[00:03] <sebas> ScottK: Hm, sounds interesting, but also not supertrivial to do nicely
[00:04] <sebas> powerdevil should offer an action to pop the applet's info panel, and the info panel needs a signal to react to
[00:05] <sebas> I'll talk to dario about that
[00:05] <sebas> (dario being the powerdevil dude)
[00:29] <ScottK> sebas: Cool.  Thanks.
[03:19]  * Daskreech waves
[03:19] <Daskreech>  Will Lucid be moving to Trunk when the Beta is released?
[03:34] <ScottK> Daskreech: Yes.
[03:37] <Daskreech> ok thanks
[03:44] <Daskreech> !info kolourpaint
[03:44] <Daskreech> ScottK: Does anyone know which lightweight painting program exists in the KDE world?
[04:16] <ScottK> No, sorry.
[04:24] <kb9vqf> Daskreech: Try kolourpaint
[04:24]  * kb9vqf is not sure if there is a KDE4 version though
[04:25] <Daskreech> kb9vqf: There is ish
[04:25] <Daskreech> It hasn't really been worked on since 4.0
[04:26] <kb9vqf> Ahh...that's one program I couldn't do without...having grown up on MS Paint I find the Gimp too powerful for day-to-day simple editing tasks
[04:27] <Daskreech>  The person maintaining it got frustrated since one of the major changes in Qt4 is Graphics and Painting which is mostly what Kolourpaint does
[04:30] <Daskreech> I know It's useful I'm looking either for a quick replacement or a way to put together some assistance for him
[04:48] <shtylman> sebas: nice work on the battery plasmoid
[04:50] <shtylman> sebas: does the detached window still stay there if you have focus follows mouse?
[04:50] <shtylman> that might be a concern (I have it enabled)
[05:25] <nixternal> !info kolourpaint4
[05:25] <nixternal> Daskreech: ^^
[05:30] <nixternal>  /wi6
[05:39] <Daskreech> nixternal: Ah thanks
[05:39] <Daskreech> nixternal: Still an unmaintained application unless the port gets rebooted
[08:06] <markey> anyone know a ShowCock 1.0 PPA?
[08:07]  * maco blinks
[08:07] <maco> do you mean choqok?
[08:07] <markey> no, ChokeOnCock
[08:07] <markey> sry
[08:07] <markey> ;)
[08:07] <maco> !ops | markey
[08:07] <markey> yes, Choqok
[08:07]  * markey blinks at COD
[08:07] <markey> but oh well
[08:08] <markey> might want to read this: http://amarok.kde.org/blog/archives/1143-Application-Naming.html
[08:09] <markey> but I disgress
[08:12] <markey> found this: 0.9.4-0ubuntu1~ppa1
[08:12] <markey> is it current?
[08:12] <markey> 2009-10-15
[08:12] <markey> hm
[08:12] <markey> good enough
[08:12] <maco> choqok is a persian word for some kind of little bird
[08:12] <markey> yes, no news there
[08:12] <maco> and with twitter being the sound birds make...
[08:13] <markey> hmm
[08:13] <markey> Choqok 1.0 Alpha2 released: “Razi”
[08:13] <markey> gonna build it myself, I think
[08:14] <markey> args, in SVN
[08:14] <markey> dear god
[08:14] <markey> ok
[08:16] <markey> hmm, Alpha 2 == 0.9.4
[08:16] <markey> might as well use the PPA :)
[08:16] <markey> neat
[08:27] <markey> wow, this works nicely
[08:27] <markey> iLike
[08:27] <markey> neversfelde++
[08:30] <tsimpson> markey: can you try to keep your comments more "family friendly" in future :)
[08:30]  * tsimpson notes this channel is publicly logged and indexed by google
[08:30] <markey> of course. I shall try :)
[08:30] <tsimpson> thanks
[08:31] <markey> pleasure
[08:31] <tsimpson> I know there has been worse in here, so I can't really berate you for the above pun :p
[08:32] <markey> who should be berated is the author of Choqok
[08:32] <markey> maybe he'll get it, maybe not
[08:32] <markey> the app is good anyway
[08:45] <Tm_T> maco: ?
[08:46] <maco> Tm_T: was a language issue. tsimpson already told him off
[08:47] <markey> regretted it deeply
[08:48] <Tm_T> oh right
[08:49]  * Tm_T ties markey son tightly to his chair
[08:50] <markey> please watch out, this could have s/m connotations, we're trying to be family friendly here, Tm_T
[08:51]  * maco headdesk
[08:54] <Tm_T> markey: please...
[09:11] <Tm_T> !guidelines > markey just to remind with love
[09:12] <Tm_T> actually, I should read those too ):
[09:17] <Nightrose> neversfelde: will you package choqok alpha 2?
[09:18] <Nightrose> oh nvm
[09:32] <Quintasan> hello
[10:04] <Quintasan> I install pulseaudio and pray it doesn't break my sound card
[10:05] <freinhard> good luck then
[10:05] <Quintasan> PulseAudio - The Ultimate Solution or Ultimate Disaster™
[10:17] <Quintasan> :/
[10:17] <Quintasan> nothing changed
[10:18] <Quintasan> FCK
[10:20] <Tm_T> Quintasan: nothing changed in where?
[10:21] <Quintasan> Tm_T: sound, I can't play sound from multipe sources at once, for ex. I need to turn amarok, opera off to play stepmania because amarok hogs the sound output :/
[10:21] <Tm_T> Quintasan: have you checked your phonon settings?
[10:22] <Quintasan> Tm_T: not really, what should I have there?
[10:23] <Tm_T> Quintasan: depends on what you want, if you want your phonon to use pulseaudio, make sure it's first interface in priority
[10:26] <Quintasan> huh, even Test sound doesn't work with Pulse :/
[10:29] <jussi01> pulse is of the devil...
[10:29] <Quintasan> screw it, I'm getting my SoundBlaster back :/
[10:50] <Quintasan> kay
[10:50] <Quintasan> everything works - Sound Blaster == win
[10:51] <Quintasan> it seems that driver or card itself sucks
[10:55] <Tm_T> Quintasan: or some mix of all components
[11:34] <neversfelde> Nightrose: yes, I will do this later this day
[11:52] <apachelogger> Sput: cant quassel do type conversion for the core lag? :P
[11:52] <apachelogger> 8000 msec sounds a bit weird
[12:17] <Quintasan> hurr durr, qt sauce is bigger than kdelibs+base
[12:19] <Tm_T> Quintasan: whole Qt tree or just Qt libraries & friends?
[12:20] <Quintasan> Tm_T: I'm syncing kde-qt
[12:21] <Tm_T> Quintasan: ok, that includes it all then I suppose
[12:22] <Quintasan> 190mb :/
[12:22] <Tm_T> Quintasan: I hope you excluded .git and friends
[12:23] <Tm_T> Quintasan: btw kdelibs and kdebase together are bigger than that atleast here (:
[12:23] <Quintasan> Tm_T: they synced faster than Qt so I assume Qt was bigger :P
[12:24] <Tm_T> nah
[12:34] <Quintasan> building!
[13:18] <apachelogger> is it me or is revu down?
