IRC Logs for #crux Friday, 2006-09-08

*** pitillo has joined #crux00:07
*** Romster has joined #crux00:45
Romsterhrmm thats nice i have a port of pyrex already in contrib and someone puts a copy in opt... without even informing me..00:46
nipuLhow dare they!01:05
*** treach has joined #crux01:09
*** richi_aut has joined #crux01:14
*** mike_k has joined #crux01:21
treach@seen revolt01:24
clbtreach: I have not seen revolt.01:24
cptn@seen the_light01:28
clbcptn: I have not seen the_light.01:28
cptnwhat a great day to start fridays... ;-)01:28
treach@seen sense01:29
clbtreach: I have not seen sense.01:29
guaqua@seen my_computer_brain01:30
clbguaqua: Error: 'my_computer_brain' is not a valid nick.  That nick is too long for this server.01:30
guaqua@seen my_bot_brain01:30
clbguaqua: I have not seen my_bot_brain.01:30
mike_k just good animation for fun01:30
*** pitillo has quit IRC01:38
marojdolan: but build() is always run in a subshell01:44
maroplus pkgmk uses $PKGMK_SUPER_VARIABLE style variable naming01:45
* maro always feels like somebody has yelled at him when he look at pkgmk01:45
*** mrpk has joined #crux01:52
*** predatorhawk has quit IRC01:56
*** mrpk has quit IRC02:08
*** mike_k has quit IRC02:17
*** richi_aut has quit IRC02:18
*** richi_aut has joined #crux02:18
*** rxi has joined #crux02:33
*** pitillo has joined #crux02:37
*** mike_k has joined #crux02:40
pitilloummm need to review my bills... my isp don?t want to send me a ip....02:51
*** treach has quit IRC02:54
*** treach has joined #crux02:58
Romsteri managed to get red alert 2 to work in wine the problem is my video is way way too slow.02:59
treachRA2, a video card killer?03:00
Romstergotta be away to get it faster i have a nvidia 4 mx420 here and using latist nvidia-x11r7 drivers.03:00
Romstershouldn't be i'm getting lously video speed :(03:00
maronvidia updated their crap for X11R7? :)03:01
Romsterit worked fine in winblows, but everything in general is slow when it comes to video..03:01
maroor well, xorg 7.103:01
Romsteri got nvidia framebuffer on int he kernel too03:01
maroRomster: bad idea, I think03:01
Romsterfist time i got the nvidia driver and was hoing it be faster than the xorg-nv driver...03:02
treachNo... I remember playing that on a matrox mystique  a long time ago, and already that card was an overkill iirc..03:02
maroshould run fine on a GF403:03
Romsterthis card worked fine in my old windows so i guess i got wrong settings, but what should i look for first.03:03
maroyes, surely03:03
cptnthe question mark key?03:03
maroif it works in windows obviously it has to work in linux as well03:03
Romstersurely just its too damn slow..03:04
marocptn: at least he uses commas03:04
treachmaybe it's just wines dx5(?) implementation that sucks..03:04
Romsteranything with video makes my pc run slow.03:04
maro(not like that other guy who only uses an ending punctutation)03:04
treachmaro: at least it seems he's fucking learnt to spell most of the time. :)03:05
Romstersays gold on the wine web site, for ra2.03:05
Romsterwould ya stop picking on me, eh!03:06
maroRomster: that's the crux of #crux03:06
Romsteri'm really just lazy, more than anything else.03:06
maroalso dumb03:06
treachI'm not picking on you, I'm making positive remarks on your progress. :)03:06
RomsterI'm not dumb...03:06
treachit's prologic who's blind.03:07
treach(almost, iirc)03:07
Romsterlol O_o03:07
* rxi pokes treach's eyes .. touche :P03:07
marotreach: ah, I wondered why his Pkgfiles looked that way03:07
Romsteryeah he can jsut see. poor fella.03:07
marowhat happened to vektori?03:08
treach@seen vektori03:08
clbtreach: vektori was last seen in #crux 2 weeks, 3 days, 9 hours, 51 minutes, and 55 seconds ago: <vektori> HEhe.03:08
rxihe's gone into hiding again03:08
Romstermaro, i've been cleaning them up when i get bored..03:09
treachhe started his studies again I think. Maybe he's just still recovering from the kick off party. :P03:09
Romsterhes at the paralymipcs atm03:10
Romsterback ont he 12th03:10
Romsterthen hes got 2 days of uni work then hes birthday on the 15th.03:10
* treach tries to imagine the marathon for dead drunk people..03:10
maroare you his girlfriend or something?03:10
rxiRomster: you hear about peter  brock?03:12
rxitreach: sit outside a pub at closing time in rural australia and watch drunk people try to make it home :P03:13
treachheh, given the distances, I guess it might come close. :)03:13
cohani want to add swap encryption in a "cruxish" way, so i have to modify /etc/rc - at least, umounting swap, encrypting it, and activating swap again with an /etc/rc.d/ - script seems not very professional03:13
cptncohan: if you don't list it in fstab, you could do without the changes to /etc/rc03:14
cohannow i have two options: load a module (dm-crypt) _before_ the mount magic, when only / is mounted readonly, before the mount -a call,03:14
cptnsince it's "special swap" anyway03:14
cohanthen swap could be still in fstab, but i don't like the idea acessing modules before mount -a,03:15
Romsterpeter  brock, the racing bloke no why?03:15
marocptn: speaking of boot script order, check out #10903:15
cohanah, ic - no swap at all, not mentioning it in fstab, manually configuring it in an /etc/rc.d-script03:15
rxiRomster: yeah .. was killed this morning at a rally03:15
*** pitillo has quit IRC03:15
Romsterah fuck03:15
cptncohan: exactly03:16
cohanand it's users choice to place "cryptswap" or whatever in the beginning of /etc/rc.conf's SERVICES to make sure swap is available when the mem-consuming stuff starts03:16
Romsterffs first steve now peter :(03:16
cptncohan: if you want your rootfs encrypted, I guess there's no way around initramfs03:16
rxiRomster: yeah its pretty sad week :(03:16
* maro considers starting a "Jihad Ubuntu" in reply to the Christian Ubuntu effort03:17
Romsterwho the fuck next...03:17
Romsterhe was a good bloke :(03:17
rxiRomster: yeah i loved him .. go holden03:17
cohancptn: root system encryption is not my issue right now - on the (root) system itself there should be no really critical data03:18
cohanit is more that i start services from other, later mounted encrypted partitions (mounting them needs manual interacting anyway, giving the passphrase),03:18
Romsterholden rocks03:18
Romsterhe'll never be forgotten...03:18
Romsterwhat a bad week03:19
cohanand this would be compromised by unencrypted swap03:19
cohanonly thingy not "cruxish" is where to place the information about the swap partition(s) if they are not listed in fstab anymore03:20
*** Romster has quit IRC03:21
cptncohan: usually, rc.d/* is supposed to be configured via rc.conf03:21
cptnso add encryped_swap=() there03:21
cptnsource rc.conf from your service script03:22
*** Romster has joined #crux03:22
cptnmaybe use <servicename>_* for the variable name03:22
cohanokay, sounds cruxish - or introducing a /etc/cryptab and providing a port for rudimentary crypted partitions support03:22
cptneither way should be fine03:23
cptnmaybe a mount wrapper would work too?03:24
cptnthen use dm-crypt in fstab03:24
cptn(as type)03:24
cptnsee mount.fuse03:25
cohani'll check that03:25
cohaneither case, a small port seems very reasonable for this approaches03:28
cptncan't really comment since I haven't played with it03:29
cptnif you want, talk to Viper_ to have an idea how to make it useful for most cases03:30
