IRC Logs for #crux Sunday, 2008-05-18

cruxbot[contrib.git]: libsndfile: 1.0.18pre20 -> 1.0.18pre2100:06
nipuLhaha, you just pasted my example00:08
nipuLi made a typo00:08
nipuLpludev00:08
predatorfreak...00:09
* predatorfreak slaps nipuL 00:09
predatorfreakDAMN YOU!00:09
predatorfreakDAMN YOU!00:09
nipuLcalm down jabroni00:09
nipuLplus it's still broken00:10
predatorfreaknipuL: << God-damn it.00:10
*** rxi has left #crux00:10
nipuLdon't shoot the messenger00:11
predatorfreakI hate hal I hate hal00:11
nipuLeven if the messenger did make a simple typo00:11
nipuLso why did you pick it up when jaeger quit?00:12
predatorfreakBecause I was using it at that point00:12
predatorfreakI've since just been maintaining it00:13
Romsteromfg...00:13
Romsteri'm glad i'm holding off on updating my system.00:14
predatorfreakArgh I need to put in policy kit.00:15
Romsternooo.....00:15
Romsteri haven't even looked though policy kit yet.00:15
Romsterhmm is http://www.mpfr.org/mpfr-current/ working for anyone?00:17
*** rxi has joined #crux00:17
*** f1y has joined #crux00:19
predatorfreakFucking hal.00:21
predatorfreakOH MY GOD YOU MUST BE KIDDING ME.00:21
predatorfreakPolicyKit REQUIRES PAM <<00:22
predatorfreakFuck that shit, I'm brute-force hacking hald.00:22
predatorfreakIrk I have to use a hal.conf from 0.5.10 for this thing to work00:26
nipuLroot needs to be included in hal.conf00:27
*** predatorfreak has quit IRC00:35
nipuLthat or put root in the plugdev group00:35
cruxbot[contrib.git]: libdivx-legacy: Pkgfile cleanup this is the same version redone to be consistant with libdivx00:37
*** predatorfreak has joined #crux00:42
predatorfreakThere, now I seem to have hald working.00:42
predatorfreakAfter much fiddling00:42
predatorfreakIt needs to allow stuff to both root and plugdev, although what the fuck is up with that plugdev stuff I have NO idea.00:43
cruxbot[contrib.git]: libdivx: Pkgfile cleanup00:43
predatorfreak/etc/rc.d/hal does NOT start hal via another user00:43
Romsterno idea. seems to work without the plugdev group.00:45
cruxbot[opt.git]: hal: Fix problems related to config, hopefully.00:46
Romsteri honestly think PAM is the worst idea ever.00:46
predatorfreakRomster: I made hal's pre-install make the non-existent plugdev group00:48
predatorfreakand I added both plugdev/root permissions so both work.00:48
predatorfreakSo hopefully that kills all the problems.00:49
Romsterpredatorfreak, shouldn't the hal.conf have a \ on the end of the line..00:49
predatorfreakRomster: It works.00:49
Romsterhope so.00:49
Romsterhmm seems \ on end of arrays is pointless?00:50
Romsterlines in a array.00:50
predatorfreakRomster: It works <<00:50
predatorfreakIt's the format they use00:51
predatorfreakSo what the fuck ever.00:51
Romstersomething i'll look into escaping new lines in bash arrays might not be needed.00:52
nipuLpredatorfreak: where is the hal.conf?00:52
Romsteryou didn't git add hal.conf ?00:53
predatorfreak...00:53
* predatorfreak cries00:53
nipuLis it late there?00:53
Romsterpredatorfreak, i think this is not your day.00:54
predatorfreakIt's 2AM00:54
predatorfreakBut primarily I got caught up trying to fix it00:54
predatorfreakthat I forgot to run git status00:54
nipuLseems like you need a HALiday00:54
nipuLXD00:54
Romsteryou should be asleep what happened to your sleeping routine.00:54
Romsterrofl cheap pun.00:54
cruxbot[opt.git]: hal: Add missing hal.conf.00:55
predatorfreakI'm probably going to get chewed out for this later00:55
predatorfreak4 hal commits back-to-back <<00:55
Romsterheh00:55
Romstercptn did a few in a row for qt4 too.00:56
Romsterso i think your safe.00:56
predatorfreakYeah00:57
predatorfreakStill less commits in a row than him :D00:57
Romsterah well00:58
jaegerpredatorfreak: the sed had to be updated a couple times with new releases01:29
jaegerit just changed the user="0" line to the plugdev group so you don't have to be root to use hal01:29
jaeger /dbus01:30
predatorfreakjaeger: The problem is.01:30
predatorfreakIt currently won't work like that.01:30
predatorfreakAs /etc/rc.d/hald starts it from root01:30
predatorfreakand dbus will refuse to give it permissions if it's not in the plugdev group.01:30
predatorfreakBut, thanks for the explanation anyway :D01:30
predatorfreakI fixed the problems that I saw, retained both options for working with hal.01:31
predatorfreakSo that it can be started via /etc/rc.d/hal and as a user in the plugdev group.01:31
predatorfreakjaeger: Ideally, hal should be started from a dedicated hal user.01:32
predatorfreakBut, as you can imagine, I've about had it with hal for tonight01:32
predatorfreakI'll implement that tomorrow.01:32
jaegerI didn't come up with that, anyway, I just adjusted it. that came from upstream01:33
jaegernot the sed exactly but the instructions to change it01:33
predatorfreakWell, whereever it came from, I've solved the immediate problems and gotten hal working.01:33
jaegernot sure what you mean, is that a new change? It's always worked as is in the past01:33
jaegerdbus had two separate sessions, a user and system, then hal starting as root just worked with it01:34
predatorfreakjaeger: Remove all the stuff related to user="0"01:34
predatorfreakand watch it fail to start.01:34
predatorfreakIt doesn't work at all from root here.01:34
jaegermust be something new01:34
predatorfreakjaeger: As I said though, the ideal thing would be to make HAL run as haldaemon.01:35
predatorfreak(I don't understand why it DOESN'T right now)01:35
jaegerwell, if that's the right way, do it :)01:35
predatorfreakjaeger: <predatorfreak> But, as you can imagine, I've about had it with hal for tonight01:35
predatorfreak:)01:35
predatorfreakTomorrow.01:35
predatorfreakWhen I don't feel like gutting hal with a knife01:35
jaegerfair enough. need to go to sleep, myself, just got home and it's 01:3501:35
predatorfreakjaeger: Good night.01:36
jaegernight :)01:36
*** jaeger has quit IRC01:36
*** mrks has quit IRC01:37
*** dropbear has quit IRC01:37
*** teK has quit IRC01:37
*** dropbear has joined #crux01:39
*** teK has joined #crux01:39
*** rxi has left #crux01:48
*** mrks_ has joined #crux01:53
*** mrks has joined #crux01:53
predatorfreakOkay, I lied, I'm just going to wrap up the hal changes today <<01:55
cruxbot[opt.git]: hal: Run as haldaemon, not root.01:55
predatorfreakFreaking 5th hal commit <<01:55
* predatorfreak waits for "YOU'RE RAPING THE GIT TREE" compliants.01:55
Romsterhehe02:11
RomsterI just went and fixed a tv set blown up power supply.02:12
Romsterturned out to be a simple repair 2 diodes and a new fuse.02:12
Romsterbe nice if all things where that simple.02:12
*** joacim has joined #crux03:16
nipuLdid you test that last commit, when i tried to run hald as haldaemon it failed03:20
Romsterdid you run the pre-install script too?03:24
Romsteri haven't updated my hal yet.03:24
predatorfreaknipuL: Works fine here03:29
predatorfreakMake sure you run the pre-install and have a plugdev group.03:29
predatorfreakI tested it before committing and it should work.03:30
predatorfreakBut the pre-install script has to be run03:30
*** onestep has joined #crux03:40
*** Rotwang has joined #crux03:42
nipuLlol03:46
nipuLdid you forget to commit the new md5sum?03:47
predatorfreakThe md5sum is up to date.03:50
predatorfreakwhat the fuck03:52
