the Desktop CD or Alternate CD?
[02:00] <Samus_Aran> LaserJock: the gui installer, I don't remember what the text installer does
[02:00] <LaserJock> I think the gui installer does a bit more work, it could be something unusual is happening
[02:00] <Samus_Aran> Moonraker12: what do you mean by "handle" the partition in the intended way ?  partition tables exist so that operating systems don't need to do any scanning, the partition table tells a possible OS what is already there
[02:00] <LaserJock> I've never had it take more than say 5-10 seconds
[02:00] <Samus_Aran> this scanning takes mere milliseconds on a modern computer
[02:01] <LaserJock> well, it does more logic than that
[02:01] <Samus_Aran> LaserJock: then you've only ever used a computer with one hard disk and probably only 2-3 partitions
[02:01] <Samus_Aran> try a ten disk setup
[02:01] <Moonraker12> > Samus_Aran: No, i cant code but i can think (Spleesh:)
[02:01] <Samus_Aran> and it doesn't cache this ridiculously slow scanning, so any time you change anything, it's back to scanning it all over again
[02:01] <LaserJock> Samus_Aran: right, that's not a common setup I guess. I have no idea
[02:02] <Samus_Aran> LaserJock: even with just 4 disks it takes about 30 to 45 seconds on a modern computer
[02:02] <Samus_Aran> there is no benefit to the end user of these slow scans, and a great deal of downside/frustration
[02:02] <Samus_Aran> no other distros do this, they all work fine
[02:02] <Samus_Aran> that's why partition tables exist
[02:03] <LaserJock> it does more than look at the partition table though
[02:03] <Samus_Aran> which it shouldn't do
[02:03] <LaserJock> you can see if the Alternate CD gives you better performance
[02:03] <Moonraker12> Samus_Aran: Yould like the possibility for the program to destroy disks instead of it taking you a few seconds longer Microsofty
[02:04] <LaserJock> Samus_Aran: well, it's doing things that most users will find useful
[02:04] <Samus_Aran> this is a problem which everyone faces.  even 10 seconds (and that's very fast for its scan) is far too long, as it will repeat this scan every time you modify the partition
[02:04] <LaserJock> I've not experienced that
[02:04] <Samus_Aran> LaserJock: name one thing it does which the user will find useful ?  (this does not include the regular partitioning)
[02:05] <LaserJock> well, I believe it looks for what other OSes are installed so it can make recommendations
[02:05] <Moonraker12> Samus_Aran: And how would you have coded it, Oh wise one ?
[02:05] <Samus_Aran> Moonraker12: how could it destroy disks ?  do other distros destroy disks ?  the answer is no.
[02:05] <Samus_Aran> Moonraker12: by reading the partition table, like every other partitioning app out there.
[02:05] <Samus_Aran> Moonraker12: no other distros have any issues
[02:06]  * ogra points Samus_Aran to #ubuntu-installer 
[02:06] <Samus_Aran> is there a listing somewhere of all the Ubuntu channels ?
[02:06] <Samus_Aran> there seems to be dozens
[02:06] <Moonraker12> Samus_Aran: Well, would you rather have an extra check or would you like to rely on the retvals of the nibblers ?
[02:06] <ogra> there surely are
[02:08] <Samus_Aran> Moonraker12: I would like the Ubuntu partitioner to behave like every other partitioning app out there.  which work completely fine.  in 18 years of x86 partitioning, I've never had an issue that a regular partitioning app doesn't handle
[02:08] <ogra> do other distros have tools like wubi or offer you to import settings from other OSes ?
[02:09] <ogra> afaik partman scans the content of the partitions as well ... tht simply takes its time
[02:09] <Samus_Aran> ogra: if it's taking so long because it's looking for settings, then it should ask the user if they wish to import settings before it spends a huge amount of time scanning
[02:10] <ogra> well, there is definately a very valid reason why it behaves as it does
[02:10] <Samus_Aran> and why would it rescan every time you change a partition ?