[13:19] <ryanakca> ScottK, Tonio__: Bangarang is in Debian NEW
[13:20] <apachelogger> talking about bangarang, is one supposed to manually add files?
[13:20] <JontheEchidna> apachelogger: #ubuntu-motu sez it's down until sunday
[13:20] <apachelogger> oh dear
[13:20] <apachelogger> so much for doing some reviews while I am stuck in upper austria -.-
[13:21] <mcas> apachelogger: Xand3r told me that you are the one to call if there are problems with project neon...
[13:22] <apachelogger> project neon is dead
[13:22] <mcas> hmpf
[13:22] <mcas> gnarf
[13:22] <Xand3r> mcas: sorry, i didn't know that
[13:23]  * apachelogger giggles over how easy it is to crash bangarang
[13:24] <mcas> than i have to do plan b ... compile the dev-env by myself
[13:56] <Quintasan> fcks
[14:08] <apachelogger> Quintasan: there there
[14:08]  * apachelogger just got stuck in bangarang :D
[14:08]  * Quintasan got stuck compiling Qt
[14:08] <Quintasan> synced wrong tree propably :/
[14:09] <apachelogger> you know, it probably is weird that I find enjoyment in buggy software, but oh hey it really is cool :D
[14:09] <apachelogger> Quintasan: haha, lol, now that is indeed very unfortunate
[14:09] <Quintasan> wut, bangarang? where are packages?
[14:09]  * apachelogger compiled
[14:09] <apachelogger> indeed I am compiling most of the time these days, and I have no idea why
[14:10] <Quintasan> srsly Tech base says -> git://gitorious.org/+kde-developers/qt/kde-qt.git qt-kde
[14:10] <apachelogger> Quintasan: wikipedia is also not always right
[14:10] <Quintasan> and it fails at webkit, #kde-devel says -> git://qt.gitorious.org/qt/kde-qt
[14:10] <apachelogger> even if it is most of the time
[14:11] <Quintasan> well I just though it's updated :P
[14:11] <Quintasan> I'm wondering why it goes at 20kB/s
[14:11] <apachelogger> it's like in math, you cant assume something is true unless you proof it to be true or to be wrong :P
[14:13] <Quintasan> why don't we have a tool for backuping configuration?
[14:34] <apachelogger> Quintasan: there once was work done on such a tool
[14:34] <apachelogger> kamino or something
[14:35] <apachelogger> never got released for real I think ... also I imagine it all a bit difficult because even though there are general rules of config and data storing the apps can override it
[14:35] <apachelogger> which of course makes for a much more complex implementation
[14:36] <Quintasan> I wanted a tools that would allow you to preserve WHOLE configuration (.kde) or certain files like Nepomuk database
[14:37] <apachelogger> .kde != configuration
[14:37] <apachelogger> .kde is a lot more
[14:37] <Quintasan> still, desktop search via krunner doesnt work :/
[14:37] <apachelogger> and indeed .kde might also include a lot of crap
[14:37] <Quintasan> so let's say .kde/share/config :P
[14:37] <apachelogger> stuff in /share/config/ might depend on /share/apps/
[14:38] <apachelogger> whereas share/apps/ can depend on share/services
[14:38] <apachelogger> which might be one of the things you would not want to have backed up
[14:38] <Quintasan> so basiacally whole .kde or nothing?
[14:38] <Quintasan> basically*
[14:38] <apachelogger> more like generally
[14:39]  * Quintasan notes that down
[14:39] <apachelogger> basically you could backup application specific data
[14:39] <Quintasan> I wanted to do that in Python but ___init___(self)
[14:39] <Quintasan> made me lol really hard
[14:39] <apachelogger> if you know what to backup, which is diffcult, because, as said, any app can make random use of kde's storage locations
[14:39] <apachelogger> Quintasan: dont do stuff in pyton!
[14:39] <apachelogger> use the bash
[14:39] <apachelogger> or the cpp
[14:40] <apachelogger> or the java
[14:40] <apachelogger> just not the python
[14:40] <Quintasan> I have a nice idea, why waste time looking on % of git sync while I can learn c++?
[14:41] <Quintasan> first let's make a playlist for coding since metal isn't exacly what I want to listen all the time :P
[14:43]  * apachelogger finds the nepomuk search in dolphin weird
[14:45] <Quintasan> I don't mean nepomuk, screw nepomuk now, it should list files from strigi index, shouldn't it?
[14:45] <apachelogger> Quintasan: yes
[14:45] <apachelogger> well
[14:45] <apachelogger> it should
[14:45] <Quintasan> and it fails
[14:45] <Quintasan> :/
[14:45] <apachelogger> doesnt here
[14:46] <apachelogger> Quintasan: go fix your sesame :P
[14:46] <Quintasan> fix'd
[14:46] <Quintasan> :P
[14:46] <apachelogger> in fact go fix java alltogehter
[14:46]  * apachelogger finds it quite a pita
[14:46] <apachelogger> Quintasan: search works just fine
[14:46] <Quintasan> damn, redland sucks, sesame sucks
[14:46] <apachelogger> BUT
[14:46] <apachelogger> not for tags
[14:46] <Quintasan> apachelogger: redland?
[14:46] <apachelogger> sesame
[14:46] <apachelogger> besides
[14:46] <Quintasan> :|
[14:46] <apachelogger> I doubt that is related
[14:46] <apachelogger> either the ontology is missing or strigi does fail to read the stuff
[14:47] <Quintasan> Indexing modules is idle
[14:47] <apachelogger> because sesame and redland can not introduce limitations to the stored data, they can only limit the performance et all
[14:47] <apachelogger> they are storage backends after all
[14:47] <Tm_T> apachelogger: hmm, why java and not python?
[14:48] <apachelogger> because it is way easier to do runtime bugs in python than in java
[14:48] <Tm_T> true that
[14:48] <Quintasan> apachelogger: in Dolphin typing Fate lists all my files tagged with Fate/Stay night :P
[14:48] <apachelogger> mostly related to python being 100% interpreted
[14:48] <Quintasan> apachelogger: but krunner sucks
[14:48] <apachelogger> Quintasan: what does this have to do with krunner?
[14:49] <Quintasan> apachelogger: doesn't krunner use the same search engine as Dolphin?
[14:49] <apachelogger> the same underlying lib
[14:49] <apachelogger> but dolphin uses a kioslave
[14:49] <apachelogger> whereas krunner probably uses either a lib or directly queries nepomuk
[14:50] <Quintasan> oh wait, it displays the results
[14:50] <Quintasan> http://imagebin.ca/view/xtECVqJp.html
[14:50] <apachelogger> also the sorting algorithms might just be very different
[14:50] <Quintasan> but the naming is PITA and opening isn't working :/
[14:50] <apachelogger> dolphin will sort all returns by name while krunner probably goes by relevance or matchiness or something
[14:50] <apachelogger> Quintasan: dont get that crap here
[14:50] <apachelogger> though
[14:50]  * apachelogger tries with tagging
[14:51] <Quintasan> apachelogger: you don't see that PITA naming in krunner results? :P
[14:51] <apachelogger> Quintasan: that is a bug in the runner
[14:51] <Quintasan> well, let me ask this
[14:51] <apachelogger> it doesnt resolve the name of the item tagged or something
[14:52] <apachelogger> Quintasan: search for a file name and it will show up fine
[14:52] <apachelogger> when looking for a tag it screws up and shows the nepomuk identifier or whatever that thingy is
[14:53] <Quintasan> apachelogger: what's the point of searching for exact file name?