*** destruct has joined #Crux03:32
mike_krxi, got link on that sad rally?03:33
cohanhe told me yesterday he did not use it for swap encryption yet03:33
cptnah, you can't use it for other partitions as you have to give it a password03:33
cohanat least, without asking for a password, encryption is not very useful ;)03:35
cohanright, swap just uses random passphrases03:35
cptnI see03:36
cohanhm, i just realize... something like crypttab (and supporting in a way similiar to other distros) should not use things in /usr/sbin - because user could be mounted later - so the corresponding cryptsetup - port would have to change its binary locations?03:37
cptnyou mean if /usr itself is encrypted?03:38
cptndoes that make sense?03:38
treachHan: if you still care about kaffeine, you could bump the version to 0.8.103:38
* tilman cares about his caffeeine03:39
cptntilman: use forums then!03:40
cohannot in my case - okay, and actually in no case where / is not encrypted as well (and this needs initramfs)03:40
cptncohan: how complex is the setup of the encrypted swap anyway?03:44
cptnmaybe it could just go into rc.local?03:44
cohanright now, it is a (very ugly) one-liner03:44
cptnyou could provide a script hiding the ugliness03:45
cptnand add /usr/sbin/mount_enc_swap /dev/hdaX to rc.local03:45
treachtinyurl ftw03:47
rxilaziness ftw03:47
*** destruct_ has quit IRC03:48
marocptn: that usage of rc.local justifies even more :)03:49
cohani like the mount.encswap or similiar best right now03:49
cptnmaro: I'm sure there's a similar use case justinging that the services are already started03:50
maro(since the encrypted partitions could likely be required for functioning of the services)03:50
tilmandamn bug tracker03:50
tilmancauses nothing bug bugs03:50
* Romster sighs03:51
rxilike bugs bunny?03:51
marocptn: such as? (not that I doubt there are any, just that I'd like to think of solutions to those)03:51
tilmanrxi: bugs bunny is a completely different matter03:52
cptnmaro: doing anything which requires the lo device being there03:52
treachthis (is) bugs tilman. :)03:53
cptnor a real internet connection even03:53
marocptn: couldn't that be done in a service?03:53
maroaka rc.d script03:53
cptnjust like the mounting03:53
*** pitillo has joined #crux03:54
cptnI mean, the mount --bind you mention in your report03:54
cohandamn, the mount.cryptswap approach lacks - because dm-mod is not autoloaded03:56
cptncohan: autoload it from the script then03:56
maroI was thinking done in the packaged rc.d scripts, not making local ones03:56
cohanbut btw... if someone uses libdevmapper as a module... and refers to those mountpoints in his /etc/fstab, where do they load it?03:56
maroif you have a concrete example of "doing anything which requires the lo device being there" it'd be easier to find the right solution03:57
cohani can of course just load the module, that works for me (tm), but is not an approach which lets this solution work on most common setups (or at least spills errors if, for example, dm-mod is fixed in the kernel, not as module)03:57
maro(something that people would want to do in rc.local, that is)03:57
marocptn: perhaps implementing rc.local as /etc/rc.d/local is the best solution? :)03:58
cohanhm, never expected "doing this thing right" could raise that many considerations...03:58
cohanmodprobe -q dm-mod is perhaps "the right way"03:59
cptnmaro: I have no example right now04:00
cptnmaro: okay, here it comes:04:04
cptngetting ping time to a server04:04
cptnor maybe two or three04:04
cptne.g. to select a sourceforge mirror04:05
marois that really things you want to do in local additions to /etc/rc?04:06
maroyes, the point being that rc.local should be local additions to /etc/rc04:06
cptnah, yes04:07
marooh well04:07
marowith such common use cases I cannot argue that $services should be started after the boot process (rc + rc.local) has completed04:08
cptnokay, I'll close the bug then04:09
Romsterbah i put uptiemd in rc so its ment of been put into rc.local...04:11
Romsterdespite the README said..04:11
maroRomster: nope04:11
marothat won't work04:11
* Romster scratces my head.04:11
marobecause services are started before the basic boot has completed04:11
Romsterok i'm confused.04:11
maroread /etc/rc.multi04:11
Romsterah i've seen local and rc but not multi..04:13
Romsterone of these days i need to go though every file and see what it does.04:14
maroyou could also just read /etc/inittab and check what files are referenced for what runlevels, but obviously that solution is only for the lazy :)04:14
Romsterif all local changes go into rc.local, that dosn't apply to sevvices on the local machine that have been installed.04:15
Romsteryes i've been into inittab before04:16
Romsterwell aditionsbe a better word.04:16
tilmanmaro: you're being a bit anal in that mail o_O04:16
marotilman: hmm, sorry, I really didn't intend to04:17
Romsterlol o.o04:17
Romsterwhats your averages for glxframes?04:18
Romster4512 frames in 5.0 seconds = 902.376 FPS04:18
maromine's ~400k04:19
Romsternot sure if that is good or a phaetic figure i have04:19
Romster400k frames?04:19
Romstergood lord makes this look shit..04:19
Romstergeforce 4 mx42004:20
maroI'm just an idiot used to reading metric04:20
Romsterit dpsn'04:20
Romsterit dosn't even run ra2 as a decent speed...04:20
maroit should have been 40004:20
marotilman: yikes, sourceid=firefox :/04:20
Romsterthen what em i ment to use for a benchmark!!!04:21
maro500 with xaa04:21
tilmani didn't bother to remove that bit04:21
tilmanso what04:21
maroRomster: benchmarks are useless for end users04:21
Romsterand wht do i need to alter to increase the speed...04:21
tilmanRomster: quake3 (demo)04:21
Romsteri jsut want it to work(tm)04:21
marotilman: not you, the firefox-google alliance :(04:21
Romsteryet my video is way too slow.04:21
tilmandoom 3 (demo) if you have good hardware/drivers :D04:21
marono, doom 3 if you have don't give a shit about freedom04:22
tilmandoom3 is as un-free as quake3?04:22
marodoes in run properly on anything but proprietary drivers?04:23
marowhat? (need to get me one of those :D)04:23
Romstererr i just tryed cube and that ran fast04:23
Romstermaybe its just a wine issue?04:23
marowhy are you using wine?04:23
pitilloanyone has doom 3 port?04:23
marodoesn't q3 run natively on linux?04:24
Romsterfor red alert 204:24
marooh... :(04:24
tilmanx800 gto something04:24
marohehe, evidently suse is shipping gcc 4.1.304:24
tilman(arb renderer, arb2 is b0rked)04:24
maropretty bleeding edge considered upstream haven't even released 4.1.204:24
marotilman: cool, had no idea they were progressing that nicely :)04:25
maroperhaps I should find my 9500PRO behind the sofa04:25
maroit's been there for 3 years04:26
Romsterhrmm think its a wine speed issue...04:27
marocptn: what's the point of a bug tracker if you just close with "As agreed upon on IRC"?04:28
tilmanwine is pretty cool these days04:28
tilmanit was such a PITA to set up and use some years ago04:28
marotilman: indeed04:29
treachno shit..04:29
treachonly, then you really needed it for various stuff. Now it's mostly just for fun. :/04:30
tilmani started playing red alert (1) some days ago04:30
tilmanbut unfortunately the (uber cool) videos are displayed in interlaced mode04:31
tilmanwhich spoils the fun >:04:31
treachoh, yeah. in game videos, so much fun...04:31
tilmannot in game04:31