[02:11] <Samus_Aran> if it was looking for settings, its job is done after the first scan
[02:11] <ogra> as LaserJock suggested, compare with the alternate installer, if the live installer behaves largely differently, file a bug
[02:11] <Moonraker12> Samus_Aran: And in those 18 years (3) you have managed not to learn coding hat so ever as you said.
[02:11] <Samus_Aran> Moonraker12: I don't know what that sentence is supposed to mean
[02:11] <ogra> Samus_Aran, its not done ... the partitions changed to be 100% safe it needs to scan again
[02:12] <Moonraker12> Samus_Aran: It means youre lying, you entered after 2003
[02:12] <Samus_Aran> ogra: it is the one that changed the partitions, it knows what the changes are, there is no need to rescan.  if hda1 had windows, and you erased hda1, obviously there's no more windows
[02:12] <Samus_Aran> Moonraker12: I've been using x86 since the 8088
[02:12] <ogra> well, talk to the devs
[02:13] <Samus_Aran> Moonraker12: you don't know me, so stop lying
[02:13] <ogra> there is definately a reason why d-i/ubiquity are not using fdisk -l
[02:13]  * Moonraker12 *Snickers*
[02:13] <ogra> and there is definately a reason it needs to scan again
[02:14] <Samus_Aran> Moonraker12: C=64 then 8088 with IBM DOS and Phoenix DOS and later onto MS-DOS
[02:14] <Moonraker12> Samus_Aran: Say something you havnt said during the past 6 years that can convince me...
[02:14] <Samus_Aran> ogra: what is the reason for scanning again ?  it is the one that changed the partitions, it knows the changes
[02:14] <ogra> Samus_Aran, ask the developer of partman, i didnt write it
[02:14] <Samus_Aran> Moonraker12: I don't even know you, why are you acting like you know me ?
[02:15] <ogra> but i know he doesnt do stuff without a very valid reason
[02:15] <ogra> if fdisk -l would be enough he would use it, be sur
[02:15] <ogra> e
[02:15] <Moonraker12> Samus_Aran: It can be good for you.
[02:15] <Samus_Aran> ogra: who is the developer of partman ?
[02:15] <ogra> he is hanging around in #ubuntu-installer
[02:16] <ogra> all people contributing to partman in ubuntu are i belive
[02:16] <Samus_Aran> Moonraker12: your English is very difficult to understand.  what can be "good for me" ?
[02:16] <Moonraker12> I piss on you
[02:16] <Samus_Aran> Moonraker12: not on IRC you don't...
[02:16] <Moonraker12> Hard to understand ?
[02:16] <ogra> though there are plenty of debian devs contributing to it as well
[02:17] <ogra> Moonraker12, please mind your tone
[02:17] <Samus_Aran> Moonraker12: your English is very poor, and you assume you know me.  as far as I know, this is the first time we've ever met (other than earlier today in #Ubuntu)
[02:19] <Moonraker12> Samus_Aran: Youre a notorious liar. Im sorry my mother language is poor. I feel rich myself.
[02:19] <Moonraker12> what an ass
[02:20] <ogra> Moonraker12, please ...
[02:20] <Moonraker12> done.
[02:20] <Samus_Aran> Moonraker12: how am I a "notorious liar" ?  we've never even met.  maybe you have me mistaken for someone else ?
[02:21] <Samus_Aran> Moonraker12: and I didn't say your mother language was poor, I said your English communication is poor.
[02:23] <Moonraker12> No, from all i can tell youre an "IMCP" from Microsoft. Let me know if you can convince me of the opposite. Otherwise, please do not speak to me.
[02:24] <Moonraker12> Im not poor on any way.
[02:26] <Samus_Aran> Moonraker12: what is an IMCP ?  I'm not anything from Microsoft.  I don't own or use any Microsoft products.  I have exclusively used GNU+Linux since 2001... look at my FreeNode cloak, given to me by lilo for helping out many hundreds of hours in ##Linux
[02:27] <Moonraker12> I have coded for the public domain etc since 1994. What as you previous nick then ?
[02:27] <Moonraker12> as/as
[02:27] <Moonraker12> was
[02:27] <Moonraker12> Bad keyboard
[02:27] <Samus_Aran> Moonraker12: you don't know me at all.  why are you attacking me ?