[14:53] <apachelogger> well, search for part of the filename then :P
[14:53] <apachelogger> anyhow, it doesnt have a point
[14:53] <apachelogger> neither does manual tagging for that matter
[14:53] <Quintasan> and it fails to list those
[14:54] <apachelogger> youd either compose a complex nepomuk query or search for a predefiend attribute of an item
[14:54] <apachelogger> like say I would search for the moby signle pale horses
[14:54] <Quintasan> it shows the Last opened files that match but let's say I want to watch 1st ep of Fate/Stay night and I didn't open it, but I did open 14,15,16,17,24
[14:55] <Quintasan> wait
[14:55] <Quintasan> HURRDURR
[14:56] <Quintasan> http://imagebin.ca/view/6IV-Zan.html
[14:56] <Quintasan> works!
[14:56] <Quintasan> dunno why but works
[14:57] <Quintasan> shit, it really does but you need to wait for a second or two :P
[14:57] <apachelogger> that is because sesame is still slow I would suppose
[14:57] <apachelogger> or because the algorithm is a bit off
[14:58] <apachelogger> oh well
[14:59]  * apachelogger feels like reporting bugs against launchpad blueprints
[14:59] <apachelogger> having specs stored in the wiki is like storing them on gobby.ubuntu.com
[14:59] <apachelogger> just that former supports somewhat pointless formatting
[15:00] <apachelogger> neither does help with discoverability, or searching or obtaining information without reading the whole freaking spec or atleast scrolling down to grasp what it might be about
[15:01] <Quintasan> apachelogger: hmm do you know how to make a hotkey to minimize all windows? I can't find it in KHotkeys
[15:01] <apachelogger> qdbus call to plasma or kwin I suppose
[15:02] <Quintasan> unclutterDesktop()?
[15:03] <apachelogger> probably not :P
[15:03] <Quintasan> wtg
[15:04] <Quintasan> I just don't know how to chain that command now
[15:04] <Quintasan> qdbus org.kde.kwin /KWin.nextDesktop() ??
[15:04] <Quintasan> nvm
[15:04] <Quintasan> :P
[15:06] <Quintasan> qdbus:11: maximum nested function level reached
[15:06] <Quintasan> wtf?
[15:06] <Quintasan> :|
[15:07]  * apachelogger syncs 3000 mails with gmail :D
[15:07] <apachelogger> poor old workstation is all outdated
[15:08] <Quintasan> lol
[15:09] <Tm_T> apachelogger: good that gmail is mostly sane
[15:09] <apachelogger> the fact that I still have to sync and configure kmail on all my machines is a PITA though
[15:09] <apachelogger> need ubuntone to sync that stuff automagically :P
[15:10] <Quintasan> apachelogger: you are using IMAP?
[15:10] <apachelogger> what else is there?
[15:10] <Quintasan> pop :D
[15:10] <apachelogger> well, I am not using unix either, am I?
[15:11] <Tm_T> apachelogger: hm?
[15:12] <apachelogger> why would one use an inferior product?
[15:13] <Tm_T> apachelogger: "because I can" I suppose (;
[15:13] <apachelogger> yes, but there is no point to it :P
[15:14] <apachelogger> the only reason pop is still around is probably because mail hosters are too lazy to implement imap all over the place, and same goes for mail cients
[15:14] <Tm_T> oh, I wouldn't even begin to count things have no point I do
[15:15] <apachelogger> which makes me raise the recommendation that one should rethink whether one would want to stick with a hoster and/or client that does not support much more sensible protocols
[15:15] <Tm_T> apachelogger: and people don't use d-imap
[15:15] <apachelogger> well, they do not, I personally find d-imap not very sensible either
[15:15] <Quintasan> apachelogger: lol I'm so stupid, you can bind plasmoids to hotkeys
[15:15] <apachelogger> but if someone wishes to use dimap, so they should, considering their client supports it
[15:15] <Quintasan> :D
[15:15] <apachelogger> then they have pop3 behaviour just better
[15:16] <Tm_T> apachelogger: yup, though, having offline copy of mail archives isn't always so bad idea
[15:16] <Tm_T> apachelogger: especially when someone is as cloumsy as I am
[15:16] <apachelogger> well, it depends on what you want
[15:17] <Quintasan> dimap? wth?
[15:17] <apachelogger> the lecturer of my internet and new media lecture also prefers to keep local backups of his mail and whatnot
[15:17] <apachelogger> that is not mutually exclusive with imap though
[15:17] <apachelogger> having access to your mail from various machines is, however, mostly mutually exclusive with pop3
[15:18] <apachelogger> unless your mail service provider does uglish things, which then again can lead to problems with your mail client anyway
[15:18] <apachelogger> eitherway imap is superior :P
[15:18] <apachelogger> Quintasan: disconnected imap
[15:18] <apachelogger> then the client would just download all messages and keep them as local copies
[15:18] <apachelogger> it has the advantage of offline access, but of course wastes diskspace
[15:18] <Tm_T> indeed
[15:19]  * apachelogger for one finds it much more sensible to keep a given amount of recent messages as local copies
[15:19] <Tm_T> that's why I keep only really important sections in dimap use
[15:19] <apachelogger> they might be good to have offline, but why would I need a message from like 5 years ago :D
[15:20] <Tm_T> apachelogger: who says you have to with dimap (:)
[15:21] <apachelogger> didnt
[15:21] <apachelogger> just saying :D
[15:21] <Tm_T> I know, and I don't even if I should sometimes... oh well
[15:23] <Tm_T> I will use more online possibilities as soon as I get good and steady connection again
[15:31] <skreech> hii mgraesslin
[15:35] <EagleScreen> hi
[15:36] <EagleScreen> what package containts poedit application?
[15:36] <apachelogger> nixternal: ping
[15:36] <apachelogger> JontheEchidna: ping
[15:36] <apachelogger> EagleScreen: ask http://packages.ubuntu.com
[15:37] <EagleScreen> true
[15:37] <EagleScreen> looking there
[15:45] <Quintasan> lol lrn2apt-file
[15:45] <Quintasan> EagleScreen: apt://poedit
[15:45] <Quintasan> urgh, no highlighting
[15:46] <EagleScreen> thanks i found it
[15:46] <EagleScreen> but i am not liking it
[15:46] <EagleScreen> i am translating .po files with a plan text editor, is it valid?
[15:48] <apachelogger> EagleScreen: if you dont break them
[15:49] <apachelogger> you could also edit them by bitshifting if you dont break them :P
[15:49] <EagleScreen> lol
[16:10] <Quintasan> UUUUU, I hate to admit it but GEdit > Kate :/
[16:11] <apachelogger> Quintasan: how dare you?
[16:11] <apachelogger> even my weirdo german flatmate knows that Kate pwns everything
[16:12] <Quintasan> bullshit
[16:12] <Quintasan> where are snippets?