RomsterMISSING   d16e51644aef3590445eb69c3e1214a9  rc.hald03:52
RomsterNEW       ef74f2a02bea1462976c9ed56f1a6b97  rc.hald03:52
predatorfreakI know I updated the md5sum and yet it's wrong...03:52
Romsterhmm03:52
*** treach has joined #crux03:52
*** jue has joined #crux03:52
*** ChanServ sets mode: +o jue03:52
cruxbot[opt.git]: hal: Somehow the rc.hald md5sum was wrong, fix this.03:53
predatorfreakGod I'm gonna get slaughtered for hal being a bitch tonight <<03:53
* Romster facepalms03:53
predatorfreakRomster: I have no idea why the md5sum was wrong03:54
predatorfreakI remember doing pkgmk -um right before committing03:54
Romsteri've had that happen before.03:54
Romsteryou did git add the md4sum too?03:54
Romstererr md5sum*03:54
predatorfreakI didn't need to it was already in existence <<03:55
predatorfreakSigh, now, let's just hope no fucking new hal releases come out for like 3 months03:55
Romsterno when you update you need to git add .md5sum03:55
Romsteror it won't be included in the commit..03:55
predatorfreakRomster: git diff catches it for me.03:55
predatorfreakDo git commit -a -m "BLAH"03:56
predatorfreakSo long as the md5sum existed BEFORE it should get updated03:56
Romsterok that doesn't explain why it got missed.03:56
predatorfreakRomster: I could swear I did pkgmk -um before I pushed03:56
predatorfreakand commited03:56
predatorfreakI dunno to tell you the truth.03:56
predatorfreakMaybe I'm going senile <<03:56
Romsterone of them mystery's03:56
Romsterlol your too young to be.03:57
juebtw, jaeger's sed worked fine with hal 0.5.10, but was broken with the update to 0.5.1103:57
Romsteryeah. which is why patch is better it'll catch errors. sed wont you have to remember to inspect the code.03:58
Romsterbut you all knew that.03:58
predatorfreakjue: I still couldn't get it to run using only a sedded hal 0.5.10 config03:59
predatorfreakSo I have no bloody idea how 0.5.10 managed to run03:59
predatorfreakand dbus bitched about plugdev not existing.04:00
juepredatorfreak: just compared both hal.conf, didn't use it myself04:02
predatorfreakjue: Honestly, I'm considering putting hal up for adoption, I do not use hal anymore myself.04:02
jueand for 0.5.10 the '48,55d' seems reasonable04:03
predatorfreakI still use dbus, but nothing I use needs hal.04:03
cruxbot[contrib.git]: gcj: code will detect if jdk is installed and omit /usr/bin/jar04:03
cruxbot[contrib.git]: skype: remove new redundant dependency dbus is now in qt404:03
cruxbot[opt.git]: sqlite3: update to 3.5.904:16
cruxbot[contrib.git]: darcs: update to 2.0.004:18
cruxbot[contrib.git]: libvorbis_autov: Add pre-install.04:23
cptnRomster: quick questin04:28
cptn*question04:28
Romsterya04:28
cptnno ports modify prt-get.aliases04:28
cptnexcept for yours04:28
cptnwhat does that tell you?04:29
Romsteriptables-imq has for ages yeah.04:29
Romsterso it's not allowed?04:29
cptnthat's not my question04:29
cptnI'm just wondering about your motivations to do these kind of things04:29
Romsterit's to avoid the core iptables being installed when i'm using iptables-imq that's been patched with the IMQ patch.04:30
Romsterif it's bad i'll pull iptables-imq to my personal tree.04:30
f1yis there a possibility to read informations about vendor, model and name of motherboard in console? dmesg, lspci and such show only chipset04:31
cptnRomster: again not my point04:31
Romsterlspci -vv maybe.04:31
Romstercptn, k...04:31
cptnI'm just wondering why you keep adding things to your ports that no one else does04:31
cptnI mean, some things might be worth considering for others too04:31
cptnand make ports more consistent overall04:31
cptnbut I just keep finding such things in your ports04:32
cptnbut only yours04:32
Romsteri think of all the possibility's that others don't think of. or real issues that i've found a fix for.04:32
cptnand then you keep them for yourself04:32
Romsterno people can easily look at my ports...04:32
Romsterthere there... not hard.04:32
Romsteri'll document it on a wiki page.04:33
cptnno, please don't04:33
cptnyou might thing they're fixes04:33
Romsterk04:33
Romsterso what do you want me to do.04:33
cptnI guess I'll start a "what do to and what not" on packaging04:34
cptnit seems that the current guidelines are not very clear04:34
Romsterok fine i'll remove it.04:34
Romsterthis shit is going to the politics04:35
cptnI call it consistency04:35
treachRomster: lol, do you know what packaging and accounting have in common? :P04:35
treachcreativity isn't very popular with the authorities. :>04:36
predatorfreakJesus christ, since I just added one of those thingies now I gotta nuke mine <<04:37
predatorfreakScrew it I'm keeping my aotuv port in my repo04:37
jueRomster: your ports are not better than others if you try to do things different, please don't do that. If you have additions or improvements discuss them on the ML before using it, please.04:37
cruxbot[contrib.git]: iptables-imq: moved to my personal repo due to crux team04:38
cptnpredatorfreak: ?04:38
predatorfreakcptn: as my libvorbis_aotuv port is basically just a packaging of the aotuv fork of libvorbis, I added a pre-install to do what you're chewing Romster out for.04:39
*** DarkNekros has joined #crux04:39
predatorfreakSince libvorbis_aotuv replaces the need for libvorbis.04:39
cptnah04:39
predatorfreakBut I don't feel like being next to get chewed out for using that.04:40
DarkNekroshi everybody ;)04:40
cptnwhy not add it, and request an alias entry?04:40
Romsteri am sick of the shit i cop here.04:40
predatorfreakcptn: It's not important enough.04:40
cptnpredatorfreak: it doesn't hurt04:40
predatorfreakand that requires a bump of prt-get, which I really don't believe is warranted.04:40
cptnbut the pre-install breaks since the alias file is not protected in pkgadd.conf04:40
predatorfreakthe pre-install is safe.04:41
cptnso updating prt-get would remove the aliases added in pre-installed04:41
cptner, pre-install04:41
predatorfreakoh whatever.04:41
predatorfreakIt's not important enough for me to give a fuck.04:41
Romsterit would be just as bad as my iptabes-imq port.04:41
predatorfreakI packaged aotuv for my own usage04:41
predatorfreakand I've even stopped using the Linux oggenc.04:41
treachRomster: you're missing the point.04:41
Romsteri'll copy aotuv for myself in my repo.04:41
predatorfreakThe windows-only Lancer SSE/aotuv oggenc2 builds are faster than oggenc by miles04:42
cptnpredatorfreak: the alias file is supposed to manipulated, so I can see we shouldn't add entries there04:42
Romsterwhat i' not missing is dont' fuck with anything only do what others do...04:42
treachwell, not really04:42
Romsteri'm*04:42
cptnRomster: well see, CRUX is not about you04:42
cptnRomster: but I think that's the problem04:42
Romsterall i ever get is bitched at04:42
predatorfreakRomster: Back it up now, I'm about to push the nuke.04:42
treachRomster:  it's more that what you're doing is a bad idea.04:42
Romsterk04:42
treachRomster: [11:40] <cptn> but the pre-install breaks since the alias file is not protected in pkgadd.conf [11:41] <cptn> so updating prt-get would remove the aliases added in pre-installed04:43
cruxbot[contrib.git]: libvorbis_autov: Removed due to drama.04:43
predatorfreakThen the dang alias file should go in pkgadd.conf04:43
Romsterit's upto the user to add a line in pkgadd.conf04:43