[02:28] <LaserJock> ok people, this is getting ridiculous
[02:28] <Samus_Aran> Moonraker12: I have used Samus_Aran for the past seven years on here...
[02:28] <LaserJock> Samus_Aran: I would suggest trying the Alternate CD and seeing how that performs in comparison as a first step
[02:29] <LaserJock> Samus_Aran: and filing a bug would be good
[02:29] <Samus_Aran> LaserJock: okay
[02:29] <Moonraker12> It is, isnt it. Samus_Aran: How come ive never seen you before on any irc network, having been on them 24/7 since atleast 2000 ?
[02:29] <LaserJock> as right now people are trying to fix critical last-minute bugs
[02:29] <LaserJock> and really don't have time to do the problem justice
[02:29] <ScottK> Moonraker12: It's really off topic for this channel.
[02:29] <LaserJock> but a bug report will help them track the issue
[02:30] <Moonraker12> Done, as i said a long time ago.
[02:30] <Samus_Aran> Moonraker12: that is ridiculous.  there are hundreds of thousands of people on IRC -- obviously you do not know all of them.
[02:30] <ScottK> Samus_Aran: It's off topic for you too.
[02:30] <lifeless> Samus_Aran: Moonraker12: both of you please stop
[02:30] <lifeless>  /ignore each other if you have to, but let us focus on the release. Thanks.
[02:31] <Samus_Aran> lifeless: does Moonraker12 normally attack everyone that comes in this channel out of the blue ?
[02:31] <Moonraker12> We are ok no.
[02:31] <jdong> what on earth is in the last 3 pages of my scrollback?
[02:32] <lifeless> Samus_Aran: you're continuing the conflict at this point; that question is itself escalatory.
[02:32] <chrisccoulson> jdong - you haven't missed anything important
[02:32] <LaserJock> Samus_Aran: we know Moonraker12 just as much as you
[02:32] <lifeless> so please, stop.
[02:32] <Samus_Aran> LaserJock: okay.  they were talking like they were a regular here.
[02:35] <Moonraker12> Samus_Aran: Settle down now. I didnt mean any harm, just probe you a bit.
[02:35] <Samus_Aran> lifeless: fine, but I really don't appreciate asking a legitimate question then being attacked and accused of being a Microsoft mole and only having used computers since 2003 and a liar.  it isn't easy to not respond to that crap.
[02:35] <Samus_Aran> anyhow, I'm out of here now, #Ubuntu-installer is the correct channel for my question.
[02:37] <bryce_> goodness
[02:38] <jdong> heh indeed....
[02:42] <chrisccoulson> slangasek - i can't remember if i mentioned to you already - i uploaded a patched version of g-s-d to my PPA a few days ago, which might give some insight as to why you see frequent low-disk space warnings (i added lots of debug messages to it)
[02:42] <chrisccoulson> would you mind trying it at some point, and showing me the debug output?
[02:46] <Moonraker12> bryce, jdong: Sorry for that. It was required on a grander scale of things plus i dont like it when people say my English / American / Europeean etc isnt good enough for them. (Thats their first line of defence).
[02:48] <Moonraker12> As they dont have a second line ;)
[02:52] <slangasek> chrisccoulson: actually, at this point I'm not sure the frequent warnings aren't simply due to the frequency with which I have to restart g-s-d due to the xrandr bug; at least, /now/ I only see it once per restart
[02:52] <slangasek> chrisccoulson: so I'd much rather focus on that bug
[02:53] <chrisccoulson> slangasek - that actually makes more sense. i've tried very hard to recreate the disk space warning problem here, and couldn't get it to break
[02:53] <chrisccoulson> did you manage to run it through xtrace yet?
[02:53] <slangasek> no
[02:53] <slangasek> will try to do so this weekend
[02:54] <slangasek> I have pinpointed a reproducible sequence as well now, which might be enlightening; will post it to the bug
[02:54] <chrisccoulson> thanks. i'll take a look at it then. hopefully it will show up something obvious
[02:54] <chrisccoulson> a sequence of events to trigger it will be useful too
[02:55] <slangasek> chrisccoulson: 1) boot, 2) plug in a VGA monitor with a higher resolution than the LCD, 3) hit the button :)
[02:55] <slangasek> (it always happens on the first time after plugging in the VGA monitor)
[02:56] <chrisccoulson> quite simple then. did you also say you were having issues with resolution reverting to 1024x768 when opening the lid too?