[16:12] <Quintasan> where is non sucking code completion?
[16:12] <Quintasan> :<
[16:13] <Quintasan> also vim > *
[16:14] <apachelogger> emacs :p
[16:14] <apachelogger> anyhow
[16:14] <apachelogger> Quintasan: snippets?
[16:14] <apachelogger> haha!
[16:15] <apachelogger> last time I worked with snippets was with a 1995 keyboard which was causing horrible typing
[16:15] <Quintasan> lol emacs
[16:15] <apachelogger> honestly, I craft up stuff faster than any snippeting or code completing
[16:15] <apachelogger> in that regard I am too lazy to wait to be lazy
[16:16] <Quintasan> apachelogger: well, I produce tons of use less code and I use include and int main frequently so it's PITA to type it every time I star something
[16:16] <Quintasan> start even
[16:16] <Quintasan> so snippets++
[16:16] <apachelogger> ehm
[16:16]  * apachelogger has templates for that kind of stuff :P
[16:17] <Quintasan> I don't bother with templates :P
[16:17] <Quintasan> easier to use snippet than look for template configuration :P
[16:17] <apachelogger> configuration?
[16:17] <apachelogger> haha
[16:17] <apachelogger> cp template.cpp main.cpp
[16:18] <apachelogger> taht is all I ever do
[16:18] <Quintasan> lol
[16:18] <Quintasan> I do vim lol.cpp and I have it done automatically
[16:18] <Quintasan> :3
[16:18] <Quintasan> I told ya, vim > emacs > *
[16:18] <Quintasan> :P
[16:18] <Quintasan>  /topic
[16:19] <apachelogger> then you still need to define a name :P
[16:19] <apachelogger> hence cp tTAB name.cpp is mostly faster anyway :P
[16:19] <Quintasan> you still need to type emacs main.cpp
[16:19] <Quintasan> so I win
[16:19] <Quintasan> :P
[16:20] <apachelogger> in vim you first need to go save and save before you can save :P
[16:20] <Quintasan> lol?
[16:20] <apachelogger> anyhow!
[16:20] <Quintasan> vim > emacs > *
[16:20] <Quintasan> :PP
[16:20] <apachelogger> emacs is gnu!!!!!
[16:20] <Quintasan> lol stallman mode
[16:21] <Quintasan> vim > stallman
[16:21] <apachelogger> if I were to make an IDE, first thing is that I would automize the creation paradigm :P
[16:22] <apachelogger> if I go like UbuntuOneJob *job; then press some fancy button and click on ubuntuonejob I expect the IDE to freaking create the templates already :P
[16:22] <apachelogger> in fact I would expect it to have them read by the time I give the pointer a name and ask me if I want to create that class
[16:22] <apachelogger> or if there is already such a class offer to include it
[16:22] <Quintasan> lol
[16:22] <apachelogger> or if I wrote UubuntoOneJob to tell me about that
[16:23] <Quintasan> maybe create IDE that thinks instead of you :D
[16:23] <apachelogger> no
[16:23] <apachelogger> that is a valid claim
[16:23] <skreech> emcs wins
[16:23] <Quintasan> or IDE instructed by brain waves
[16:23] <apachelogger> everything that is related to not findings something is politics
[16:23] <apachelogger> I expect an IDE to care about politics, so I can concentrate on functionallity
[16:23] <Quintasan> that would be awesome - Brain instrutced IDE
[16:24] <skreech> apachelogger: project follows mind?
[16:24] <apachelogger> aighto
[16:24] <apachelogger> I mean, maybe I am alone with this, but by the time I write code I have an exact structure of it in my mind
[16:25] <apachelogger> so the IDE should ultimately assist with bringing that to 1s and 0s
[16:26] <apachelogger> if I write this->foobar(); and foobar does not exist, the IDE can already add it to my header as private and add a template function to my current cpp
[16:27] <apachelogger> https://wiki.kubuntu.org/Kubuntu/Specs/LucidPackaging <-- shouldnt that be updated considering KO is no option for cd inclusion?
[16:28] <apachelogger> hm
[16:28] <apachelogger> "Roman wants to print to a network printer. He fills in the details in the printer settings kcontrol module and prints off his masterpiece."
[16:28] <apachelogger> the first part of the second sentence would be flagged as bug by me TBH
[16:29] <skreech> apachelogger: what about fobar() ?
[16:30] <apachelogger> skreech: I go "omg!!!!" ... add an additonal o and the ide goes "omg!!!!" and changes all occurances because this->fobar(); was the only reference anyway
[16:31] <apachelogger> if I have used foobar() twice already and accidently change one to flubar() the IDE should jump all over me telling me about the problem at hand
[16:31] <apachelogger> semantics ftw!
[16:36] <skreech> :-)
[16:43] <ScottK> New k3b alpha could stand packaging.
[16:44] <nixternal> apachelogger: pong?
[16:45] <Tm_T> ScottK: hmmm, how recent it is?
[16:46] <ScottK> Tm_T: Mentioned on planet.kde.org today as new.
[16:48] <Tm_T> ScottK: roger, planning to build with muse support?
[16:48]  * ScottK isn't planning on packaging it.  Hopes someone else will.
[16:48] <Tm_T> ScottK: oh, ok
[16:50] <apachelogger> ScottK, nixternal: shouldnt we revise the spec https://wiki.kubuntu.org/Kubuntu/Specs/LucidPackaging concerning koffice?
[16:51]  * apachelogger doesnt really get what the spec is out for
[16:51]  * ScottK how no opinion on koffice.
[16:51] <ScottK> how/has
[16:52] <Tm_T> apachelogger: what about koffice?
[16:52] <apachelogger> I dont get what the spec wants from KO
[16:52] <apachelogger> or what KO wants from the spec
[16:53]  * apachelogger starts wondering why kubuntu is around
[16:54] <Tm_T> apachelogger: spec doeans't want anything, it just states that Lucid will have KOffice
[16:55] <apachelogger> that is an incredibly useless statement though :P
[16:56] <apachelogger> oh wellz
[16:57] <Tm_T> apachelogger: how so?
[16:57] <apachelogger> cause karmic also had ko? :P
[16:57] <Tm_T> apachelogger: but was it Qt4 version? I remember atleast Jaunty had old 1.6 series
[16:58] <apachelogger> no it was not
[16:58] <apachelogger> here comes the reason why I think the spec needs to be revised
[16:58] <apachelogger> or rather should have never been written the way it was written
[16:58] <apachelogger> "KOffice 2.1 will go into main replacing the KDE 3 version."
[16:59] <Tm_T> apachelogger: hmm, agree
[16:59] <apachelogger> though, really that is because how the spec process works and how it is applied
[17:00] <apachelogger> it is trying to stack initial project considerations onto ubuntu where there really is no consideration needed, what I am saying is that how specing works these days is virtually useless to us
[17:01] <skreech> What is useful to us?
[17:01] <skreech> Or better how can speccing become useful to us?