treachthat is not the same thing as "romster is an idiot who only breaks stuff and does things different"04:43
predatorfreakImagine if someone had a personal repo full of replacement ports.04:43
predatorfreakand decided to use alias to override.04:44
Romsteri get that impression i break shit.04:44
Romsternow i'm going todo other shit because i don't feel like being here right now in my state.04:44
treachhappens, but I don't think that's the problem really.04:44
cptnRomster: again, your commit is silly04:45
treachRomster: the problem I'd say is the way you implement creative solutions without checking the implications with others04:45
predatorfreakFrankly, I think the whole prt-get alias shit is a giant blasted hack.04:45
cptnRomster: instead of fixing the technical problem04:45
cptnyou are pissed04:45
cptnand remove the port04:45
Romstercptn, i could say the same about all them sed lines but i HAVE NOT...04:45
Romsteryues i'm god damn pissed.04:45
cptnwhy didn't you?04:45
predatorfreakand what we REALLY need is a decent freaking full-featured Pkgfile format <<04:45
Romsterevery day it's romster romster don't do this you should do that.04:45
predatorfreakI should just start writing out a new pkgmk and say fuck it.04:45
treachRomster: take a walk, chill and return later.04:46
treachand try not to take everything personal04:46
Romsteryeah i could make a new Pkgfile format but no one will accept it no point.04:46
predatorfreakcptn: The best solution to this whole thing is to have the concept of "provides" in the Pkgfile.04:46
cptnpredatorfreak: yes, which needs meta data in .pkg.tar.gz04:46
predatorfreakSimilar to PKGBUILDs, spec files, portage virtuals, etc.04:46
Romsteri can rattle what i want in a package manager but it's pointless.04:47
predatorfreakcptn: Which we've been horribly slow to actually implement...04:47
cptnpredatorfreak: yeah, but eventually it may happen :-)04:47
predatorfreakcptn: I'll probably be dead by then <<04:47
predatorfreakI mean, I understand development power is kinda stretched atm.04:48
predatorfreakBut, I figure this is one of the better things that could be done rather than complaining when people have to hack around the current system.04:49
predatorfreakIf we had a proper system in the first place, hacks wouldn't be needed.04:49
cptnpredatorfreak: well, I was complaining about Romster's way to packaging04:49
juepredatorfreak: if you want a full-featured package system you might consider using another distro04:49
cptnpredatorfreak: not that one fix04:49
cptnpredatorfreak: it just reflects badly on contrib04:49
predatorfreakjue: I like this distribution just fine, thank you very much.04:49
predatorfreakI don't see how a decent package manager conflicts with CRUX' goals or anything.04:50
predatorfreakI think it'd help A LOT04:50
predatorfreakI'm already in the process of thinking out a new pkgmk which will implement depends, inter-repo depends and other things natively.04:50
predatorfreakBut I'm still planning that and don't want to display a half-baked idea <<04:51
cptnRomster: if you have complaints on sed line, please let me know04:51
treachpredatorfreak: did you get the pdf I started on?04:51
cptn*lines04:51
predatorfreaktreach: ?04:51
treachpredatorfreak: I started out a little paper a while back on the problems with the package management in crux.04:52
predatorfreakLink? XD04:52
treachnever got around to finish it properly though, but I know tilman got a copy, and I think I sent you one too.04:52
predatorfreakI don't remember receiving one..04:53
treachk. I'll have to dig it out of the backups and finish it up some day. :>04:54
treachessentially I was discussing the disparity of the various tools used for package management, and how some of the things prt-get does is to be considered "hackish" IMO04:55
predatorfreaktreach: I'm hoping that once I properly formulate my ideas for a new pkgmk, I can get a working implementation in a week or so.04:56
predatorfreakand have something to display04:57
treachbetter take your time thinking things through.04:57
predatorfreaktreach: I intend too.04:57
treachlots of stuff to consider, and we don't need another ""frog perspective" solution :>04:57
predatorfreaktreach: I intend to merge, to some extent, the concept of repositories into pkgmk.04:59
predatorfreakSo that repository depends can be used instead of the hackist priority setup of prt-get.04:59
predatorfreakI.E. opt/qt4 or kde/qt405:00
predatorfreakor something like that05:00
cptnmmmh05:01
cptnthat may conflict with "provides"05:01
predatorfreakcptn: We have a concept of provides?05:01
predatorfreakand I intend fully to bork prt-get in the process of making this.05:01
*** rxi has joined #crux05:01
treachmmh, and if you use a proper db, like sqlite for instance, you could do away with the mega hackish "cache files" too.05:01
predatorfreakIt's not intended to be a backwards compatible thing.05:01
cptnwell, there's http://crux.nu/Main/PkgutilsIdeas05:01
cptnand http://crux.nu/Main/PkgutilsAttributes05:01
cptnthey describe some of the existing ideas and discussions05:02
predatorfreakcptn: None of which have really gotten off the ground XD05:02
predatorfreakI'm trying to think of a rather KISS style format and setup.05:02
predatorfreakcptn: No offence to your code either, but it's still rather primitive :D05:03
cptnyou mean the attr stuff?05:04
treachprt-get was created to help out with stuff pkgmk didn't solve, and it suffers from it IMO.05:04
predatorfreakcptn: Yes, sadly.05:04
cptntreach: a number of things are implemented in the long layer, yeah05:04
predatorfreakthe Attribute stuff is right in the direction I want to head, though05:05
cptnyet, to come to the starting point of the discussion, the tools won't help if they're not used properly05:06
predatorfreakcptn: Do you happen to have any time to work with me on extending the attributes stuff?05:07
treachguys, if you promise not to cut me open I could give you my little pdf.. it's not done by far, really. but there are quite a few thoughts in it.05:07
cptnI hope so :-)05:07
predatorfreaktreach: Gimme :P05:07
predatorfreakI make no promises though :D05:08
treachwait, I'll have to find the backup, it's somewhere around here.. :p05:08
predatorfreakOOHHH IT'S ALL IN SHELL!05:09
predatorfreakWoohoo!05:09
cptnthat was really just a prototype05:09
predatorfreakcptn: I don't have to try and learn a ton of C/C++ while hacking at it :305:09
cptnto get a feeling05:09
tilmanw t f\05:09
predatorfreakYeah, but that's all I need to test my ideas05:09
tilmanYou're getting this mass mail because you had a vulnerable SSH key[0],05:09
tilmanor two, or six, on freedesktop.org.05:09
tilmano_O05:10
predatorfreaktilman: Didn't generate it on Debian did you? :)05:10
predatorfreakcptn: In some ways, I actually like the pkgadd.conf stuff.05:12
predatorfreakProperly formalised, it could be a way to quickly add hooks into pkgmk.05:12
predatorfreakTo facilitate quickly adding things like sha256 or something like that05:12
predatorfreak(Just quickly looking over the code here)05:13
treachpredatorfreak: did you get the link?05:14
predatorfreaktreach: Yes, reading now05:14
treachok.05:15
predatorfreaktreach: Also, you are not infinitely wise, only Per was infinitely wise :P05:16
