IRC Logs for #crux Thursday, 2011-12-29

*** horrorStruck has joined #crux00:36
*** v33 has joined #crux01:19
v33sup guys01:19
Romsternot alo01:25
Romstert01:25
v33oh? why not?01:36
Romsterrelaxing now was hacking away at stuff earlier.01:44
v33oh? sounds like fun...how are you spending your holidays ?01:46
Romsterhttp://www.youtube.com/watch?v=Ou1uF9BKbKo01:49
Romsterhacking at bugs now i've had enough for the time being, went for a short rive watch some movies. browse some youtube :D01:50
v33dear god. how much porn can that store? xD01:52
Romstermore than a lifetime i'd say01:55
v33haha. i wanted to pick up some books on backtrack01:57
v33it seems really interesting01:57
Romsteri want a nice hdd enclosure but i can't decide what one and i need more money yet to02:02
v33i have an enclosure kit lol02:05
*** v33 has quit IRC02:55
*** ardo has quit IRC03:54
*** ardo has joined #crux04:03
Romsterhttp://www.youtube.com/watch?v=kignGE77l_I04:46
Romster1951 UNIVAC I - Mercury delay line Memory for computers04:47
*** vaddi has joined #crux04:50
Romsterhttp://www.youtube.com/watch?v=QrgTtZXuj4w awesome vortex cannon05:00
*** ardo has quit IRC05:37
*** vaddi has quit IRC05:49
*** mike_k has joined #crux05:58
*** gb_away has quit IRC06:36
*** gb_away has joined #crux06:37
*** gb_away has quit IRC06:57
*** gb_away has joined #crux07:48
*** ardo has joined #crux08:09
jaegeramazing how much things have changed since that video... I have more than 70TB in far less rack space than that08:12
Romsteryeah08:25
Romsternogagplz and i are attacking the rest of the wine optional deps.08:25
Romsterdbus-glib-32  freeglut-32  lcms-32       libgpg-error-32  libv4l-32  prelink-32  util-linux-ng-3208:25
Romsterelfutils-32   hal-32       libgcrypt-32  libsdl-32        openal-32  sane-3208:25
Romsterso far08:26
Romsterexcept libsdl-32 was for something else.08:26
Romsterso when its all done a README on what each does.08:26
jaegerdo you actually USE all of those or just for fun?08:26
Romsterwell he just wants to cover all the bases so in future it's not because of a missing port.08:27
Romsterprelink would be important for some games with copy protections08:28
jaegerso not really, heh08:28
Romsterdown to wine http://paste.pocoo.org/show/527314/08:28
Romsterwell it'l be me maintaining them ones so it wont be any extra workload on you.08:29
Romsterand they can be in the README for extra functionality.08:29
jaegerI'm not complaining, I just find it odd to do work for work's sake08:29
jaegerwithout a true purpose except someone might want it in the future08:30
Romsteri may have a use for prelink and i might use usb compat later on but dunno about the rest.08:30
Romsterat any rate i'm learning more about multilib08:30
jaegernice :)08:31
Romsteri know wine is pretty big08:32
Romsterwhen you look at how many things it can use, and who knows we might even get more users in crux multlib because of it.08:33
jaegerperhaps08:33
Romsterthough honestly most wine users use ubuntu and know very little so that maybe wishful thinking.08:33
jaegerI admit I'm surprised at how little interest there is in multilib08:34
Romsteroh i had a interst in it ages ago jsut didn't understand it enough to do it from scratch.08:34
jaegernor did I at the time, but you and I seem to be about the only ones now who care... aside from your friend08:35
Romsterand nogagplz08:36
Romsterso there is 3 currently.08:36
Romsteroh you already included him lol.08:36
Romsterwell x86-64 has had how many downlaods 80 something i think tek or frinnst said int he past few months? or have you, how many have download that multilib iso jaeger ?08:37
Romsteri need coffee brb08:37
jaegerno idea on that, haven't checked yet08:38
jaegerI bet it's not many08:38
jaeger11 total downloads, 4 unique IP addresses08:41
jaeger2 of those are my work and home IPs08:42
*** vaddi has joined #crux08:44
Romsterso that counts me and my friend...08:44
Romsterno more testes to multilib?08:44
Romstertesters*08:44
Romsteri guess they prefer pure.08:44
Romsterthough as jue said we could make pure x86-64 the offical and just have glibc or what have you ready to overlay multilib if they need it.08:45
jaegeryeah08:45
jaegerI think frinnst said he might test it but that sounded pretty noncommittal to me08:46
*** horrorStruck has quit IRC08:46
Romsterhe maybe waiting for us to iron the bugs out perhaps08:46
jaegercould be08:47
Romsterwell if it all falls flat over i'm still gonna use multilib population of 3 :D08:49