tilmanthey shelled out millions of $$$ for bad actors etc04:32
treachheh, speaking of funky vids.. I remember some odd game where the videos where shown upside down if you played with certain drivers.04:33
treachThat's a nice touch.04:33
cohanmount.swap does not work out, because it is not executed for mount -a, and defining a new type (e.g. cryptoswap) is mounted, but not used by swapon -a04:35
cohanreally a pity, just providing a mount.swap would have been _very_ cruxish ;)04:36
cohanwait, it still works... just needs two entries per swap, one for the encryption, one for the swap04:41
cptnmaro: well, you said that you cannot argue against it04:48
cptnso if for bug reports which are unconfirmed by the devs, once the reporter agrees that it's not a bug I see no reason to keep it open04:49
cohanare there also mount-helpers for umount?04:52
cohanthe same one should be called, i believe04:55
marocptn: it makes it totally useless for later reference04:55
*** richi_aut has left #crux04:57
*** mike_k has quit IRC05:00
*** Auge^ has quit IRC05:01
cohanperhaps things like that rc.local stuff are not "bugs", and should be collected on a "system design considerations" page?05:10
marocohan: it was marked as a feature request05:10
cohanthose page could list pros/cons of an implementation, and explain why it is as it is, and is there as a reference for later discussion/reevaluation05:11
marocohan: great idea, IMO05:11
marois there a public wiki yet?05:11
Romsterwtf saving a kwrite file caused kwin to crash in kde O_o05:13
*** Auge^ has joined #crux05:14 nice05:14
Auge^the familyname of my hoster (a friend)05:15
tilmanAuge^: "re." means "regarding". the irc idiom only works in german ;)05:15
tilmanbut i'll shut up now05:15
marotilman: actually it doesn't05:15
maroit's latin05:15
tilmandoesn't what?05:16
marothat's why "Re:" is correct and "RE:" isn't05:16
cptntilman: it doesn't german!05:16
tilmandieser satz kein verb!05:16
maro<tilman> Auge^: "re." means "regarding"05:16
tilmanit doesn't???05:17
tilmanmaro: sources? :D05:17
cohanre. and re/re: are completely different perhaps? ;)05:20
cohani thought in emails RE: was for REsponse05:20
cohanand in skat, re is was follows contra05:20
treachREply. Many possibilitys.:)05:20
cohanarghl, screw it, typoday - wanted to say "re is what follows contra"05:21
tilmantreach, cohan: probably true. the german v of outlook replaces it with "AW" for "antwort" ("reply")05:21
cptn.oO(AW: Re: AW: Re:)05:21
marothat's why outlook is incorrect and most unix MUA's are correct05:21
Auge^notes is more than incorrect... by the way05:22
tilmanmaro: ah05:22
cohani think i slowly get this mount-thingy work cleanly, just a05:22
*** the-ruediger has joined #crux05:22
marotilman: now you know the source of those "Re: RE:" threads :)05:23
cohani just need to figure out who is responsible for /etc/mtab to also do clean unmounts05:23
cohanAW: is also the default on too localized linux distros/muas05:23
marohave symlinking /etc/mtab to /proc/mounts been considered?05:24
maro(busybox considers a "real" /etc/mtab obsoleted and the only argument it has to support it is for kernels which don't have a /proc)05:27
Hantreach, would you like to take it over?05:29
cohani am aware of this, i still wonder how filesystems mounted by helper scripts get listed in /proc/mounts05:29
cohanat least, if they do not get an entry there, umount -a will not call them05:30
treachNo, I'm not official in any form :)05:30
tilmanmost httpups aren't official05:30
marocohan: you mean if the kernel doesn't know of the mount? :P05:30
treachtilman: be carefull.. Catapultam habeo. Nisi pecuniam omnem mihi dabris, ad caput tuum saxum immane mittam. ;)05:30
tilmanalea jacta est!05:31
tilmanaeh, sunt05:31
treach(I have a catapult. Give me all of your money, or I will fling an enormous rock at your head.)05:31
cohanno, the correct latin phrase is "alea jacta est", only the german translation is pluralized05:32
cohanotherways, it should read "aleae jactae sunt", iirc05:32
*** Romster has quit IRC05:32
* tilman mostly slept through his latin classes05:32
cohanbut my latin ... is actually not existing ;)05:33
cohanjust this alea-jacta-est-thingy somehow stuck in my mind05:33
cohanmaro: yes, it is about mounts the kernel is not aware of - like those "mounts" setup by cryptsetup05:34
cohanwell, they are actually mappings... nethertheless, they should be removed - or my approach to set up this mappings within a mounthelper is also "stupid"05:34
*** Romster has joined #crux05:34
treachtsk. the romans didn't have "j". Didn't you see Indiana Jones? :P05:35
tilman! :D05:35
Romsterfar out i locked up kde desktop pfft05:37
cohanj/i is essentially the same ;) still, writing iacta looks a bit like a aero company05:37
marocohan: noone else have dealt with this problem before?05:37
treachRomster: I guess something more is amiss there.05:37
cohanmaro: dunno - still googling. and other distros make extra stuff and /etc/crypttab for it05:37
Romsteron latist version too..05:38
Romsterkwin started it all after i saved a dile with kwrite.05:38
Romstervery odd05:38
treachcohan: j wasn't added until 16th century iirc.05:40
cohanmaybe - clerics did many other ugly things with latin as well ;)05:40
*** bd2 has joined #crux05:41
cohanhm, perhaps i am trying to solve the wrong things. cryptsetup relies on libdevmapper - so a cruxish way to establish and user device-mapped stuff should be introduced first?05:43
cohanand this symlinking of /etc/mtab to /proc/mounts seems to be not a perfect idea, regarding to man mount:05:48
cohanIf you are not so unwise as to make  /etc/mtab  a  symbolic  link  to05:48
cohanoh dear, this turns out to be a little bit more complex - with loopbacks and devmappings beeing no mounts etc.06:06
maroin the absence of vektori, would someone look at
maro(who knows, maybe even apply!) :)06:21
maroor should I just bug it and wait for vektori to return?06:22
cptnthere's no patch header!06:22
marocptn: is it hard to imagine what it does?06:22
tilmanmaro: looks good, though the reordering of the sysconfdir would bug me ;)06:22
marotilman: I did it because --mandir has to be passed in the new ver06:23
tilmani figured ;D06:23
maroRomster: when would vektori return? :)06:23
cptnmaro: at least it's longer than three of the patches in shadow :-)06:25
tilmanlighttpd does:06:25
tilmankill `cat /var/run/`06:25
tilmanmaro: so if you stop and start the uptimed service, your uptime will still be lost, right?06:26
marocptn: obviously, it removes the README since the /etc/rc modification isn't needed anymore :)06:26
*** brointhemix has quit IRC06:27
marotilman: I'm not sure it's lost, I just know what vektori's argument for having it directly in rc was that it should only be run once - OTOH at least arch creates the bootid on every start06:27
marobut arch being arch, I'll investigate it ;)06:27
maronope, it isn't lost06:28
tilmandid you contact vektori already?06:29
marono, the patch is 10 minutes old and vektori have been away for 2 weeks06:30
treach@seen jaeger06:30
clbtreach: jaeger was last seen in #crux 4 weeks, 1 day, 16 hours, 22 minutes, and 23 seconds ago: <jaeger> don't know KDE, myself06:30
treachI got a mail from him yesterday.06:30
marobut I'm all for not tankering with other maintainer's ports unless it's security related06:30
maroso I suggested bug'ing it instead and wait for him to return06:31