treachI know. but I usually am smart enough to realize when I'm too stupid. :>05:17
tilmanhttp://www.fooishbar.org/blog/tech/fdo/dsaKeys-2008-05-17-03-16.html05:17
cptntilman: wow05:18
predatorfreaktilman: Blame Debian ;)05:18
nipuLi have a modular base alternative to pkgmk i've been playing with, not sure how much use it is to the current discussion05:21
nipuLhttp://git.die.net.au/cgit/misc/ppkgmk/tree/ i haven't touched it in a while though05:22
predatorfreaktreach: Okay, my only comment is: Let's completely hammer out attributes and during a C implementation, add sqlite :)05:23
predatorfreakBecause, yes, having an actual database for a backend makes a lot of sense05:23
treachI'm a bit curious about cptn's view.  ;-)05:23
cptnheh05:23
nipuLi concur05:23
treachcptn: I realize one could read that as a bash of prt-get, but it's not meant that way.05:24
cptnI'll read it in a minute05:25
treachok. :)05:25
predatorfreakGoing to shower, then I'm going to comprehend the sparingly commented pkgutils-attr05:26
predatorfreakand hopefully start hacking on it later.05:26
cptntreach: I remember the ports tree as a DB approach05:38
cptnI think I like the fact that it's files, since I can browse in there05:39
cptnbut that's probably just because I'm used to it :-)05:39
joacimI'm thinking plain old files is easier to manage too05:40
cptnone problem is that you don't have to maintain storage code anymore05:41
cptnhowever you'll have to write new network -> DB sync code05:41
cptnto make ports -u efficient05:41
*** mwansa has joined #crux05:42
treachyeah, I know. It's listed in the "problems" section, if not in those exact words. :-)05:43
treachnot sure why not having to maintain storage code is a problem though..?05:45
cptnerr, I put that wrongly05:46
cptnI understood that as advantage05:46
cptnbut you trade it for the effort of maintaining the network code05:46
cptnwhich is currently done by the rsync folks05:46
treachah. ok.05:46
tilmanso, if any of you are using DSA keys with crux.nu, please replace them05:47
predatorfreakcptn: Well I can see just how primitive this prototype is :D05:49
predatorfreakI'm going to try to work in a concept of repositories either later today or tomorrow as well as preliminary dependency checking05:50
cptnhey, it was just to show that it's not just a theory :-)05:50
predatorfreakand build a mock ports tree based off a modified pkgutils-attr within a week or so.05:50
predatorfreakcptn: After I have my ideaers prototyped, a semi-useful implementation to prove my ideaers and such, I'll publish it and announce it to crux-devel.05:51
cptncool05:52
*** discomaxus has joined #crux05:56
*** maxus has joined #crux06:02
predatorfreakcptn: A really quick and braindead idea for avoiding duplicate attributes, just looking over your todo list, would be to run uniq over the output.06:09
predatorfreakSo that only unique lines are written.06:09
predatorfreakcptn: Also, presumably, the only way duplicate attributes could get added is a bad pkgmk.06:12
*** mxs has quit IRC06:12
*** mxs has joined #crux06:12
cptnheh, I don't actually remember the code06:13
cptnuniq sounds fine of course06:13
cptnpredatorfreak: I think it was about pkgadd.conf06:16
cptni.e. that you don't register two hooks for an attribute06:16
*** discomaxus has quit IRC06:22
*** maxus has quit IRC06:23
*** mwansa has quit IRC06:42
*** onestep has quit IRC06:49
*** Rotwan1 has joined #crux06:53
predatorfreakcptn: Back and I see what you mean06:56
predatorfreakIt shouldn't run the same hook twice, in my opinion.06:56
cptnjust keep in mind that the code was not meant to be used for the real implementation06:56
cptnI only made it to show the idea06:57
predatorfreakcptn: Oh trust me, after I finish hacking this06:57
predatorfreakIt'll basically be a rewrite06:57
cptnand I'm not sure if that TODO's up to date :-)06:57
predatorfreakcptn: I intend to produce a prototype that could be immediately used, but obviously would be better as a reference for a C or somesuch implementation06:58
cptncould be06:59
cptnhowever, since building packages takes usually quite some time, using a bash script might not be that much of a speed overhead07:00
cptnI think Han has sh-only pkgutils, right?07:00
predatorfreakYes.07:02
predatorfreakOnly it's really really old07:02
predatorfreakcptn: oh and hey, Gentoo implement portage in Python07:02
predatorfreakSo we can do shell, I suppose :P07:02
cptnor ruby :-)07:02
predatorfreakcptn: Too lazy! :D07:02
cptnI think it's good if the prototypes are easy to hack07:03
cptnat this stage, I think ideas and goals differ quite a bit among devs and users07:03
cptnpreviously, some attempts died because we couldn't agree07:03
cptnI think that pkgutils-attr prototype made jdolan quite angry for example :-)07:04
predatorfreakcptn: I think it's about time we got serious about this stuff.07:05
predatorfreakI'm going to even stop giving a damn about SHA256 for 2.5 and push for a proper, working pkgutils-attr for 2.507:06
cptnI agree about the getting serious stuff, but then I think a good start would be to define goals07:07
nipuLmultiarch07:07
cptne.g. I don't think the repo deps are worth implementing07:07
cptnwhereas I might want stuff you're not really interested in07:07
cptnand so on07:07
predatorfreakcptn: What about things like GNOME and KDE?07:08
predatorfreakWhich override opt ports and such.07:09
predatorfreakI think that repo deps would be very useful for cases like that07:09
*** Rotwang has quit IRC07:09
jueI agree with cptn, before we start implementing something we should have agreed about the features we want/don't want07:09
nipuLi just fix that one by ordering prtdirs in prt-get.conf07:09
predatorfreaknipuL: I view that as hacky :D07:09
cptnI always considered that rather elegant07:10
nipuLme too07:10
nipuLvery kiss07:10
predatorfreakI don't, repo deps could be done rather easily and eliminate the need.07:10
nipuLwill it also eliminate dep hell?07:11
predatorfreakBesides, in order to properly do dependency verification in pkgmk, it has to have a concept of repositories anyway07:11
predatorfreaknipuL: For GNOME and stuff? The only way to eliminate that is for them to have sane depend lists..07:11
cptnthat's another point07:11
cptnwhether dep handling has to do in pkgmk07:11
nipuLi also like that pkgmk isn't aware of deps or repos07:11
nipuLdo one thing well07:12
predatorfreakcptn: Why in Gods name should we waste time implementing attributes if we only use them in a glorified prt-get update?07:12
predatorfreakWe might as well use what we have now...07:12
juepredatorfreak: I'd suggest you write down your ideas somewere so we can discuss them07:12
predatorfreakThe idea in using attributes should be to remove the need for prt-get and move the functionality into pkgutils07:13
cptnpredatorfreak: check the pkgutils ideas07:13
cptnattributes is really just a format thing07:13
cptnbut is simplifies lots of things, including meta-data07:13
predatorfreakGreat, I'm already in a whole different ballpark than what you're aiming for.07:13
nipuLi saw is as more of a generic way of giving packages extra information07:13
cptnI'd like to keep the layering as it is07:13
nipuLnot just for prt-get07:13
predatorfreakI don't want to keep separated tools.07:13
predatorfreakI want to merge functionality into a proper package manager.07:14