[17:01] <apachelogger> no doubt, writting down ideas and visions and concept is very useful, not much help with the struggles of ubuntu at large though
[17:03] <apachelogger> skreech: need to think about that a bit more
[17:03] <apachelogger> documentation of ideas alone certainly cant be it :P
[17:04] <apachelogger> technically the large meta-specs of kubuntu dont help much either, from my POV all the stuff in the packaging spec should be an own spec
[17:04] <apachelogger> only put together form a general idea of what we want to do packagingwise in lucid
[17:08] <skreech> So sort of a special case spec that always gets written?
[17:08] <apachelogger> well, each portion of the packaging spec can cause problems or be endangeroud by unexpected things
[17:09] <apachelogger> like say KOffice 2.1 is not mature enough to be considered worth using by the enduser
[17:09] <apachelogger> this stuff needs to be mentioned and discussed, which is really what a spec should be used for
[17:10] <apachelogger> not only an outline for the general path, but precisely document how the general path can change if something goes wrong
[17:10] <apachelogger> say some case arises that prevents us from moving Qt 3 to universe
[17:11] <apachelogger> what would we do? and can such a case arise at all? and how important is that moving in context of all lucid specs?
[17:11] <apachelogger> in the end you should be able to walk through specs and boot whatever you deem impossible
[17:12] <apachelogger> ...at any given poin in the cycle...
[17:12] <ScottK> apachelogger: One use of the specs is when we need MIR for stuff, "Needed for approved spec" is one of the reasons that's acceptable.
[17:12] <apachelogger> so say in december you walk through the specs and notice that we can not deploy nepomuk just now
[17:12] <apachelogger> obviously that changes how lucid will appear
[17:13] <apachelogger> but you need to know how large is the impact on users, on us, on canonical
[17:13] <skreech> apachelogger: In short the specification is not specific enough
[17:13] <apachelogger> and without having these things written down you can not say .. ah well, cant deploy nepo, book spec
[17:13] <apachelogger> skreech: it is not specific at all IMHO
[17:14] <apachelogger> it is like a longer todo
[17:14] <apachelogger> a pre-development todo, as ScottK indicates
[17:14] <skreech> Is that just a matter of how it's used or is it made to be generally non specific ?
[17:14] <apachelogger> which is exactly why it should be more than just a todo
[17:14] <apachelogger> skreech: both
[17:14] <apachelogger> the fact that specs are in wiki but tracked in lp doesnt help
[17:16] <apachelogger> actually that is the biggest thing that needs to be resolved IMHO, without having ultimate control over the data (and thus the data represenation to the user) it is very very very diffcult to make sane use of specs for anything but pre-development todo lists
[17:16] <skreech> so that technical aspect needs to be taken up with the Ubuntu team?
[17:17] <apachelogger> it needs to be taken up with no one, it just needs to be fixed :P
[17:18] <skreech> Ah wonderful :)
[17:53] <JontheEchidna> apachelogger: pong
[17:55] <JontheEchidna> apachelogger, I didn't announce the doodle on ML cuz Qt 4.6 rc broked kmail
[17:57] <JontheEchidna> or rather, it broke most of my kioslaves
[18:02] <Nightrose> JontheEchidna: make one of your minions do it then ;-)
[18:03] <JontheEchidna> Well, tbh I forgot about delegating it <.<
[18:04] <JontheEchidna> So I suppose it's still slightly my fault
[18:04] <JontheEchidna> ...aside that I'm crazy enough to run pre-alpha crack
[18:10] <nixternal> pfft, pre-alpha crack from the repos? have fun with trunk...though today it is running very well
[18:11] <JontheEchidna> Qt 4.6 beta -> rc1 transition. Otherwise things went quite smooth for pre-alpha
[18:11] <nixternal> I will have to agree with that
[18:11] <nixternal> I haven't found anything broked on the netbook or desktop yet
[18:11] <skreech> zWhat you need is git kernel with trunk KDE compiled on a fresh from the servers bootstrapped gcc ;-)
[18:12] <nixternal> haha
[18:12] <nixternal> LFS w/ KDE
[18:20] <ghostcube> hmm just to mention if you start an apturl download and the kpackagekit window is open and you click close it closes .|
[18:20] <ghostcube> :S
[18:20] <JontheEchidna> amichair: ping
[18:20] <ghostcube> it doesnt minimize lol
[18:21] <JontheEchidna> apturl uses install-package I believe, not kpackagekit
[18:21] <ghostcube> hmm but the window closes :) and this may shouldnt happen or
[18:21] <ghostcube> it does continue installing
[18:34] <amichair> JontheEchidna: poong
[18:34] <JontheEchidna> amichair: nvm, I figured it out
[18:35] <JontheEchidna> amichair: I do have some comments now though :P
[18:35] <amichair> JontheEchidna: okily-dokily
[18:36] <JontheEchidna> The Krazy2 code checking utility was complaining about the use of showNormal() for showing the dialog
[18:36] <JontheEchidna> From what I gather, you're trying to restore the dialog with the same geometry/coordinates
[18:37] <amichair> JontheEchidna: what's Krazy2?
[18:37] <amichair> u can change it to show() too
[18:37] <JontheEchidna> From my experiments doing m_dialog->show(); m_dialog->activateWindow() are fine
[18:37] <amichair> actually there's still a bug
[18:37] <amichair> after the first notification, the dialog is not brought to front, it can stay hidden behind other open window
[18:38] <JontheEchidna> hmm, yeah
[18:38] <amichair> JontheEchidna: the qt docs say to do raise+activateWindow+show, but it still doesn't do it right
[18:38] <JontheEchidna> is there any way we could update the dialog without hiding it?
[18:39] <JontheEchidna> perhaps put a circular "busy" widget in front.
[18:39] <amichair> JontheEchidna: the problem is not there - even if u close the dialog and another notification pops up, the new dialog is not brought to top
[18:39] <amichair> the hide-show during update is to prevent the geometry from changing
[18:40] <JontheEchidna> Krazy is KDE's code checking utility: http://englishbreakfastnetwork.org/krazy/
[18:40] <JontheEchidna> I have a local install so I can check my various projects
[18:40] <amichair> and what does krazy say about showNormal()? why is it bad? docs say it's just like show, but restores the window if it's minimized, etc.
[18:41] <JontheEchidna> here's what it says: http://paste.ubuntu.com/330483/
[18:42] <JontheEchidna> the behavior is very window-manager specific, which is why we're seeing all these bugs
[18:42] <amichair> hmmm... not much of an explanation
[18:42] <JontheEchidna> KWin has focus stealing prevention, which is probably why it isn't brought to front
[18:43] <amichair> I used it since show() didn't help, and from the qt docs it sounded like showNormal would do better if the window is minimized
[18:43] <JontheEchidna> technically it's not minimized but hidden
[18:43] <amichair> there are two cases, hidden and minimized
[18:43] <ghostcube> the techbase has more info about the link shown there
[18:43] <amichair> it sounded like using showNormal would solve the latter
[18:44] <amichair> I hoped it would solve the former too, but it didn't
[18:44] <Quintasan> hmm 6GB is too low to compile Qt :/
[18:44] <amichair> in any case, show() is ok there too, seeing that the bug is not solved. didn't find how to do that yet, if at all possible
[18:44] <Quintasan> anyways to extend *.img file?
[18:45] <JontheEchidna> Is there a way we could both: a) destroy the dialog on close and create a new one on a new notification and b) update the dialog upon hook additions without user action and without closing the current dialog?