tilmantreach: err, jaeger isn't vektori? ;)06:31
treachtilman: uh.. elevate your perspective a bit. :)06:31
tilmanmaro: agreed, do that06:31
tilmanmaro: do you have sufficient access rights to assign it to him?06:31
*** mike_k has joined #crux06:33
treachtilman: in this instance he is. Principially, at least since he's in a black box and we can't know if he's jaeger or vektori right now. :D06:33
tilman@seen vektori06:34
clbtilman: vektori was last seen in #crux 2 weeks, 3 days, 13 hours, 17 minutes, and 24 seconds ago: <vektori> HEhe.06:34
tilmantreach: errrr. okej06:34
treachOk, we'll leave Mr Heisenberg out of this. :)06:35
tilmanyou mean schroedinger?06:36
treachthe point was, that just because somone isn't here, or even on irc, and hasn't been so for a long time doesn't necessarily mean you can't reach them.06:36
treachtilman: yeah.06:37
tilmanokay, but we know jaeger is around on irc06:37
tilmanand we know vektori just isn't06:37
treachthe point still stands.06:37
marotilman: nope06:38
marotilman: I don't have any privileges at all, anywhere06:40
maroshould probably learn suck ass better :)06:40
maro :)06:42
maroyou were already there :)06:43
marocptn: !!! :D06:44
cptnnot done yet :-)06:45
* maro sends cptn virtual beer06:45
cptndoes that make me virtually drunk now (pun intended)06:46
cptnlocally, it's not %p1 and %v106:47
cptni.e. the version separated from the path06:47
marocptn: how do you feel about removing the -v code when a script has been implemented? :)06:48
cptnmaro: check the man page :-)06:48
maroah, I'm so slow :)06:48
marohm, what in the case of multiple dups? multiple lines of output?06:50
cptna -> b06:50
cptna -> c06:50
cptnthat's at least what it did before06:50
marook, cool :)06:50
maro(perhaps a -> b, b -> c?)06:51
tilman"being more sensitive"06:51
marono idea what code changes that would require, and obviously a minor detail06:51
cohanah, screw it, crypting the swap cannot be done perfectly cruxish as long as /etc/rc and /etc/rc.shutdown or crux at all are devicemapper-aware,06:53
cohanbut i believe making crux overall devicemapper and loopback-aware or introducing something like /etc/crypttab support is not cruxish at all06:53
cptnyet maybe worth noting somewhere?06:54
marocptn: not at all, just commit with a message "added crypt stuff" ;P06:55
cohanyes, i'll now build a port for  my mount.cryptswap with a README pointingthat out06:55
*** Romster has quit IRC06:55
marocohan: how heavily would they need to be modified?06:56
cohanat least, on standard setups, it works very nicely by just adding mounts.crypttab and making one addtional fstab entry for every swap partition06:56
*** Romster has joined #crux06:57
cohanjust if you use a swapfile (it is also supported) you will need to trigger some manual devmap removal and unlooping on shutdown to make a clean unmount06:57
cptnso maybe the rc.conf + service approach would help there?06:58
cptnsince it's run on shutdown06:58
cohanmaro: i don't know yet, i played with several scenarios - but always ended up that libdevmapper needs to be somehow integrated if you want to "do it right" - i'll try to write down my findings06:58
cohancptn: yes, but i don't like the idea of turning off swap somewhere in rc.conf - or it must be the last/first entry06:59
cohanthat feels like the wrong place for sth. like swap, imagine some memory intensive action takes place afterwards, whyever..07:00
cptnalthough if your services already swap at startup, you'll run into bigger problems later on07:01
cohanthats definitly right, of course - but i fear swap needed on/during shutdown more07:03
cohanbut no problem with placing "cryptswap" service as one of the first in SERVICES, it will then be called as one of the last on shutdown, i believei?07:04
cohanyes, looks like07:04
tilmanmaro: i replied to your mail when my connection died07:07
tilmanso now i have to go through it all again :'(07:07
cptncohan: have a good look at R_SERVICES07:07
cptnIOW, yes, the shutdown order is reversed07:09
cohanyes, just saw that, thats why i said: < cohan> yes, looks like07:10
marotilman :/07:10
*** the-ruediger has quit IRC07:13
*** the-ruediger has joined #crux07:15
Auge^who uses bttv-driver with xawtv? which versions? i just get a black picture with: This is xawtv-3.95, running on Linux/i686 (
*** clare has joined #crux07:17
marocptn: a small addition to - fc-cache, mkfontdir & friends07:17
cptnadded, thanks07:18
maroand the gnome "???" could be "gconftool-2 --makefile-install-rule"07:18
tilmani hope to get to that tomorrow07:19
tilmani'm itching07:19
marowoot :)07:19
cptntilman: note that there was some resistance towards this07:19
cptntilman: last time I brought that up in an IRC meeting07:19
marogeneric support for install scripts could turn out more useful07:19
tilmancptn: i'll read through that before writing any code07:20
maroat least much more flexible - then the auto stuff could be implemented at the pkgmk level07:20
Auge^noone with bttv here? :/07:21
*** koefz has joined #crux07:22
Auge^ok... it seems the nvidia driver is the problem :/07:30
tilman\o/ :D07:35
*** the-ruediger has quit IRC07:36
tilmanwe need planet.crux.nu07:40
tilmanall the cool kids have planets07:40
rxiexcept poor pluto07:41
*** koefz has quit IRC07:41
*** mrpk has joined #crux07:41
guaquapluto is being robbed from its planethood :<07:43
treachapparently a conspiracy among the enemys of pluto. :/07:43
rxii wonder how big the nut ring for that was :P07:44
tilmanwilliam lewis ftw07:44
guaquaanyone using mailcheck_imap-script for irssi?07:45
tilmanno, i've written a leeter mailcheck-imap app07:45
tilmanbut thanks for asking ;)07:45
Hantreach, enemies07:45
tilmanthank god we have a spellchecker bot07:46
guaquatilman: care to share it?07:46 (embrace2)07:46
HanYou systematically make that mistake.07:46
tilmanits dependencies will make you cry though, i think07:46
treachtilman: would that be that plugin to gkrellm? Oh, wait it was supposed to be for rss, wasn't it? :p07:47
*** koefz has joined #crux07:47
tilmanrdf maybe07:47
tilmanor both. or maybe rdf==rss, who knows ;)07:48
guaquai could get that for linux07:48
guaquabut now i have a windows desktop07:48
tilmaneverytime you boot a windows desktop, god kills a kitten07:49
rxii almost killed a kitten on the way to work :P07:50
*** PainBlade has left #crux07:50
tilmanthat reminds me07:51
tilmanmaro: i saw a badger corpse (roadkill) on the way to work some weeks ago07:51
tilmani almost picked it up to install linux on it :D07:51
rxihehe i thought you were going to say it reminded you of maro :P07:51
cptnmaro: BTW 'dup' didn't report multiple occurances so far07:52
guaquawhat desktop do you people use in linux?07:52
cptnmaro: I'll look into that though07:52
treachguaqua: have a look in the ports, I guess most of them are there.07:53
guaquawell just curious :/07:53
guaquai couldn't add shortcuts to e17 so i guess i07:54
guaquai'll have to go for something else :)07:54
rxitheres no decent themes for e17 yet07:55
tilmanwinter is good you fool07:55
treachtoo bright07:55
guaquaHan: :)07:55
tilmantreach: :D :D07:55
tilmantreach: i think i've heard lots of comments like "e16 only has dark themes, bah!"07:55
treachtilman: I actually belong among those who kind of like the Solaris CDE theme. However, the gay who came up with the pink crap should taken out and shot at dawn.07:56