cptnI think prt-get right now implements to much in its code base07:14
cptnwhich should move to a lower layer07:14
predatorfreakSo we install pkgutils/ports and get a full set of utilities.07:14
cptnbut that doesn't mean that it should be one with pkgmk07:14
nipuLa lot of the information gathering should move to pkgutils07:14
predatorfreakIncluding a wrapper around pkgmk to basically find a port in a repo and build it.07:15
predatorfreakWhere as pkgmk resolves depends and factors them in, etc.07:15
predatorfreakThat way people like me, who only use prt-get for system updates, actually get a tool that knows what depends matter07:15
*** Rotwan1 has quit IRC07:16
treachpredatorfreak: you could do what I did -- well, you should probably complete it -- and then post it somewhere. ;-)07:16
*** onestep has joined #crux07:16
predatorfreaktreach: Make a PDF?07:16
cptnpredatorfreak: you know, keep in mind that we've actually discussed this twice in a real life meeting, and countless times on IRC07:16
treachlengthy discussions on irc are kind of hard to follow and keep track of what was said etc.07:16
predatorfreakBy what I'm seeing, I'd have a better time just forking and keeping the same format07:16
cptnpredatorfreak: I think there are some ideas how it should be07:17
predatorfreakand mine are conflicting wildly.07:17
cptnprobably07:17
cptnbecause opinions differ :-)07:17
nipuLwe've lost a few people over the years because of this07:17
cptnor "taste" in Per lingo07:17
predatorfreakI do not want to implement a glorified portage07:17
treachpredatorfreak: but you're not in core, so you lose by default. ;)07:17
nipuLhan comes to mind07:17
predatorfreakand have a useless independent building tool07:17
predatorfreakLike ebuild.07:17
predatorfreakpkgmk is currently only useful for one package at a time and I want to extend it to factor in other packages.07:18
predatorfreaktreach: That's my problem <<07:19
cptndepends on what one's goals are I guess07:19
cptnback in the days, it was quite clear that prt-get wouldn't make it on the ISO07:19
cptnbut it was still handy to me, and for others too07:20
*** Rotwang has joined #crux07:20
nipuLbuild it and they will come07:20
predatorfreakcptn: I find it handy, sure.07:20
predatorfreakFor updating the system.07:20
treachcptn: but crux isn't quite the same today as it was back then.07:20
predatorfreakEverything else I want a smart pkgmk.07:20
predatorfreakand currently prt-get is a tad braindead with sysup.07:20
cptntreach: no, but if predatorfreak is working on his own pkgutils and it doesn't make it into core, that doesn't mean that he _lost_07:20
predatorfreakI.E. not recalculating dependencies and shit.07:21
treachcptn: ah, that was just a joke. :)07:21
predatorfreakcptn: It means I wind up like Han.07:21
nipuLjust write it, if it's good people will use it07:21
predatorfreakWith my reimplementation and the official version07:21
cptnpredatorfreak: what's wrong with that?07:21
treachcptn: that said, I *really* think the entire package management needs an overhaul.07:21
predatorfreakcptn: It defeats the blasted point.07:22
cptnI mean, no one broke Pkgfiles to make Han's pkgutils break or things like this :-)07:22
cptnso he should have been happy07:22
predatorfreakAt that stage, I really might as well fork everything else.07:22
predatorfreakSay fuck it and steal your trees07:22
predatorfreakthen rebrand it <<07:22
treachcptn: for instance there are a lot of things in the prt-utils scripts that really belongs in the package management, and not in some random user contributed script07:22
nipuLNETWOSIX!!07:22
predatorfreakSo that I have total control over it.07:22
cptnpredatorfreak: and depending on how you do that, that could be fruitful for both sides07:23
nipuLyou already do have total control07:23
predatorfreaktreach: I also want to implement a uninstall-unused-depends style option on uninstall.07:23
treachok07:23
predatorfreaknipuL: Not really.. I've been trying to push my SHA256 stuff upstream for awhile.07:23
predatorfreakand I haven't gotten anywhere with that.07:23
predatorfreakI can code it, but getting stuff merged is annoying.07:24
predatorfreakFrankly, I love CRUX, but am annoyed when working on things not-related-to-maintaining-ports.07:24
nipuLyou can still use sha256, git is good at that stuff07:26
predatorfreakAnyway, enough ranting.07:26
predatorfreaknipuL: Sure, I can, but without upstream acceptance.07:26
predatorfreakThe point is moot.07:26
predatorfreakMe and Romster both use SHA256 but since it's not upstream it's purely in our personal repos.07:27
nipuLi don't think there even is a point07:27
nipuLi've been using multiarch pkgutils and repos for amd64 quite happily for about a year now, i could care less if they get merged07:27
predatorfreaknipuL: The point for me is why should I bother to implement this stuff if I'm the only one who uses it?07:29
predatorfreakSHA256 is nice and all.07:29
predatorfreakBut if I'm the only one using it.07:29
predatorfreakThen really it's worth nothing.07:29
nipuLclearly you do not understand the tao of programming07:29
predatorfreakPersonal tools are fine when used like that.07:29
predatorfreakBut my point isn't to fork pkgutils it's to actually produce good improvements <<07:30
nipuLi like personal tools07:30
nipuLgreat tech support, friendly userbase07:30
cptnpredatorfreak: improvements are sometimes just subjective07:31
cptnsee the Romsterisms07:31
predatorfreakcptn: Yeah, I've gotten that.07:31
treachfrom a user perspective, I think details like using sha256 or md5 is pretty moot. but as I said before, the tools themselves would need an overhaul with regard to consistency and functionality.07:31
predatorfreakand frankly at times I have the strongest urge to fork CRUX, but I'm too nice to do that.07:31
cptnas long as it's a friendly fork...07:32
treachIMO the fact that these various tools were never designed to work in a coherent way from the start causes a lot of problems.07:32
cptnI always felt that the ability to fork was one of the strength of open source07:32
treach*work together*07:32
nipuLperhaps you could spoon instead07:32
Romstercptn> see the Romsterisms < yeah bloody figures...07:33
predatorfreakcptn: I don't want to fork just because my ideas aren't liked.07:33
Romsteri may just fork haven't decided yet.07:33
predatorfreakI guess I can just make the CRUX variant of Portage V Paludis for Gentoo07:33
predatorfreakpkgutils v pkgutils-ng :D07:34
Romsteri'm giving crux one last go with the suggestions i got from jue.07:34
nipuLpkgutils v pacman07:34
cptnpredatorfreak: it worked on the upper layers at least, see prt-get vs ilenia07:34
predatorfreakilenia is deadz07:34
predatorfreakSince like.. forever :d07:34
predatorfreak:D*07:34
cptnI thought PPC only ships ilenia?07:34
predatorfreakIs CRUX PPC still alive?07:34
cptnreplacing both ports and prt-get07:35
nipuL#crux-ppc is pretty dead07:35
nipuLempty in fact07:35
predatorfreakcptn: Don't take this the wrong way, but ilenia looks better than ports/prt-get :D07:36
cptnagain, a matter of opinions I guess07:36
nipuLopinions are like arseholes07:36
nipuLeveryone has one, but they all stink07:36
nipuLexcept mine of course07:37
nipuLwhich smells like roses07:37
predatorfreakcptn: Well, it seems to me that merging the tools into one is a good idea07:37