[18:46] <amichair> JontheEchidna: iirc that's exactly how it is now...
[18:46] <JontheEchidna> for the dialog to update it has to be hidden
[18:47] <JontheEchidna> it would be nice if it just automatically updated the dialog without me having to click the notification again
[18:47] <amichair> when closed, the dialog is destroyed and later recreated, and when notifications change, the dialog remains but is updated with new notifications
[18:47] <amichair> JontheEchidna: oh, right, that last bit would be nice
[18:48] <JontheEchidna> oh, yeah. it is doing the first bit
[18:48] <JontheEchidna> but doing the last bit should be enough to not need to do fiddly tricks to make sure focus and window stacking is correct
[18:49] <amichair> JontheEchidna: strange thing is, the window focus bug happens after the window is destroyed and recreated, i.e. on a different instance of the dialog!
[18:49] <amichair> I actually did not see it happen when just updating the dialog, though it might happen
[18:50] <ghostcube> can i somehow see what the installer does after i closed the window o.O
[18:50] <amichair> also the update hide/show is there because without it, when updated, the dialog changes size, a bit strange. there's probably a way around it, but it doesn't make much of a difference, as it does function correctly
[18:51] <amichair> JontheEchidna: what ghostcube just wrote reminded me - it would be a nice feature to have a 'details' button so u can see what's gonna happen when u click 'run action'.
[18:52] <ghostcube> :)
[18:52] <JontheEchidna> The description should show all the average user needs to care about
[18:52] <JontheEchidna> imo
[18:52] <amichair> but as always, I think we should get everything working perfectly before thinking about enhancements :-)
[18:53] <amichair> JontheEchidna: another issue: waiting for confirmation that the action ran successfully, followed by disabling/deleting the notification so it won't come up again
[18:53] <amichair> the problem is when running with Terminal=true - I haven't yet found a way to get the process result code
[18:53] <JontheEchidna> I have pondered how to do that
[18:54] <JontheEchidna> using a KProcess blocks all of kded until the KProcess finishes, but that's the only way I can get an exit code from the app
[18:55] <amichair> I played around with it a bit, but couldn't find a way of getting the result at all when running in terminal
[18:55] <amichair> e.g. using kprocess with 'konsole -e sh -c' gets close, but it still doesn't return the process exit code from within the terminal
[18:56] <JontheEchidna> oh yeah, the terminal stuff makes that even more tricky
[18:56] <amichair> maybe there's a reusable gui-console component we can use? like when update-manager shows the terminal in a 'details' box
[18:56] <JontheEchidna> update-manager is gtk
[18:57] <amichair> yeah, I just mean maybe there's a component like that somewhere that we can use?
[18:57] <ScottK> There's a KDE front end for the upgrader (the gui version of do-release-upgrade)
[18:58] <amichair> there's also update-manager-kde, but it's python, so still not exactly what we need
[18:58] <amichair> is there such a console-in-dialog somewhere else? anywhere in kde?
[18:58] <JontheEchidna> upgrade hooks are in practice few and far between. I wouldn't spend too much time worrying about this
[18:59] <amichair> JontheEchidna: so we disable/remove the hook right when clicking 'run action'?
[18:59] <amichair> and just hope it works :-)
[18:59] <JontheEchidna> That was what I have been wanting to do, yes
[19:00] <amichair> ok then.
[19:00] <amichair> so how is it removed? delete the hook file? write it down somewhere? need sudo?
[19:01] <JontheEchidna> we don't exactly have a mechanism for deleting the hook files yet
[19:01] <JontheEchidna> disabling the run button for the GUI would probably be good enough for now
[19:01] <amichair> I meant to ask how it's supposed to be implemented
[19:01] <JontheEchidna> oh
[19:01] <JontheEchidna> hmm
[19:01] <amichair> JontheEchidna: but then when the next notification pops up, the old ones (that already ran) will show up again...
[19:01] <JontheEchidna> in the past seen hooks were stored in config files
[19:02] <JontheEchidna> but then that notification will never show up again even if it needs too (e.g. a firefox restart notification)
[19:02] <amichair> the spec in the wiki should be updated with this info
[19:02] <JontheEchidna> in theory it would have the same name, but this is now a month later with the latest firefox and the notification wouldn't show because we've already "seen" it
[19:03] <JontheEchidna> update-notifier (gnome) saves the md5sum of each hook
[19:03] <amichair> JontheEchidna: the file timestamp can help there... but still, is this specified anywhere?
[19:03] <JontheEchidna> we could save the md5sums of the notifications we've seen and use that to determine whether or not to display it
[19:03] <amichair> or are we inventing a 'memory' mechanism on our own?
[19:04] <amichair> md5+timestamp will work
[19:04] <amichair> so these hooks are never ever deleted? by anyone?
[19:05] <JontheEchidna> I don't know. The update-notifier tool has a hook-delete.c file in the souce, so maybe...
[19:05] <amichair> update-notifier is the gtk frontend?
[19:06] <JontheEchidna> yeah, it handles update notifications and hook showing
[19:07] <amichair> hmm... then why did we write it again from scratch? don't they already have functional components that can be reused, and just implement the gui part?
[19:08] <JontheEchidna> It's written in C+tightly integrated with GTK/Gnome
[19:08] <amichair> what a shame
[19:08] <JontheEchidna> *...in C plus it is tightly...
[19:08] <amichair> that's just duplicate work
[19:08] <JontheEchidna> well, KDE was here first ;-)
[19:09] <amichair> they/we could have the functionality separated from the frontend, and each do it's own gui
[19:10] <JontheEchidna> I'll be back in a bit
[19:10] <Quintasan> GRR
[19:10] <Quintasan> anyone living in trunk
[19:10] <Quintasan> ?
[19:51] <Quintasan> omfg, TinyCoreLinux -> 10mb with X server
[19:51] <Quintasan> :D
[20:06] <ryanakca> Quintasan: cool :D
[20:07]  * ryanakca waits for the alternate CD to finish filling the 160GB HD on his new EeePC with random data so that he can get on with installing Kubuntu :)
[20:18] <Quintasan> ryanakca: you are encrypting your hdd?
[20:21] <ryanakca> Quintasan: Yeah
[20:36] <Quintasan> JontheEchidna: what should I know except basic and advanced data types, pointers, memory allocation in C++ before moving to OOP?
[20:37] <shtylman> plymorphism
[20:37] <shtylman> *polymorphism
[20:37] <Quintasan> :O
[20:37] <Quintasan> wtf
[20:37] <amichair> that would be under OOP, no?
[20:37] <shtylman> aka inheritance
[20:38] <Quintasan> just to be sure, I need to know this to learn Qt?
[20:38] <amichair> Quintasan: under pointers u should include also references, pass by value, pass by reference, the differences etc.
[20:38] <shtylman> Quintasan: yes
[20:38] <shtylman> also... templates is important for container classes
[20:39] <amichair> Quintasan: have u learned other languages before C++?
[20:39] <shtylman> polymorphism is what lets you have a QPushButton that inherits methods from QWidget
[20:39] <Quintasan> amichair: only Pascal but it doesn't count, does it? :P
[20:40] <amichair> Quintasan: why not? it means u have a bunch of basics covered :-)
[20:40] <Quintasan> well, I found the basics fairly easy, but when I was pointers I was totally like "WTF?!!?"