rxitreach: why wait til dawn?07:57
treachIt's traditiona, I think. :)07:57
rxiahh .. we just shoot first and ask questions later :P07:57
Auge^any ideas how to set the commandline-options (-nogl,-nodga) in xawtvrc? dga = false/0/no doesn't work :/08:01
cohancrux dsl users out there - how do you ensure that pppoe is started at boottime, and even recovers (ie connects later) when adsl is not available at startup?08:01
guaquamaybe i'll start using linux on desktop more now that i don't use dc08:02
guaquacohan: doesn't require any special tricks for me08:02
guaquai just run ddclient as daemon too08:02
cohanbut ddclient does not trigger any connection? or don't you use rp-pppoe ?08:06
guaquai do use rp-pppoe08:06
guaquabut i guess it reconnects by itself08:06
cohanand you start pppoe-start on startup?08:07
guaquai'll take a look08:07
cohanpppoe-start is nice, it _usually_ also reconnects - but it bails out if on startup no dsl connection is possible,08:07
cohanor if somewhere in between there is no dsl available for a longer period of time08:07
guaquacould be08:08
cohanright now, i use a cronjob to run pppoe-start periodically, just to be sure - but i don't like it08:08
guaquai've had it fail a few times too08:08
guaquanetbsd had a nice ifupd08:08
rxirp-ppp has an option to autoreconnect if the connection s dropped08:09
cohanrxi: but it fails if the connection is _not_ initially working08:11
cohanat least, pppoe-start finishes then08:11
cohanthis left some gateways unreachable in the past after power failures, there dsl needed an hour or two to be available again08:12
rxithats why i use proper adsl routers like cisco, etc08:13
cohanokay, extra hardware and extra ip (mis)configure trouble08:13
cohanand cost intensive - then i don't need a linux gateway at all, could just use a plain cisco or even cheaper dslrouter08:14
*** Dudde is now known as B\Dudde08:15
rxiand easy to use, etc08:16
rxiif your using a box solely as a gateway its going to cost more to run in power than a adsl router08:17
cohanthats right. but running a router AND a box - i somewhat dislike the idea. i love the NAT and traffic control I have when running linux as a gateway08:19
cohanplus, my gateways mostly do isdn and/or cheap fileserver stuff08:19
cohanand the cronjob thingy works nice, just ist not very beatyful. that's why i asked08:21
Auge^fine. xawtv finally works... but not with xv (§$#* nvidia driver!)08:33
tilmanRomster: want a desert fox?08:36
tilmanto cheer you up cause someone put pyrex in opt :)08:38
Romsterheh oh well, a simple email would of been nice, i've got it moved out of contrib on next update now.08:40
*** pitillo_ has joined #crux08:43
Romsteri didn't have pixal shader on and i had wine on win xp, changed to 98 now and turned on pixel shader dunno if that'll speed up ra2 or not08:43
Romsteri'll try it later.08:43
Romsterwondering if i can doa global install of ra2 too, to like /var/.wine/drive_c/... so that not just the users files in ~/.wine get sued but also the global ones too.08:45
Romsterarn't even sure if thats possable i'll have to google around.08:45
Romsterfood is more important atm :)08:45
*** the-ruediger has joined #crux08:52
*** destruct has quit IRC08:55
*** pitillo_ has quit IRC08:55
*** pitillo has quit IRC08:56
Auge^f***ing nvidia...09:00
Auge^bttv vs. xawtv vs. nvidia09:01
RyoSmh never used those together so cant tell09:03
Auge^be lucky.09:03
*** rxi_ has joined #crux09:11
*** destruct has joined #Crux09:13
Romsteri got nvidia to work for once, b4 i was missing all my fonts on it.09:13
Auge^ah, back to version 7676 - much better09:15
cptnAuge^: you did run gl-select, did you?09:16
Auge^i didn't... but the problem is based on DGA, not GL09:17
jjpkNvidia provides its own files when you use the proprietary drivers. By default x11 files are used, so you have to run it.09:19
*** B\Dudde is now known as Dudde09:30
*** clare has quit IRC09:31
*** Brzi has joined #crux09:35
*** aon has quit IRC09:43
*** aon has joined #crux09:44
*** clare has joined #crux09:46
Auge^mh. but what about DGA? if i want to use DGA with nvidia? i cannot find an appropiate answer for this question... just a lot of how to run apps without DGA :/09:47
*** Brzi has quit IRC09:47
Romsteri just realised that on kmix that the vertical and horzintal radio buttons ont he configure options are reversed!!! on version 2.609:52
*** Romster has quit IRC10:04
jjpktor ftw.10:10
*** Romster has joined #crux10:19
cohanhm.. some initial cryptswap version is out now:10:20
cohanthx for all the hints, still not too lucky10:22
cohanperhaps a rc.d-only-version is better, which just uses swapon -s to see all swap partitions and "remount" them encrypted?10:23
*** the-ruediger has quit IRC10:25
cohanthat would mean no fstab- and mount-fuzzling at all10:25
cohandrawback: whenever swapoff -a ; swapon -a is called during system operation, you end up with non-encrypted swaps at all10:27
cohanups, nasty Pkgfile error - fixed now10:30
*** Husio has joined #crux10:36
*** mike_k has quit IRC10:43
*** mrpk has quit IRC11:19
*** MrX has quit IRC11:32
*** MrX has joined #crux11:35
*** mrpk has joined #crux11:39
*** mrpk has left #crux11:40
*** Romster has quit IRC11:44
*** laod_ has quit IRC11:46
*** laod has joined #crux11:48
*** the-ruediger has joined #crux11:49
*** lasso has joined #crux11:59
treach .. dang it, why couldn't it be "turkey"? :D12:10
tilmansifuh has been here, recently12:11
tilmandid you notice that?12:11
treachwell, I saw it in the log.12:11
jjpkYeah, he yacked about Steve Irwin's death.12:12
treachstill "a no-longer-crux-user" as well. :)12:12
jjpkDidn't someone who used to come here try to get him a job in sweden with a faked university degree? :D12:13
tilmani think i remember that12:13
tilmanvaguely :D12:13
*** Husio_ has joined #crux12:14
jjpkOh right, it was some random swede who kept changing his nick on irc.12:14
treachSSHHH don't speak of the devil.12:15
jjpkThen he came bitching about it here, asking if there was a forum or a ML to publically expose. ;)12:15
jjpktreach: don't tell me you are afraid.12:15
treachkopete doesn't haven an ignore function... :/12:16
jjpkThere is nothing but humor in that.12:16
treachright. It's totally lame.12:16
jjpkIt does not have ignore? o_o12:16
treachnot for irc.12:16
jjpkoh ok.12:16
treachapparently the devs are total kooks..12:17
aonirc is where it's needed!12:17
aonyeah, sifuh12:19
aoni remember him12:19
aoni even talked to him when he was here! :o12:19
jjpk <- why am I not surprised? :D12:19
treachhe's a legend in his own lifetime. You've gotta give him that at least. :p12:19
aonhehe :)12:20
treachcigar? Crack pipe, more like it.12:20
treachMan, I *really* can't stand windows anymore, even though they actually have essentially fixed my initial main gripe with it. (bsods)12:21
*** Romster has joined #crux12:22
tilmanwe left because of the stability problems, and stayed because of the convenience12:22
aoni wish there were proper edas for some other OS than windows12:22
aonalthough there could be some for hpux or solaris or something12:23
treachtilman: indeed. For instance, I'll never understand whats so fucking hard about keeping everything related to a user in his $HOME...12:23
treachone would think it's pretty simple, but alas apparently totatlly out of reach from redmond..12:24
tilmanthat's not only redmond's fault12:25