cptnI understood that07:37
cptnI think that decoupling the network sync from package building is good07:38
predatorfreakcptn: Basically that's how it is..07:38
predatorfreakilenia is just a glorified frontend.07:38
predatorfreakIt doesn't replace pkgmk.07:38
nipuLdecoupling network sync? does that mean have a seperate tool to gather sources?07:38
cptnports -u07:39
cptnto sync the ports tree07:39
cptnnot having ilenia do that07:39
cptnanyway, my main difference with the concept of ilenia is that "highest version wins" by default07:40
cptnwith an way to add preferred repositories07:40
rxican someone give me the vim command to find and replace /etc/ports with /usr/ports/_files07:40
cptnand it does more things automatically, which may be what people want nowadays07:40
cptnpredatorfreak: for example the "braindead" sysup no adding new dependencies, that's because people wanted to actively ignore dependencies07:41
predatorfreakcptn: Make a switch for it..07:41
predatorfreak--no-deps07:41
predatorfreakor equivalent07:41
cptnpredatorfreak: we wouldn't _force_ them to install them, so it would be braindead to pull them in during sysup07:41
predatorfreakcptn: See above..07:41
cptnI'd rather store the dependencies in pkgdb07:41
cptnso you know whether a dep is new, or was ignored07:41
predatorfreakThat's what every other sane package manager does.07:41
predatorfreakThey follow the deps and offer the ability to ignore them.07:42
predatorfreakSeems like the "smart" thing to do.07:42
cptnI thin that's also in the TODO list in git07:42
treachrxi:  %s/\/etc\/ports/\/usr\/ports\/_files07:42
treach:p07:42
cptnI'm just saying there's a historical reason why that was right at some point in time07:42
rxitreach: thanks mate :)07:43
treachnp07:43
rxii always get my / and \ mixed up07:43
cptnrxi: why not use sed for that? :-)07:43
nipuLrar!! my kingdom for a blank cd07:44
rxicptn: id still have the same problem with my dislexia :P07:45
cptnwell, sed supports different separator chars07:45
rxiahh07:45
cptnso you can do 's|/etc/ports|/usr/ports/_files|g'07:45
cptnno need to escape the slashes07:46
rxicool .. i remember that next time ports updates :)07:46
rxiRomster: what happened to your repo?07:49
Romsterhttp://romster.dyndns.org:8080/linux/ports/crux/07:53
Romstermoved07:53
Romstersince prologic is gone.07:53
Romsterand i've emailed viper to get it altered on the portdb and nothing yet...07:53
rxiahh07:53
Romsterwho must i email to get the portdb updated?07:53
rxijust reregister it?07:54
Romstergotta send a email to do that too.07:54
Romstercan't self register it how the web site is.07:54
cptnwell, the website says contrib-admin@crux.nu, but that goes to viper too07:55
Romsteroh last time i emailed viper directly and got it altered.07:55
Romsterand talked to him on irc.07:55
Romsterlast time he droped in here i told him to please check your emails.. and why.07:56
Romsterbut still no change.07:56
*** onestep has quit IRC07:59
cptnRomster: this is not intended as a flame bait :-)08:00
cptnbut regarding the post-install of gcj et al...08:01
Romstersee what i mean about asking and emailing and ML and nothing happening.08:01
cptndo they really need to be so complicate?08:01
Romstereh i guess it's over the top i wanted to be sure i caught all errors.08:02
*** mwansa has joined #crux08:02
Romsterhave you even looked over my other scripts.08:02
cptnno08:02
Romsteror just pick on the only ones you see recently :)08:02
Romsterthought they were simple enough and not over complex and easy to read. and the end user doesn't need to read them they also are informative.08:05
Romsterif you think you can do it in a better way give me a example of how i should be doing them cptn.08:06
Romsterand i may change them.08:07
cptnokay, will do08:07
Romsterit's what i thought up and tested and worked.08:08
Romsterand gave error signals where needed so pt-get can say if it failed or passed.08:08
Romsterand gave the user installing it feedback so they know what's happend.08:08
Romsteri do edit them from time to time if i see a improvement.08:09
Romsteri thought they were neat as they are now.08:09
RotwangRomster: why you use sh not bash?08:10
Romstersomeone might not have bash installed but some other shell.08:14
Romsterso it's wise to not use specific shell features and be compatible with the standard.08:15
cptnhow realistic is that, if pkgmk depends on bash?08:15
cptnif you write a post-install for a CRUX port?08:15
Rotwangtrue08:16
Romsterah but that may change.08:16
Romsterever thought of that too.08:16
cptnor we could only have bash, but no sh08:17
cptnever thought of that?08:17
predatorfreakcptn: Any distro that removes the /bin/sh symlink is full of idiots.08:18
predatorfreakSeriously.08:19
Rotwangheh ;]08:19
ryuocptn: what ya been smokin dude? /bin/sh is critical.08:26
cptnryuo: just an example of something that's unlikely to happen08:26
cptnanytime soon08:27
predatorfreakcptn: If it works without needing bash features.08:27
predatorfreakIt's really stylistic nitpicking08:27
predatorfreakcptn: Hell, I could nitpick your Pkgfile.attr stuff because you don't properly quote the depends.08:28
*** Falcon|_ has joined #crux08:28
predatorfreakEach item should be wrapped in single-quotes.08:28
cptn?08:29
predatorfreakcptn: depends=(blah blah blah)08:30
predatorfreakTechnically speaking that gets split at the space.08:30
predatorfreakBut you should do depends=('blah' 'blah' 'blah')08:31
ryuosince when did crux support depends in anything but # Depends?08:31
ryuothat sounds like archs method08:31
predatorfreakryuo: I'm talking about his pkgutils-attr stuff08:31
ryuooh.08:31
cptnpredatorfreak: what does that have to do with not using feature since they might go away?08:32
predatorfreakcptn: Since when was basic bourne compatibility going AWAY?08:33
predatorfreakBash is an EXTENTION of bourne shell.08:33
predatorfreakSeriously.08:33
cptnso your point is?08:33
cptnyou should not use bash because of that?08:34
predatorfreakcptn: <predatorfreak> cptn: If it works without needing bash features. <predatorfreak> It's really stylistic nitpicking08:34
predatorfreak<<08:34
predatorfreakYou're not getting it.08:34
cptnno08:34
cptnthe argument was that you should use sh instead of bash08:34
predatorfreakIt's stylistic nitpicking unless it actually breaks shit.08:34
cptnsince bash could go away08:34
cptnnot sure how your arguments fit in there08:34
predatorfreakcptn: No Linux distro is dumb enough to not ship bash.08:34
cptnah08:35
ryuobut bash is a staple on any linux system08:35
cptnnow we're getting somewhere08:35
predatorfreakNor is any Linux distro dumb enough to not ship /bin/sh.08:35
cptnagreed08:35
predatorfreakcptn: My point is.08:35
predatorfreakSince BOTH are a given08:35
predatorfreakYou're nitpicking on coding style.08:35
ryuoi usually use /bin/sh instead of bash though for my shell scripts08:35
ryuonever really found a difference between the 2...08:35
predatorfreakIt's really the authors preference as to whether or not to use bash-specific features.08:35
cptnpredatorfreak: agreed08:35
cptnpredatorfreak: but then make that clear08:35
predatorfreakPersonally, I do, I find arrays to be very useful08:35
nipuLit makes a difference on non linux systems08:36
predatorfreakcptn: I thought I did.08:36