[20:40] <shtylman> haha
[20:40] <Quintasan> was :S
[20:41] <Quintasan> s/was/saw
[20:41] <amichair> there are pointers in pascal too...
[20:41] <shtylman> also.. I wouldn't just say memory allocation... I would say memory management
[20:41] <shtylman> in general is important
[20:42] <Quintasan> amichair: I was 11 back then and after that came the age that I was using computer for playing (12-14 yo)
[20:42] <Quintasan> 2 years of wasted time :|
[20:42] <Quintasan> rgreening: \o
[20:42] <Quintasan> I could get into MOTU easily if I had not wasted TWO years :/
[20:42] <amichair> I also started off with pascal, actually (well, after basic and assembler). even dabbed in delphi for a project or two a while back
[20:43] <rgreening> hey Quintasan
[20:44] <Quintasan> lol assembler
[20:44] <Quintasan> I looked at it but decided to not learn it
[20:44] <Quintasan> well I wonder if I should apply for MOTU now
[20:46] <amichair> Quintasan: while I barely had to use it in practice (other than for... umm... stuff I won't mention here), it gives a great understanding of how things actually work, which is very helpful for other languages as well, since they all ultimately translate to the same low-level functionality
[20:46] <amichair> assembler is one-to-one how the hardware-level works, more or less
[20:47]  * kb9vqf used to write device drivers for Windows 9x in assembly...ugghhh
[20:47] <Quintasan> write drivers? omg
[20:47] <kb9vqf> :p
[20:47] <Quintasan> I feel so little compared to everyone :<
[20:48] <amichair> kb9vqf: I see 3 ugghhh's in that sentence :-P
[20:48] <kb9vqf> :)
[20:48] <kb9vqf> The Bad Old Days
[20:48] <amichair> well, 4, including the actual ugghhh
[20:49] <amichair> Quintasan: at some point you'll also need a good grasp of threading, but that can wait until you feel comfortable with the basics
[20:49] <Quintasan> I'm called a hacker by my friends, you guys would be, well... hmmm I'm lacking adjective :P
[20:49] <Quintasan> amichair: you mean using multipe cores?
[20:49] <amichair> old?
[20:49] <amichair> :-P
[20:50] <kb9vqf> 13 years ago?  Yeah
[20:50] <Quintasan> :S
[20:50]  * maco points out that "assembler" is not a language but rather what turns assembly into binary
[20:50] <amichair> I was referring to that missing adjective :-)
[20:50] <maco> and assembly is the language
[20:50] <Quintasan> I was still learning to walk back then propably :P
[20:51] <maco> assembler : compiler :: assembly : C
[20:51] <maco> er wait.. assembler : assembly :: compiler : C
[20:51] <Quintasan> yay, qt built successfuly!
[20:52] <Quintasan> Riddell: add me to #kubuntu-bunker invites list please!
[20:52] <amichair> http://en.wikipedia.org/wiki/Assembly_language#Related_terminology
[20:53] <amichair> maco: ^^
[20:53] <maco> amichair: huh. alrighty then. im gonna agree on the confusing bit then
[20:53] <Quintasan> I feel lost at C++, I'm using Stephen's Prata Primer Plus as a reference but it looks like a damn long way to OOP
[20:54] <maco> because ive written assembly and a compiler but never written an assembler
[20:55] <amichair> it's confusing only as strict definition... in practice, it's always clear what someone's talking about :-)
[20:55] <rgreening> Can anyone point me to the kubuntu-dev process and who are currently approved kubuntu-dev?
[20:55] <maco> i thought the kubuntu-dev stuff didnt go into effect until archive reorg?
[20:56] <amichair> Quintasan: threading is not about multiple cores, it's about different pieces of code running 'concurrently'
[20:56] <amichair> Quintasan: which may be physically (e.g. on multi-cores) or logically (on single cores, even old ones)
[20:57] <maco> i saw the email about ubuntu-desktop being setup and i know kubuntu-dev is going through that process too... is that done now?
[20:57] <amichair> Quintasan: it's the mechanism that allows the computer to do several things 'at the same time' as far as the user, application, and code is concerned
[20:59] <Quintasan> amichair: ahh, okay, I guess I will follow the book though, I thought I could jump to OOP after pointers, data types and functions
[20:59] <Quintasan> :P
[21:00] <amichair> Quintasan: u can always jump around, as long as u get things covered eventually :-)
[21:01] <Quintasan> there should be a flowchart for learning c++ :PP
[21:03]  * rgreening wonders why I bother some days
[21:05]  * kb9vqf would help rgreening if he knew the answer...
[21:07] <rgreening> heh, Im feeling in a non-transparant mood :)
[21:33] <Quintasan> when is kde 4.4 scheduled to be released?
[21:36] <amichair> February
[21:36] <kb9vqf> http://techbase.kde.org/Schedules/KDE4/4.4_Release_Schedule
[21:36] <amichair> yep, that's where I just looked :-)
[21:36] <kb9vqf> :)
[21:40]  * amichair is really annoyed by his desktop freezing anytime there's a bunch of disk access
[21:45] <amichair> is there an equivalent of windows' FileMon? something to monitor and log which processes access which files all the time?
[22:04] <Mamarok> amichair: there is ksysguard
[22:05] <Mamarok> and ksystemlog for log files
[22:11] <amichair> Mamarok: how do I use them to log which processess access which files all the time?
[22:21] <Mamarok> that depends what you want to log
[22:21] <Mamarok> amichair: Linux logs stuff all the time and writes it to the /var/log/ folder
[22:22] <amichair> Mamarok: does it log all file accesses by all processes? that would be a helluvalot of overhead if it were done by default...
[22:23] <Mamarok> no, of course not, but there are log files for X, for hte kernel, for cronjobs, etc
[22:24] <amichair> that I know, but it doesn't help me with what I'm trying to do :-)
[22:24] <Mamarok> if you just want to see what is using up your CPU or memory use ksysguard, it shows you in real time and you can filter by CPU, memory, etc
[22:24] <Mamarok> well, if you you would tell me *what* aou actually want to do...
[22:24] <Mamarok> you*
[22:24] <amichair> and if I want to see what is using up my disk access? per file? per process?
[22:25] <Mamarok> did you even try to use ksysguard?
[22:25] <amichair> I want to log disk accesses, so that every time any processes accesses any file, it is logged.
[22:27] <amichair> yes, I use ksysguard all the time, but haven't seen how to make it do that.
[22:27] <amichair> I don't even see how to make it log the data which it does display...
[22:28] <Mamarok> hm, I never did that, I guess you would need to write a script that writes the log file, monitoring disk access by file is rather uncommon
[22:29] <Mamarok> and this would not be ksysguard anyway, but ksystemlog
[22:30] <Mamarok> I don't really see the reason for doing that, could you explain?
[22:31] <amichair> I thought ksystemlog only views logs, not creates them - but I never looked into it.
[22:32] <amichair> well, for the full story - I'm having terrible desktop freezes caused every time there's a bunch of disk access.