tilmanyou have to blame app developers, too12:25
treachbah. that's just passing blame.12:25
tilmanwindows has this 'documents and settings' folder that can be used for per-user data12:25
treachexactly, and it's a nuciance.12:26
*** Husio has quit IRC12:26
treachnusiance, even.12:26
*** Husio_ is now known as Husio12:35
*** acrux_ has joined #crux13:04
*** acrux has quit IRC13:05
*** Husio has quit IRC13:15
*** predatorfreak has joined #crux13:21
*** koefz has quit IRC13:43
*** sen_ has joined #crux13:45
*** treach has left #crux14:00
*** treach has joined #crux14:06
*** acrux_ has quit IRC14:06
*** acrux has joined #crux14:07
treachprt-get listorphans -> glibc I guess I won't be needing that then. ;)14:18
jjpkMust be waste. ;)14:18
jjpkprt-get said so!14:18
treachyeah. :-)14:18
RyoS <- what could have caused this error? :S14:20
treach(makes me think of one of my more fatal n00b mistakes, a loong time ago when it removed that package on slackbox while trying to slim it down... >.< )14:20
jjpktreach: ah those n00b days. I *almost* miss them.14:21
bd2RyoS, it's not recommended, but you could try to comment "make check" in gcc's Pkgfile14:23
RyoSi will try that14:23
RyoSbut i am not sure why it brakes the installation14:23
bd2RyoS, do you try to make own gcc-4.1.1 port? I saw cptn have one14:24
RyoSautogen is not there.. but where should it be?14:24
RyoSno not my own14:24
treachbrake? Brings to a grinding halt, I'd say.14:24
RyoSits jw's port14:24
*** Husio has joined #crux14:24
bd2RyoS, right. jw == cptn14:24
RyoSso just change the make check to make -k check?14:28
bd2RyoS, yes. Or (better), install that port
RyoSwhy didnt han add a .footprint.. i just looked with prt-get fsearch for autogen14:30
bd2don't know why it's neither in contrib nor provided in opt, and gcc 4.1.1 port don't include it in "Depends on:"14:30
* Han checks14:30
cptnwell, I build 4.1.1 on august 2614:31
treachinteresting that glibc htus has an indirect buildtime dep on guile..14:31
treachs htus/thus14:31
cptnand don't have any autogen port installed14:31
bd2cptn, maybe it's side effect of autotools changes?14:31
Hanautogen depends on guile :-S14:31
HanIt's not a part of autotools14:32
RyoSi see14:32
cptnof the 4 gcc 4.1.1, only Han catches a non-zero return value of make check14:33
bd2btw, there are lot dependencies for the "make check" ->
RyoSi see14:34
Hancptn, which is pretty clever since I don't expect anyone to install guile and autogen and what not just to run a check on make check.14:35
HanBut it's nice if it works ;-)14:35
cptnwell, my port doesn't run fixinc...14:36
Hanermh `fixinc'? That's the target which requires guile?14:36
cptn"Necessary to run `make check' for fixinc."14:36
Hancat gcc-4.0.3-fixinc.patch14:37
cptnunder the guile dep on the link bd2 posted before14:37
Hanhmmm clever :-)14:37
cptnRyoS: can you post a full buildlog?14:37
Hancptn, if you replace the version patch with my ``sed -i "s/$version/& (CRUX)/" $name-$version/$name/version.c''14:38
HanI'd be happy. :-)14:38
*** predatorfreak has quit IRC14:39
RyoSsorry cptn, didnt log it14:39
RyoSi will install autogen now14:39
RyoSbetter, i am installing it right now14:39
RyoStry to build it again after that14:39
aonnew cdrkit \o/14:40
bd2Han, just curious, can you show your pkginfo -i | wc -l ? ;-)14:40
Hancptn, and I think you should claim maintainership of that port. Otherwise people keep bugging Per.14:40
cptnHan: mmmh, any reason not to use the VERSUFFIX define?14:41
cptnor did you just mean to prefer sed over a patch?14:41
bd2Han, hahaha. Do you really use them all? ;-)14:41
Hanno of course not14:42
Hancptn, yes, using sed in this case is much easier maintainable.14:42
cptnI'd just prefer to use the VERSUFFIX way14:42
cptnas it seems cleaner to me14:42
HanIt will also cost you 5 minutes of annoying work whenever you have to bump that version ;-P14:43
cptnnot necessarily14:43
bd2Han, so, you just making ports and maintaining them?14:43
cptnI usually don't rename the patches14:43
cptnso unless the do big changes to version.c, no changes are needed14:44
cptnbut I agree it's more convenient14:44
cptn(via sed)14:44
Hanbd2, errr it's not so clearcut14:44
bd2some people can't maintain ports they *use* every day, but you're maintaining ports that you really don't use. That is knighly :-)14:46
RyoS <- this is what autoget finally fails on14:46
RyoSdependies compiled fine14:46
cptntreach: have your crystal ball handy?14:47
bd2RyoS, paste more output. It's not sufficent14:47
treachcptn: sorry, I think you still haven't returned it since you borrowed it the last time. :)14:47
Hantreach, say did you want to take over the kaffeine port?14:48
HanI actually don't use it at all.14:48
treachHan: No, I said I didn't since I'm trying to keep my lousy ports off the radar as far as possible. :)14:48
Hanah, ok. >:-)14:49
HanAnyone else?14:49
RyoSenough output bd2?14:49
cptnSCM_ sounds like scheme14:50
treachcould it be something present in an earlier version of guile?14:53
RyoSmh? dont know, wouldnt think so14:54
treachguile contains a whole bunch of stuff like that, but I can't find those particular references in the current version.14:54
RyoSmh.. weird..14:55
cptnI just build gcc again... no problem here14:56
cptn(Except for a footprint missmatch which I'm looking into now)14:56
RyoSwhich version of guile do you have installed?14:57
cptnneither do I have autogen14:58
RyoSso you just skipped the make check?14:58
HanNo, he patched away the check target which needs guile14:58
RyoSoh, how?14:58
cptnRyoS: it's all in the port in my repo14:59
cptnhas been for quite a while14:59
Hancheck the port14:59
RyoSi see14:59
cptnso if it keeps failing, please log the complete build14:59
RyoSi will try it again14:59
cptnor use Han's gcc which handles the situation gracefully :-)14:59
cptnwell, the port15:00
RyoSwell, the warning in the description scared me away actually :P15:00
RyoSDescription: latest stable version of gcc, _NOT_ recommended for daily use.15:00
cptnwell, if you wanted the tried and true, you'd stay with core/gcc, right?15:01
RyoSwell, i wanted to rebuild this version from core, failed for me..15:01
RyoSbtw: why is libiconv not in core anymore?15:01
Hancptn, disgracefully. ;-D15:02
cptnRyoS: was it ever?15:02
RyoSi thought it was15:02
cptnmost apps are fine with glibc's iconv support15:02
HanRyoS, the version in core should build, or else you should report what went wrong15:02
RyoSif i would knew what went wrong han, i wouldnt be so clueless now ^^'15:03
RyoSwell, i am building cptn's port now with the patches15:03
HanWhat was first, the chicken or the egg?15:03
RyoSHan: first build failed because i deleted libiconv before building15:03
Hantsk tsk tsk :P15:03
RyoSthe second one failed for some unknown reason for me15:03
cptn$ locate libiconv15:04
RyoSmaybe make check too15:04
cptnwhy would it fail because of that?15:04
RyoSi habe absolutley no idea15:04
*** Husio has quit IRC15:04
*** sen_ has quit IRC15:04
RyoSwith the standard libiconv port from yhafaris repo i sometimes needed to edit Makefiles manually adding -liconv to the LIBS...15:05
Hanbecause he removed it while he was building.15:05
RyoSwithout it went better, but some still try to use it, and cant find it (of course)15:05
*** sen_ has joined #crux15:05
RyoSthis is how it failed before with gcc 4.0315:06