[02:58] <slangasek> chrisccoulson: yeah, still having some sort of problem there; lower priority for me
[02:59] <chrisccoulson> slangasek - i noticed this in the gnome-desktop code a few days ago, which might just be a coincidence:
[03:00] <chrisccoulson> * Pick some random numbers. This should basically never happen */
[03:00] <chrisccoulson> output->pref_width = 1024;
[03:00] <chrisccoulson> output->pref_height = 768;
[03:00] <slangasek> heh
[03:00] <chrisccoulson> anyway, sleep time for me now!
[03:00] <slangasek> ok, 'night :)
[05:22] <otrewyi191> hi
[09:20] <spaettzAway> xx
[09:37] <billisnice> my gimp will act like it is loading and then disappear.
[11:59] <slangasek> ccheney: hmm, just noticed (on upgrade) that the latest OOo-common also bumped xfonts-mathml from a suggests to a recommends... hoping that doesn't cause any space problems or introduce any regressions, since this wasn't included on the images at all before.
[12:00] <doko_> yup, it wasn't just bug fixes, but an ooo-build and debian merge as well :-/
[12:02] <slangasek> yeah.
[12:46] <pengo> hi
[12:46] <pengo> can someone please re-open this bug as it's still getting reports? https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/150872
[12:53] <slangasek> pengo: the latest comments in the bug talk about USB sticks, which would be an entirely separate bug and should be reported separately
[12:53] <pengo> slangasek: no, that's the same bug
[12:54] <pengo> slangasek: caused by the same fault anyway.. i'm pretty sure there are duplicates that mention USB
[12:57] <pengo> it's a very annoying bug
[12:57] <pengo> oh well good night
[13:37] <logari81> totem user experience:
[13:37] <logari81> 1) u try to open a wmv file after a fresh installation of karmic
[13:37] <logari81> 2) u are prompted to install additional codecs "Windows media video 8 & 9"
[13:37] <logari81> 3) the installation of the corresponding packages ends successfully
[13:37] <logari81> 4) u get the following error message
[13:37] <logari81> http://files.ubuntu-gr.org/forum/logari81/Screenshots/Screenshot-Totem%20bug.png
[13:37] <logari81> 5) u try to reopen the file
[13:37] <logari81> 6) it works
[13:38] <logari81> is there any chance to get this annoying defect fixed until the end release?
[13:44] <evand> logari81: please file a bug using `ubuntu-bug totem`
[13:47] <slangasek> TheMuso: I don't see any explanation in 450214 of why e2fsprogs is thought to be misbuilt?
[15:00] <logari81> evand: done #459772, I think it is the kind of detail the 101 papercuts project was thought about, it is a pity that a new user will have to confront with such an error message.
[15:03] <evand> logari81: I think it's a normal run of the mill bug, not a usability bug, which is what the papercuts project is tasked with.
[15:04] <logari81> evand: ok, it is not a usability issue, u are right
[15:23] <cjwatson> slangasek: 450214> TheMuso told me on IRC that after he'd fixed the swap problem he found that mke2fs failed in the same way, and that rebuilding e2fsprogs with -O2 fixed it for him
[15:52] <flower> why does /etc/apt/preferences not work in Ubuntu (jaunty), at least this example doesn't work: http://pastebin.com/me179d8a
[15:52] <flower> it works in Debian
[16:06] <slangasek> cjwatson: ack, accepted then
[16:22] <Keybuk> sladen: still not around?
[16:36] <flower> and nobody seems to be able to answer my question...