[22:32] <Mamarok> how much ram do you have? and what's the size of your swap?
[22:32] <amichair> while it's something very wrong on its own - no amount of disk access should make the mouse pointer freeze, I can let that go for now, and at least try to figure out *who* is doing all the access.
[22:33] <Mamarok> you might want to look at the memory usage, since low memory triggers a lot of disc access for swapping
[22:33] <amichair> I have 4g of ram, which should be plenty
[22:34] <Mamarok> so do I, but sometimes you can get into trouble anyway, especially when several "hnugry" apps are running
[22:34] <Mamarok> like downloading/upgrading and compiling with both cores
[22:34] <amichair> I've used an equivalent utility in windows which logs every file access by every process to analyze disk-related problems in the past, and my question was whether such a tool exists in linux
[22:34] <kb9vqf> You can view instantaneous disk access with iotop
[22:34] <Mamarok> you did eliminiate a memory leak of an application?
[22:35] <kb9vqf> Not sure about logging though
[22:35] <amichair> kb9vqf: yes, but it's not good enough, since it doesn't log, and when the desktop is frozen, I can neither start it nor view it. that's why I'm looking for a proper logger.
[22:36] <kb9vqf> Well, one hacky way to get around that is to run iotop through ssh on a different computer; when yours freezes there will be a frozen iotop display on the second computer
[22:36] <Mamarok> amichair: run it with output to a file
[22:36] <amichair> I have 17gib of swap, it seems
[22:36] <kb9vqf> That is, one computer as a terminal
[22:37] <Mamarok> but you will end up with a giant file not easy to read
[22:38] <amichair> Mamarok: it will be large, but if I keep it on for say 15 minutes, enough to catch a freeze or two, it should be easy enough to analyze. it's the same amount of activity in windows when I used FileMon (which does the logging), and it was always helpful.
[22:39] <amichair> right now I see 3.1/3.6 gib in use. I've seen freezes with 2.6/3.6 as well. there's plenty of ram, and I'm just swapping, say, konsole/kate/firefox. nothing that should be swapped out to disk in the first place.
[22:40] <amichair> I also ran a long SMART test on the disk, just in case, and it shows no physical error with the disk.
[22:41] <amichair> hmmmm
[22:42] <amichair> plasma-desktop is taking up 750M - sounds a bit much for a desktop...
[22:42] <Mamarok> well, if you have a lot of plasmoids...the plasma stuff is not running separate instances :(
[22:43] <amichair> I have a desktop folder, comics strip, and post-it note.
[22:44] <amichair> and this happened before the post-it note too, which I only added a week ago
[22:45] <Mamarok> I guess it's plasma, never had such a load for that
[22:45] <Mamarok> but keep in mind that on Linux, the more RAM you have, the more will be used
[22:45] <amichair> can u elaborate on that?
[22:45] <Mamarok> since it helps speeding up stuff, it's the res that counts
[22:46] <Mamarok> do you have htop?
[22:46] <Mamarok> much better than top btw, no idea why top still is default
[22:46] <amichair> thanks for the tip, trying it out...
[22:47] <Mamarok> there you can see the memory use in more detail
[22:48] <amichair> looks nice!
[22:48] <amichair> I was actually really missing lots of details in ksysguard
[22:48] <Mamarok> you can sort by different parameters with F6
[22:49] <Mamarok> well, ksysguard tells you the same, you just have to click on the column headers a lot
[22:49] <Mamarok> easier with htop IMHO
[22:49] <amichair> Mamarok: I like the graphs, trends often reveal imporant info :-)
[22:51] <amichair> is it showing processes or threads? seem to be lots of duplicates
[22:52] <Mamarok> amichair: threads, you can filter by hierarchy though
[22:53] <amichair> u mean F5/Tree?
[22:56] <ryanakca> Hmmm... bangarang works much much better when Nepomuk uses the sesame2 backend than when it uses the redland backend... compare 20-30s to an album with redland to a second with sesame2
[22:57] <amichair> Mamarok: well I still have no answer to my problem or my question, but I learned a new useful tool - so thanks :-)
[22:58] <Mamarok> you are welcome :)
[22:58] <Mamarok> play around a bit with htop, it really is a great tool
[22:58] <amichair> I am :-)
[22:59] <amichair> btw, what's VIRT/RES?
[22:59] <amichair> total / currently-in-ram?
[23:00] <Mamarok> amichair: virtual or resident memory (or residual, not sure of the term)
[23:00] <Mamarok> so if something uses a lot of virtual memory that can be neglectable if the res part is low
[23:00] <amichair> but if used, might imply lots of swapping?
[23:00] <Mamarok> like X right now here uses 1.2 GB virtual memory, but only 400 Mg or res
[23:01] <Mamarok> no, swapping would start only if the res memory is too high
[23:01] <Mamarok> which leaves less for other processes
[23:02] <amichair> oh yes, I meant from the point of view of a single process - if it has high virt and very low res, that means it's got lots of swap in use?
[23:02] <Mamarok> no, just that there is a lot of memory available :)
[23:03] <Mamarok> check the percentage of total memory it uses
[23:03] <amichair> if a process uses 1G virtual, with 100M res, does that mean it's using 900M in swap disk?
[23:04] <Mamarok> no, not at all, or do you see a high swap load in the graph above?
[23:05] <Mamarok> virtual memory != swap
[23:05] <amichair> swp shows 2G/17G or so
[23:05] <Mamarok> it has just been allocated more virtual memory because there is more available
[23:06] <Mamarok> not much, depending on how many apps are running
[23:06] <amichair> but I'm asking just to try and straighten out the terminology, as converted from windows
[23:06] <Mamarok> well, the terms should be the same, I don't think the RAM is named differently, it's the handliung that is different
[23:06] <Mamarok> -u
[23:06] <amichair> ok, so virt includes memory pages which haven't necessarily been accessed, or stored anywhere?
[23:06] <Mamarok> exact
[23:06] <amichair> oh, ok
[23:07] <amichair> so what's the actual used page count called?
[23:08] <Mamarok> that's what so much nicer done with Linux, you can even tell the system the priority of the applications
[23:08] <Mamarok> right, "page" is a bit strange, didn't hear that since dos times
[23:08] <amichair> priority as in memory use? coz cpu/scheduling priority u can do on all OSs, I think
[23:09] <Mamarok> memory allocation in Linux is much more dynamic than in Windows
[23:09] <amichair> page is the basic unit of virtual memory
[23:09] <amichair> it's the smallest atomic unit that is physically allocated in RAM and/or swapped in and out to disk (swap file)
[23:10] <amichair> usually 4k
[23:10] <Mamarok> well, see the two columns PRI and NI?
[23:10] <amichair> ya?
[23:10] <Mamarok> PRI for priority, wich is basically the same for all GUI apps
[23:11] <Mamarok> NI for nice level, where you give jobs different priorities
[23:14] <Quintasan> yay, running trunk
[23:14] <Mamarok> Quintasan: nice :)
[23:15] <Quintasan> I wonder if I should move it to my main machine not to KVM :P
[23:32] <Quintasan> hmm, how can I get bigger resolution than 1024x768 in KVM? I have tried messing in xorg.conf but with no result