cptnRyoS: did you by change force installation of libiconv?15:07
cptncomes from glibc15:07
RyoSi didnt force it15:07
cptnis this file still there?15:07
RyoSits not there15:07
cptnwell, yhafri's port has it too though15:07
cptnso unless you forced installation of libiconv from yhafri, you couldn't install it15:08
RyoSits there15:08
cptnwell, if it's still there, then I can't see how you install libiconv15:09
cptnunless you still have that port installed15:09
*** Husio has joined #crux15:09
cptnas it overwrites both /usr/bin/iconv and /usr/include/iconv.h15:09
cptnwhich are in glibc as well15:10
HanI bet you can't repeat this problem15:10
RyoSoh you dont know me :þ15:10
HanI've seen a friend of mine getting my computer to hang with two keystrokes.15:11
RyoSlol ^^ thats what my lil cousin did to me :þ15:12
HanShe told me she had bad vibes with PC's and I didn't believe her.15:12
*** kingruedi has joined #crux15:13
jjpkmaybe computers can smell fear.15:13
RyoSyou better do if a woman says so15:13
tilmandid she go for ctrl-alt-backspace? ;)15:13
tilmanor magic sysrq + b?15:13
cptnPower-off, Any key15:13
RyoS.. *G*15:14
cptnah, I remember the first Mac keyboard I used had a power switch :-)15:14
HanOh well, I rescued 77 vacation photographs of her. She had a bad memstick and they seemed lost.15:14
HanEver since she kind of worships me. ^^15:16
*** the-ruediger has quit IRC15:18
*** Husio has quit IRC15:21
*** Husio has joined #crux15:27
deus_exAuge^: wrt DGA : nvidia no longer supports dga, hence xawtv breakage.Try tvtime or xdtv as tv app.15:40
Auge^yeah... now i am using an old bttv driver to use xawtv... but i will try some other tv apps...15:43
deus_ex'Old' bttv driver?Aren't you using bttv from kernel?15:45
Auge^args... old nvidia driver of course :)15:45
Auge^new bttv driver.15:45
Auge^too late for brain usage =)15:46
deus_exYou need more coffee, perhaps ;)15:48
deus_exI use latest nvidia, bttv and xdtv with no problems.15:49
Auge^not yet... i have flu...15:49
RyoS=======> Building '/usr/ports/packages/gcc#4.1.1-1.pkg.tar.gz' succeeded.15:49
Auge^i guess so... mplayer with newest driver works well too :)15:49
RyoSyay :)15:49
cptnRyoS: :-)15:51
treachand now for the complaints about stuff that won't build. ;-)15:52
cptnRyoS: so what did you do?15:52
RyoSi first removed your patches.. :þ sry, leaving them as they are worked fine now15:52
RyoSwhat did you say included iconv.h?15:52
RyoSok thanks15:53
RyoSshould i recompile my kernel after gcc update?15:53
cptnno need to do that, no15:53
*** Husio has quit IRC15:53
RyoSi see, ok :)15:53
deus_exAuge^: This might be of interest to you:
deus_excptn: Congrats on new prt-get release :)15:57
deus_exDoes it make coffee yet?15:59
treachit finds orphans at least. like glibc. ;D15:59
deus_exPoor glibc, an orphan.Sad,really...16:00
cptnerr, listorphans isn't really new16:00
Auge^deus_ex: i know this thread... and i've build a package nvidia-old which works well on kernel :)16:00
treachcptn: Ok, I hadn't noticed it until now, so I presumed it was.16:01
cptndeus_ex: thanks; it's just a couple of small changes this time16:01
cptnjust wanted to push out those changes, some have been in subversion for quite some time already16:01
deus_exAuge^: Good for you.16:01
cptntreach: it was added in 5.1216:02
cptnin april, if we can trust the changelog :-)16:02
treachheh, that's what happen when you are satisfied with the way things are. :)16:02
deus_excptn: Is there a roadmap for prt-get somewhere?16:03
cptndeus_ex: not really16:03
cptnprt-get is currently a rather low priority16:03
bd2C rewrite? :-)16:03
cptnunfortunately, I like working on it16:03
cptnbut there are a couple of other things to do16:03
cptni.e. there are still a couple of non-patched ports in core16:04
deus_excptn: I'm sure you can handle it :)16:05
cptndeus_ex: that's a list of things I wanted to do at some point in time16:06
cptnnot everything is really wanted by me anymore16:06
cptnand some require major changes16:06
cptnbut it's probably the closest thing to a roadmap16:06
deus_ex"might people like ansi colors?" Yes, colors are good :)16:08
deus_exWhat is 'dependency injection', in a nutshell?16:09
deus_exHow does that work?16:10
cptnwell, let's say a new version of GTK starts depending on dbus16:10
cptnprt-get sysup wouldn't "inject" dbus16:10
cptnsince it's not an update per se16:10
cptnit may however be that building gtk wouldn't work then16:10
cptnso sysup --insject-deps would install all new dependencies before updating the rest16:10
cptnit's one of those annoying side effects of our dependency model16:13
cptni.e. not enforcing installation of dependencies16:13
treachthat sounds like a feature of limited value, honestly, since not forcing deps is a major feature of the current dependency model :-)16:14
treachAll IMO obviously.16:14
deus_excptn: Does sysup outputs that it needs to install foo in order to update something?16:15
cptndeus_ex: well, it doesn't inject dependencies yet16:15
cptnbut once implemented, it will definitely have a test mode16:15
cptnso you'd see what would added16:15
Hancptn, don't stray too far. :-)16:16
cptnHan: well, the lack of time ensures that :-)16:16
treachRemember KISS is waiting with a hammer behind every corner. :P16:16
cptnit's more for big updates, like gnome 2.14 -> 2.1616:16
cptnI guess such beast don't really work well without such mechanisms16:16
cptnI fear we might also run into similar problem with x1116:16
cptntreach: heh, I guess prt-get has many non-KISS features already16:17
cptnI just try hard to make them optional16:17
treachX11 and gnome is not KISS, maybe crux should be limited to shell apps only. :P16:17
Hancptn, it seems like httpup is allergic to c-ares, an optional libcurl dependency16:17
Hangot to do some work to verify that.16:18
bd2speaking of time... cptn, have you discussed pkgutils C rewrite "design issues" mentioned by tilman once here. Is some drafts ready as now?16:18
cptnbd2: there's nothing beyond the what's in the wiki16:18
cptnunless tilman has worked on it privately16:19
bd2I see..16:19
tilmanif drafts are ready the'll be announced on crux-devel16:19
cptnbd2: as far as I'm concerned, I've been mainly working on handing over infrastructure stuff to others lately16:20
bd2tilman, can we look into unfinished drafts? :-) just for feeding the curious :-)16:20
cptndocumenting setup etc.16:20
Hancptn, c-ares is very nice btw:
tilmanthere's no unfinished drafts16:21
bd2tilman, ah.. sorry. I've misunderstood you then.16:21
cptnHan: I'll have a look16:22
deus_exI thought that you'll be making libpkgutils, keeping pkgutils as scripts ?16:22
Hanhmmm works fine on Linux16:22
cptnHan: I actually merged your Makefile suggestions in httpup 0.4.0i16:23
tilmandeus_ex: maybe we'll call the new shit 'thenewshit' just to piss everyone off16:25
tilmanincluding USE flags16:25
tilmanand a CRUX user representive team16:25
HanYou did! :-)16:25
HanLets seeee...16:26
treachs /team/comitté/16:26
cptntreach: :-)16:26
tilmanall that16:27
tilman Checkout e_utils/core16:28
HanIt looks very familiar indeed. =)16:28
tilmanhttpup-repgen-ignore going bonkers?16:28
tilmancan someone tell Han about the 2M+ core file he's putting in contrib?16:29
treachtilman: actually I think you've managed the impossible, and escaped a black hole. :-)16:30
tilmanyou're making me feel dumb and confused way too often ;)16:32