cptnpredatorfreak: but when someone on this channel says one should use sh over bash08:36
predatorfreaknipuL: When I'm writing portability minded shell scripts, yes.08:36
ryuonipuL: if your shell script only works on linux kernel functions, what difference does it make?08:36
cptnbecause it should go away08:36
nipuLtoo many #!/bin/sh's that use bashisms08:36
predatorfreakcptn: My remarks didn't imply that I believed that.08:36
cptnbut Romster's08:36
ryuoi've written shell scripts for changing cpu governors before for example ;)08:36
predatorfreakcptn: My remarks were just comments on your argument shit with Romster.08:37
cptnokay, I guess I failed to make my point before08:37
cptnall I meant is that "because XYZ might happen" is not a strong argument08:38
cptnsince ABC might happen which made your efforts pointless08:38
predatorfreakcptn: and my point was.08:38
ryuocptn: so what? if it does, just update it and fix it later.08:38
predatorfreakYour original shit about him using /bin/sh shit was nitpicking stylistically.08:38
predatorfreakThere was no real point to bringing it up in the first place if the code works.08:38
* cptn scrolls back08:39
*** Falcon| has quit IRC08:39
cptn" so it's wise to not use specific shell features and be08:39
cptn                  compatible with the standard."08:39
predatorfreakoh wait08:39
predatorfreakThat was Rotwang.08:39
cptnthat's just not generally valid08:39
predatorfreakSorry.08:39
predatorfreakGetting tired.08:39
cptnit makes sense to be standard compliant08:40
cptnand whereever possible, I don't rely on bashisms08:40
predatorfreakcptn: I'm faulting you for what Rotwang said.08:40
cptnbut using that as a general argument to not use bash in crux stuff...08:40
predatorfreak<Rotwang> Romster: why you use sh not bash?08:40
cptnpredatorfreak: okay08:40
predatorfreakThat's what I was remarking about08:40
* ryuo grabs some popcorn and watches the show of sliding text.08:40
predatorfreakryuo: We should just stage cage fights in #crux <<08:41
ryuodont you mean steel cage match? ;)08:41
predatorfreakRomster v tilman, me v cptn and jue at the same time :P08:41
predatorfreakryuo: Are you a native English speaker?08:41
nipuLand i'll bring a bag of spark plugs to throw at people i don't like08:41
ryuopredatorfreak: thats what they call it in Texas <_<08:41
nipuLLIKE YOU!08:41
Rotwangpredatorfreak: it was only a question08:42
predatorfreakryuo: stage a cage fight basically means that.08:42
ryuoand isn't "<<" Romster's signature move?08:42
Rotwangdont be so tense you people08:42
predatorfreakryuo: Yes.08:42
predatorfreakI stole it.08:42
ryuopredatorfreak: thief :-p08:42
Rotwanggo out and chill08:42
predatorfreakryuo: I also stole \o/08:42
ryuowhy are you people even arguing over something that may or may not matter to crux?08:42
predatorfreakand \o08:43
predatorfreakryuo: Tradition.08:43
predatorfreakWhen there's nothing better to do.08:43
ryuoso what if crux's stuff is "portable" to other distros? its meant to be used with this distro, so...08:43
predatorfreakFling poo.08:43
ryuoi dont think it matters one way or the other :-p08:43
Romsteri chose to use sh over bash, maybe one day i'll use bash over sh if i depend on bash features, it's my god damn choice now go watch a movie or something.08:43
cptnRomster: http://paste.lisp.org/display/6089708:44
cptnuntested, just posting it since you're here08:45
cptnerr, wait08:45
DarkNekrosthere are 3 fi :S08:45
Romsterwhat the no { } brackets either.08:46
cptnand the set -e is missing08:46
predatorfreakcptn: Double check your code! :D08:46
cptnhttp://paste.lisp.org/display/6089808:46
nipuLhehe08:46
nipuLgit log hal08:47
Romsterno quiting of vars is bad too.08:47
predatorfreaknipuL: ... Fuck you man!08:47
Romsterquoting*08:47
predatorfreak:D08:47
Rotwang#!/bin/sh -e08:47
DarkNekroscptn, what's the difference betwen test and [ ]?08:47
predatorfreakcptn: Also, thinking about it, since CRUX is a source based distro basically.08:48
cptnDarkNekros: it's shorter?08:48
RotwangDarkNekros: same thing08:48
predatorfreakWhy don't we integrate pre-install/post-install into the Pkgfile?08:48
Romstercptn, for rights that shit should be standard in pkgutils not hacked in a post-install.08:48
predatorfreakI.E. pre-install function, post-install function.08:48
DarkNekrosI thougth it was about the interpreter :P08:48
predatorfreakSo that when calling build.08:48
predatorfreakYou do pre-install08:48
predatorfreakbuild08:48
predatorfreakpost-install08:48
DarkNekrosthanks cptn & Rotwang ;)08:49
nipuLwhy not more stages08:49
cptnhere's the context BTW:08:49
cptnhttp://crux.nu/gitweb/?p=ports/contrib.git;a=blob;f=gcj/post-install;h=071290f20fc40fe35f8e3e53a1bf7d63af7000da;hb=a7d731bd9aea869ba418cbd8ae3fbfbf3565ca9608:49
predatorfreakObviously, post-remove needs to be handled by the fancy thingiemabob.08:49
cptn^ that paste is a replacemen for it08:49
predatorfreakBut that's about it.08:49
predatorfreak(plus pre-install/post-install should get renamed to pre-build, post-build)08:49
Romsterand user and group should be handled in the Pkgfile and tracked.08:50
nipuLthat's why pkgadd hooks would be nice08:50
predatorfreakYes.08:51
*** ryuo has left #crux08:51
predatorfreakIn fact.08:51
*** ryuo has joined #crux08:51
nipuLi'd rather see that than adding it to pkgmk08:51
* predatorfreak adds to ideas for pkgutils-ng08:51
Rotwangand there should be useflags!08:51
predatorfreakRotwang: No.08:51
* Rotwang runs08:51
Romstereww08:51
nipuLlet's switch to a make based system08:51
Romstergo use gentoo if you want useflags08:51
Rotwangive had an idea for cruxish useflags08:51
* nipuL runs the other way08:51
Rotwangbut no one likes it :{08:52
predatorfreakRotwang: Bad boy.08:52
nipuLRotwang: i think lot's of people have tried and failed08:52
* Romster looks over cptn's code.08:52
nipuLmyself included08:52
cptnRomster: I've added a warning for the case where the symlinks exist08:54
cptnsince they might be pointing to the wrong binary08:54
Romstercptn, you didn't even check if ccache or distcc is installed.. yeash.08:54
cptni.e. if the user made them himselv08:54
Romsteri have all that error checking now.08:54
cptnRomster: no, because if the binary is there, and the masquerading dir is there I don't actually care all that much if it's installed from ports08:55
Romsterah you check for the existance of the directory..08:55
cptn_and_ the binary08:55
Romsterwhat about removing cleanly nothing allows for that.08:55
cptnindeed08:56
cptnthere's another approach08:56
cptngcj-distcc-bindings08:56
cptnas a port08:56
Romsterhmmz08:56
cptnit's just an overhead, but cleanly removable08:56
cptnand will also update properly if the binaries move08:56
Romsterthat seems like a better way and rankly i liked them how the Pkgfiles used to be...08:56
Romsteri couldn't see the problems with it.08:57
Romsters/rankly/frankly08:57
Romsteri'll make a metaport for it and install symlinks i like that idea better08:58
Romsterif no one objects SPEAK UP NOW!08:58
cptnmetaport?08:58
Romsteror forever hold your peace.08:58
cptnI'd just add an entry in the gcj readme08:59
cptnsince it only affects those that have distcc/ccache08:59
Romstercreate symlinks in gcj-distcc-bindings ports and put the depends on distcc and gcj08:59
cptnah, yeah08:59
*** Falcon|_ is now known as Falcon|08:59
Romsteraka metaport.08:59