[16:36] <Keybuk> flower: this isn't a channel where you ask questions, try #ubuntu
[16:38] <flower> Keybuk: you're right, I asked it there... but they asked me if I meant /etc/apt/sources.list maybe :/
[16:38] <ccheney> slangasek: oh :(
[16:39] <flower> and if I'm right that it aint working properly in Ubuntu it seems to be more like a dev topic
[16:39] <Keybuk> flower: no, see the topic
[16:40] <Keybuk> if you think something isn't working right, file a bug
[16:40] <flower> mmh ....
[16:40] <flower> ok....
[18:07] <apw> pitti, slangasek, bug #422536 appears to not be a kernel issue at all, but an apport false positive ... pitti i wonder if you could confirm, and let us know if this can be fixed in apport easily
[18:58] <bluefox_> how much RAM do I need to run karmic?
[18:58] <bluefox_> Mem:   3800308k total,  1164784k used,  2635524k free,   115840k buffers
[18:58] <bluefox_> Swap:   787072k total,        0k used,   787072k free,   418048k cached
[18:58] <bluefox_> this is what I Get logging into Gnome and running xterm and top immediately.
[18:59] <bluefox_> Nautilus has a 780M VMA space with 149M resident and 22M shared, it's the most memory hungry process on the list; next is Xorg with 232M, 91M, 28M
[18:59] <bluefox_> so the top two make up about 250M of RAM usage.
[19:24] <JanC> bluefox_: if you have enough swap (and enough time), 64 MiB is enough  ;)
[19:25] <JanC> 256 MiB should be enough for basic use
[19:26] <bluefox_> JanC:  Just logging in takes me 680MB of non-cache
[19:26] <JanC> and virtual memory includes mmap'ed files & binaries
[19:29] <bluefox_> JanC:  I know what VMA consists of
[19:29] <bluefox_> JanC:  lookat the used/buffers/cache numbers.
[19:29] <bluefox_> 1164784 - (115840 +  418048)
[19:29] <bluefox_> 630896
[19:31] <JanC> bluefox_: mmap'ed files & binaries aren't included in "buffers" & "cached" AFAIK, while they can be "removed" from RAM if needed (as they exist on disk anyway)
[19:32] <bluefox_> IIRC "cache" is VFS cache and buffers are mmap() buffer cache
[19:33] <bluefox_> but point taken.
[19:33] <JanC> well, the live-CD supposedly boots with 256 MiB RAM & no swap  ;)
[19:33] <bluefox_> memory management and analysis is crap on linux :(
[19:34] <bluefox_> it's like "there's X ram in use but we don't really know how much is needed and how much is stale crap"
[19:34] <JanC> well, you certainly need more numbers than just those
[19:36] <bluefox_> JanC: I still say the real test is the Quake test.
[19:36] <bluefox_> I used to run Quake on 64M Windows 98SE o_o
[19:36] <bluefox_> 32M was comfortable, I had a lot of ram :D
[19:37] <bluefox_> (Quake used to run on a 33MHz 486DX with 8M of RAM, though)
[19:37] <JanC> and memory management isn't that bad actually, e.g. Windows doesn't mmap libraries
[19:37] <bluefox_> yeah
[19:38] <bluefox_> JanC: the OOM is broke bt.
[19:38] <bluefox_> btw.
[19:38] <hyperair> what's wrong with mmap?
[19:38] <JanC> Windows will copy libraries to RAM and then from there to the swap file, thus you have 2 copies of the DLL on disk  :P
[19:38] <bluefox_> I made the mistake of leaving Evolution open for a few hours and it eventually ate several gigs of ram, then my PC froze
[19:39] <hyperair> fun
[19:39] <bluefox_> it froze after I noticed Evolution's RES in top, and started typing 'killall' :P
[19:39] <bluefox_> killa-- aww damn.
[19:39] <JanC> Evolution like to freeze from time to time, not related to RAM  ;)
[19:39] <bluefox_> no, my WHOLE MACHINE froze.
[19:40] <bluefox_> it became unusable
[19:40] <hyperair> well livecds cannot be used as benchmarks imo
[19:40] <JanC> I have Evolution open for > 1 month sometimes...
[19:40] <hyperair> while they're meant to be somewhat usable, a significant portion of your RAM is used for storing files.
[19:40] <bluefox_> it takes me that long just to open Evolution....