treachI'm not trying.. I've got to improve my communicative skillz, I guess. :/16:33
Hanhmmm for some reason it's not portable. I have to dig deeper16:33
treachtilman: Either I misunderstood something, or he actually responded to you, so I guessed he might have unignored you. I was quite possibly wrong though.16:34
Auge^ah. nice. package creating is very simple :)16:35
treachDo you still think it looks harder than making packages for pkgsrc? ;D16:35
* Han has nightmares of making ports for openbsd16:36
tilmantreach: i think he responded to cptn (-> makefile merge)16:36
treachHan: tilman told us you've pooped in contrib.. are you aware of that?16:37
treachls -lh /usr/ports/contrib/e_utils/core16:37
treach-rw-r--r-- 1 root root 4.5M Sep  8 23:29 /usr/ports/contrib/e_utils/core16:37
Hantreach, there are other ways of bringing that message.16:37
Hanerare humanum est16:38
Auge^treach: i thought and i think so... crux ports system is much simpler than netbsd's one ;)16:38
HanAuge^, as defined by the word `crux' ;-)16:39
Hantreach, fixed16:39
treachHan: of course it is. I wasn't trying to be mean, just a somewhat misdirected attempt at humor I fear. :)16:39
Hancptn, both the httpup and prt-get new releases work ok on OpenBSD16:40
HanAuge^, I ported cruxports to openbsd :-)16:40
HanI bet it's a nobrainer to port them to netbsd as well.16:41
Auge^so... i have webspace and a port... let's see :)16:42
HanThat's real mens talk!16:42
Auge^just one dependency: x11... mh16:42
HanMost people don't mention dependencies like that.16:43
cptnHan: good to know, thanks16:50
RyoS=======> ERROR: Footprint mismatch found:16:54
RyoSMISSING   -rwsr-xr-x      root/root       usr/lib/pt_chown16:54
RyoSNEW       -rwxr-xr-x      root/root       usr/lib/pt_chown16:54
RyoS=======> ERROR: Building '/usr/ports/packages/glibc#2.4-1.pkg.tar.gz' failed.16:54
RyoScan i ignore this mismatch?16:54
Hanhow the hell did you get _that_ mismatch?16:56
RyoSi have no idea o_O16:56
RyoSits from your port btw ^^'16:56
treachit's that magic black thumb again.16:56
RyoSi said before you dont know me...16:56
HanAnyway, if you are really lazy you can, but you have to chmod 4755 that file16:56
RyoSthats what i will do *G*16:57
HanRyoS, have you set some funny fstab options on that partition?16:57
RyoSnot really ._.16:58
* Han looks at RyoS with the evil eye16:58
RyoSno you cant16:59
RyoSyou are a nice guy16:59
* RyoS tries to hide16:59
Hanhmmm funny. irrsi claims it's 00:00 while screen claims it's 23:5917:01
RyoSits a lie! :þ17:01
RyoSyou run irssi on another pc17:01
HanNo, honoustly the same PC17:01
treachscreen a bit... disconnected?17:02
HanSat Sep  9 00:03:59 CEST 200617:04
Hanindeed. irssi uses the right time.17:04
Hanperhaps it's just a low refresh-rate.17:05
RyoSwhat should i do know? *points at nopaste*17:05
cptnRyoS: follow the instructions in the error message17:05
cptnas if it was a white rabbit17:06
HanFollow the white rabbit!17:06
HanThat white rabbit. ^^17:06
tilmana white, cute desert fox17:07
treachRyoS: use -kw. You're playing with the semi-big boys now, sure you don't want to back down to the provided packages on the iso? ;-)17:10
RyoSi know17:12
RyoSnever ;D this is the last option...17:12
RyoSbo:~$ firefox17:12
RyoS/usr/lib/firefox/firefox-bin: error while loading shared libraries: cannot open shared object file: No such file or directory17:12
RyoSuh :S17:12
RyoSwhat the hell?17:12
RyoSno app startx17:13
HanLooks like your system is foobar17:13
treachnah. reinstall the glibc and gcc from the iso, and it should be ok. FU maybe, but not BAR. ;)17:15
RyoSworks again... downgraded from the iso..17:17
treachlesson conclueded.17:17
RyoSbut how can i now upgrade without banging the system..?17:17
Hanwait until the new release. :-)17:18
treachbanging the system? That sounds extremely dirty..17:19
RyoSi know treach.. :/17:19
RyoSwell, without glibc.. can i still update gcc right?17:19
* treach takes a decisive step away from RyoS17:19
RyoShey i am clean17:19
RyoSis there no way to update glibc..?17:28
treachNo. It's done with black magic.17:29
tilmani think people have pulled that off before17:29
RyoSblack magic.. i see..17:29
RyoSmaybe i should involve my schools headteacher then?17:30
RyoSright ._.17:32
*** predatorfreak has joined #crux17:32
Auge^mh, no "portspage" in ports? is there some similar script?17:43
treachit is in the ports17:43
tilmani think it's not17:44
Auge^ <- says no17:44
treachfunny. fsearch says yes17:44
tilmanIt's now included in the prt-utils package.17:44
treachAuge^: prt-get is very handy. we have been arguing with cptn that there should be an obligatory quizz on the prt-get manpage before the installation proceeds, so far without much sucess though.. :-)17:48
Auge^which name should i use? my nickname or my real name?17:54
tilmanup to you...17:55
*** lasso_ has joined #crux17:58
Auge^so. next step.18:00
*** Romster has quit IRC18:01
*** acrux_ has joined #crux18:05
*** Romster has joined #crux18:07
*** lasso has quit IRC18:11
*** acrux has quit IRC18:11
*** acrux_ has quit IRC18:19
*** acrux has joined #crux18:20
Auge^now i have uploaded my httpup repository and subscribed to crux-contrib... so... feel free to test my port(s)...18:20
treachif you want to do some selftesting there is prtverify18:25
Auge^mh, where to find?18:29
treachsame place as portspage..18:29
treachfsearch is a great tool..18:29
treach(prt-get fsearch)18:30
Auge^mh...great tool... but can't find prtverify18:37
*** kingruedi is now known as the-ruediger18:38
treachthing:~% fsearch prtverify18:38
treachFound in /usr/ports/opt/prt-utils:18:38
Auge^mh... it seems like my ports are 1 week older than yours *oops*18:40
Auge^ok =) my mistake...18:41
treachshit happens :]18:42
Auge^no errors =)18:44
Auge^time for bed... good n8!18:45
treachheh, I know enough german to understand that.18:51
*** laod_ has joined #crux18:53
* tilman hits the sack as well18:54
treacha good idea indeed. good night.18:55
*** treach has left #crux18:55
*** the-ruediger has quit IRC19:08
*** brointhemix has joined #crux19:22
*** the-ruediger has joined #crux19:36
*** lasso_ has quit IRC19:50
*** the-ruediger has quit IRC19:55
nipuLgreetings, from the world of tomorrow!!19:57
*** predatorfreak has quit IRC20:11
*** predatorfreak has joined #crux20:12
*** Romster has quit IRC20:46
*** clare has quit IRC21:03
*** brian|lfs has joined #crux21:11
*** aon has quit IRC21:28
*** brian|lfs has quit IRC21:28
*** aon has joined #crux21:29
*** brian|lfs has joined #crux21:29
*** brian|lfs has joined #crux21:31
*** alancio has joined #crux21:31
alancioanybody home?21:33
nipuLnope i'm at work21:53
alancioyou must work at a NOC21:54
rxi_nipuL: you work pretty late .. i fucked off at 1222:01
*** Romster has joined #crux22:07
nipuLi finish here in about 15 minutes, but then i have a quick onsite job after that22:09
nipuLthen i can finally have a weekend22:09
alanciowhat time is it there?22:14
alancioin australia?22:14
*** deus_ex has quit IRC22:22
*** deus_ex has joined #crux22:30
*** laod_ has quit IRC22:38
*** alancio has quit IRC22:54
*** mrks has joined #crux22:57
*** mrks_ has quit IRC23:02
*** morlenxus has quit IRC23:05
*** morlenxus has joined #crux23:06
*** treach has joined #crux23:36
* treach rubs his eyes... "I could have sworn was the latest when I went to bed.."23:38

Generated by 2.11.0 by Marius Gedminas - find it at!