Romsterbut with some symlinks.08:59
cptnsounds good to me09:00
Romsterfinally something i like.09:00
Romsterand looks less hackish too.09:00
Romsteryet more work for me todo..09:00
Romsterhope someone appreciates my work.09:00
* Rotwang appreciates09:02
RotwangRomster: with all your ports and ideas you could throw a new ditro :P09:03
Rotwangdistro*09:03
Romstermight happen.09:04
ryuoi can see it now... Romster OS09:04
ryuo:-p09:04
Romsterhah....09:04
ryuoman i remember the crap that happened on IRC when i was younger XD09:13
ryuopeople threatened each other with nukes XD09:14
ryuosome proggie that tried to crash the targets computer09:14
rehabdollme and a couple of classmates used winnuke on _EVERY_ connected computer on our network09:20
rehabdollwe lost network access to our lab after that :D09:21
*** Viper_ has joined #crux09:21
Rotwanghttp://www.swollenpickles.com/wp-content/uploads/ramsey1.jpg09:47
*** predatorfreak has quit IRC09:49
cruxbot[opt.git]: madwifi: forgot release bump for solving FS#27209:56
*** jdolan has joined #crux10:01
*** ChanServ sets mode: +o jdolan10:01
rehabdollRotwang: :D10:25
*** mxs has quit IRC11:06
*** blueCommand has joined #crux11:07
cruxbot[opt.git]: rox: updated to 2.811:11
*** jdolan has quit IRC11:14
*** blueCommand has left #crux11:14
*** jaeger has joined #crux11:15
*** ChanServ sets mode: +o jaeger11:15
jueRotwang: I have a simplification for your cream port, http://sh.nu/p/2442311:54
*** f1y has quit IRC11:59
*** Viper_ has quit IRC12:01
Rotwangjue: have you tested it?12:04
Rotwangah i see it i they added DESTDIR variable since i didnt noticed it ;]12:08
Rotwangjue: thanks12:08
juenp12:08
Rotwangbut still script interferes outside $PKG x|12:13
jueyou mean the call to vim --cmd 'echo $VIMRUNTIME' --cmd 'quit ?12:18
Rotwangi mean Cleaning up previous versions...12:19
Rotwangheh and this one alsa12:20
Rotwangalso12:20
Rotwangit writes to /tmp and that shouldnt happen, no?12:21
*** Viper_ has joined #crux12:25
juegenerally why not, but the way cream does it is a security hole12:28
Rotwangill simply install everything inside Pkgfile omitting script, of course i might assume that everyone has vim installed in /usr/share/vim !?!12:30
*** namenlos has joined #crux12:30
juesure, that's where core/vim installs it12:32
jueRotwang: don't get me wrong, I don't use cream, just saw the big patch for a vim extension and looked shortly over install.sh. Just leave it as it is.12:37
Rotwangok ok ;]12:39
cruxbot[contrib.git]: cream: removed ugly patch12:46
*** namenlos has quit IRC12:51
*** jue has quit IRC13:03
*** Viper_ has quit IRC13:34
*** sepen has joined #crux13:41
sepenhey13:42
*** phi- has joined #crux13:44
*** phi- has left #crux13:44
cruxbot[contrib.git]: wbar: new port13:49
*** mrks has joined #crux13:51
*** mrks_ has quit IRC14:09
cptnRotwang: wbar looks cool14:12
Rotwangwbar + awesome = :D14:13
cptnis that the dwm clone?14:13
cptnor wmii?14:13
tilmanit's yet another tiling wm14:15
cptnah yeah14:16
cptnnever got the hang of these14:16
tilmansame, openbox ftw14:18
treachcptn: based on dwm originally iirc14:18
Rotwangyeah, actively developed14:19
treachforked for (*shock*) having a config file. ;-)14:21
*** ryuo has quit IRC14:23
tilmani'm writing some haskell code14:23
tilmanand i'm feeling infinitely dumb14:23
treachcptn: speaking from my own experience, dwm was really nice as long as you didn't have to use applications like the gimp, dia or pidgin. :)14:26
cptnheh14:28
cptnwell, whenever I see screenshots they're (obviously) always very nice14:29
cptnbut I kinda got used to the free placing of windows...14:29
cptnbut the concept makes sense14:29
Rotwangi use mostly floating layout (not tiling)14:30
cptnah14:30
treachyeah. My favorite part of the concept is that you don't have to move your hands off the keyboard just to move a window out of the way14:30
treachin fact once one has gotten used to it, one realizes really how much that entiere process slows down your brain.14:31
treachtype, damn I need to see that other window, move hand to mouse, use mouse to move window, move hand back and continue....14:32
treachnot really good from a cognitive perspective..14:32
Rotwangin case your mose is broken you still are able to use X :D14:34
treachif all you use is terminals and vim/emacs I'd say dwm is a clear winner14:34
jesse_gimp/pidgin/dia work just fine if you configure them to float at all times in dwm.14:35
treachthe concept begins to fall apart when one starts using multi window applications or stuff that likes to live in the systray14:35
treachjesse_: not really imo. I guess you could say it works, but it's not  very nice IMO.14:36
jesse_Yeah, in principle they do not fit well with a tiling wm.14:36
jesse_It's back to the typical mouse usage. Something of a hack due to their design.14:39
*** DarkNekros has quit IRC14:41
*** DarkNekros has joined #crux14:43
cruxbot[contrib.git]: doxygen: Cleanup and updated 1.5.5 -> 1.5.614:53
*** j^2 has quit IRC14:57
*** boe has joined #crux15:26
*** pitillo has joined #crux15:38
*** Rotwang has quit IRC16:35
*** pitillo has quit IRC16:37
*** maxus has joined #crux16:40
*** boe has quit IRC17:29
*** mwansa has quit IRC17:37
DarkNekroshttp://www.freesoftwaremagazine.com/columns/bizarre_cathedral_517:52
*** DarkNekros has quit IRC18:16
*** predatorfreak has joined #crux18:17
*** joacim has quit IRC19:43
*** Dudde has quit IRC21:13
*** mavrick61 has quit IRC21:13
*** mavrick61 has joined #crux21:14
*** Dudde has joined #crux21:15
*** mernil has joined #crux21:37
mernilhi all! :-)21:37
*** mernil has quit IRC21:37
nipuL12:24 <@tilman> and i'm feeling infinitely dumb22:00
nipuLthat makes me feel less retarded about my own haskell experiences22:00
jaegerI took haskell in college22:01
jaegeryuck22:01
nipuLi took scheme and prolog22:02
nipuLah i love screwing with telemarketers22:04
nipuLwanted to give me a free nokia mobile22:04
nipuL"nokias, suck, i want a free sony ericsson"22:05
jaegerheh22:05
nipuL"actually i've already got one, it's awesome, how about i give it to you, all you have to do is stop call me"22:06
^racer^you can send it to me, and I will promise not to call you22:32
*** morlenxus has quit IRC23:05
*** morlenxus has joined #crux23:05
*** jaeger has quit IRC23:13
cruxbot[contrib.git]: gcj: remove install-scripts23:55
cruxbot[contrib.git]: gdc: remove install-scripts23:55
cruxbot[contrib.git]: gcc34: remove install-scripts23:55
cruxbot[contrib.git]: gcc-fortran: remove install-scripts23:55
cruxbot[contrib.git]: gcj-distcc-bindings: initial commit23:55
cruxbot[contrib.git]: gcj-ccache-bindings: initial commit23:55
cruxbot[contrib.git]: gdc-distcc-bindings: initial commit23:55
cruxbot[contrib.git]: gdc-ccache-bindings: initial commit23:55
cruxbot[contrib.git]: gcc34-distcc-bindings: initial commit23:55
cruxbot[contrib.git]: gcc34-ccache-bindings: initial commit23:55
cruxbot[contrib.git]: gcc-fortran-distcc-bindings: initial commit23:55
cruxbot[contrib.git]: gcc-fortran-ccache-bindings: initial commit23:55
cruxbot[contrib.git]: mingw32-gcc-distcc-bindings: initial commit23:55
cruxbot[contrib.git]: mingw32-gcc-ccache-bindings: initial commit23:55
cruxbot[contrib.git]: mingw32-gcc: remove masquerade symlinks in favor of bindings23:55
predatorfreakHOLY FUCK23:56

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!