[19:40] <JanC> with > 3 GiB of mail in IMAP accounts
[19:40] <bluefox_> heh, it actualyl does take me 10-15 minutes to open evo
[19:40] <maco> JanC: wow. ive never had >1 month uptime. crap...3 days is a long time for me
[19:41] <hyperair> JanC: seriously? i'm always running evolution --force-shutdown.
[19:41] <bluefox_> I have more mail stored in Thunderbird, which opens quick
[19:41] <hyperair> JanC: stupids thing hangs whenver the network acts up.
[19:41] <hyperair> stupid*
[19:41] <bluefox_> two accounts in Evo, which takes 10-15 min to open
[19:41] <JanC> yeah, Evolution sometimes doesn't like network hickups
[19:42] <JanC> and it opens quick enough, closing it takes much longer  ;)
[19:42] <JanC> (syncing with the IMAP server first)
[19:43] <bluefox_> JanC: I can't get evo to open quick
[19:43]  * bluefox_ maybe strace -t it, or strace -r, to get timing information and wtf it'sd oing.
[19:45] <bluefox_> JanC: lots of lseek() and read()
[19:46] <JanC> that's normal, but shouldn't take 10 min  ;)
[19:46] <bluefox_> itpops up a box"Do you want to makeevolution your default mail client?" and I Click no, then the waiting begins
[19:46] <JanC> eh
[19:47] <JanC> it has to download a lot of messages maybe?
[19:47] <bluefox_> and tons of disk activity
[19:47] <bluefox_> okthat was quick
[19:48] <bluefox_> 14:43:49 to 14:47:39
[19:48] <bluefox_> so evolution's faster now than it was pre-karmic
[19:48] <bluefox_> MUCH faster.
[20:24] <arand_> partman-basicmethods is part of the installer (udeb package), and I want to test a recent patch for it, can anyone hint me how?
[21:14] <evand> arand_: anna-install openssh-client-udeb; scp your changes in.
[21:14] <evand> before you get to the partitioner, of course
[21:16] <arand_> evand: ok, cheers
[21:23] <evand> arand_: sure thing
[21:24] <evand> arand_: for what it's worth there's also #ubuntu-installer, where we handle installer development
[21:25] <arand_> evand: Yea... I just bothered cwatson into patching a partman bug and now I've no clue how to test the patch :/
[21:30] <evand> heh, I'm in and out tonight, but I can give you a hand if you run into any trouble and I'm still around.
[21:45] <TheMuso> This is weird. Both a juanty and karmic box here have gone back to eastern Australian standard time, instead of being at eastern daylight savings time... Weird.
[21:53] <TheMuso> hrm ok seems my phone is an hour fast, not my machines.
[21:54] <lifeless> TheMuso: its 0800
[21:54] <lifeless> (well just before)
[21:54] <TheMuso> lifeless: Yes, I know.
[21:54] <lifeless> TheMuso: cool; just offering external data
[21:55] <TheMuso> My phone seems to have decided to act weird on me for some reason.
[21:55] <maco> theyre sneaky like that
[21:56] <TheMuso> Its to do with having changed the time on my phone manually due to $telco not doing it correctly via auto-update at the dayliht savings time. Now that has finally come through, and its put my phone out an hour.
[21:58]  * dtchen frowns at libsdl/src/audio/{alsa,pulse}/ in Karmic
[22:00] <lifeless> TheMuso: are you with vodafone?
[22:02] <TheMuso> lifeless: no telstra
[22:03] <lifeless> TheMuso: I wish voda would do times in .au
[22:03] <lifeless> they epically fail here
[22:22] <pgraner> pitti: Bug 422536
[22:54] <TheMuso> slangasek: re hotkey-setup, it was inherrited. I can upload a new meta to remove it, if you wish to purge it from main etc.
[22:54] <TheMuso> s/main/the archive/
[22:58] <wamty> ubuntu is a whole lot better when it doesn't ask you for the password every time you need to change something or install something.
[22:58] <wamty> why it does that?
[23:02] <ahe> wamty: for security reasons but #ubuntu is a much more appropriate place to ask such questions
[23:21] <TheMuso> /c/c