jaegerI still plan to use it08:51
Romstercool well the old saying is "if you build it, they will come" and it may happen.08:54
jaegerheh08:56
Romsterwe are ignoring opencl i don't think anything even uses that at all.08:58
rmullI vote for pure being the default09:11
rmullObviously because I have no need for multilib09:11
rmullBut mostly because if multilib is something you can just add later, that's good enough09:11
Romsteryeah might need some small changes to core in pure though.09:17
Romsterelse it'll be a toolchain recompile when you overlay multilib.09:17
Romsternot sure of the extent of this yet it's early days yet09:17
juehello09:19
Romsterhi jue09:20
*** mike_k has quit IRC09:21
jaegershouldn't need much toolchain recompiling if the toolchain is replaced as part of the conversion09:23
jaegerheyo, jue09:23
jueRomster: that's no completely correct, I said that having a multilib toolchain as the default would be nice if there's not much or better no impact on the system at all09:25
jue... and it wouldn't hurt to ship a glibc-32 package with out iso09:26
jues/out/our/09:26
juethat's at least my opinion ;)09:26
jaegerI do remember you saying that, it seemed a reasonable approach to me09:27
jaegeron a side note regarding the ISO I'd like to add parted to it at some point09:27
jaeger(for GPT support)09:27
jaegeras UEFI and >2TB drives become more common fdisk is coming up short09:27
jue+1 for that09:27
juethere was at least one other user requesting parted as well09:28
thrice`I rather like pure64 as the default, with the multilib upgrade as an easy option.  but i'm bias, since i've never needed anything 32-bit yet :-)09:28
jaegerjue: ok09:29
Romsteryea i do remember that jue09:30
jaegerI guess most users need 32-bit less than I do09:30
jaegerI don't use very many 32-bit things but I do use them pretty often09:30
Romsterjaeger, why not gdisk ?09:30
juethrice`: what's the problem if your gcc is capable to produce 32bit stuff?09:30
Romsterhttp://www.rodsbooks.com/gdisk/09:31
thrice`jue, as a package, not much.  but I thought on updates, it makes building gcc take much longer?09:31
* jue looks at jaeger 09:31
jaegerRomster: gdisk has some odd behaviors, such as automatically converting mbr to gpt, etc.09:31
jaegerNot something I'd like if I weren't expecting it09:32
Romsteri haven't timed it but i'm building a gcc42 port for a older program as multilib and it seemed to take abut the same time as it did on pure 32bit.09:32
jaegerfor the pure64-with-multilib-toolchain thing, as I see it the only bloat there is glibc-32, which isn't big09:32
jaegergcc without any -m64/-m32 options defaults to -m64 in this case09:33
Romsteryou kidden jaeger i thought that's what gparted did in the past. without asking for permission09:33
thrice`maybe I'm not remembering right.  I thought at least when I tried on gentoo, a pure64 gcc built quicker than a multilib gcc09:33
jaegerso if you install the toolchain over a pure64 system but don't change things like pkgmk.conf it should act basically the same as pure6409:33
jaegerpure64 gcc *will* build faster than multilib09:33
jaegerit's building at least one entire stage less09:33
thrice`right, I thought it was pretty significant09:34
jaegerhow often do you build gcc, though?09:34
thrice`yeah, sure, it's easy to lock too :_)09:34
jaegerWell, I certainly don't want to force multilib down anyone's throat09:34
jaegerif pure64 is the popular choice, that's how it goes09:35
Romsteri find having the option for 32bit should one need it pretty good compromise. at the very least for thsoe not using 32bit is only a gcc and glibc-32 *shrugs*09:36
Romstermany a ML vote on it.09:36
Romsters/many/maybe/09:36
jaegerI'm content to leave pure64 the "default" and just update the wiki article about converting09:37
Romsterthat's the other option.09:40
juehmm, if offical CRUX makes the switch to x86_64 we should try to combine our efforts, and I don't see a good reason why we shouldn't use the multilib toolchain09:40
Romsterjue has a valid point we could just overlay multilib opt core contrib and have what ever as 32bit as users need.09:41
juethat doesn't mean that we must have official mulilib-repos09:41
Romsterhmm09:42
Romsterjaeger, you intend to move them multilib to crux.nu later on and me and nogagplz gaining access? or just me to maintain wine, zsnes etc.09:43
jueRomster: why do we need a overlay for core/opt?09:43
Romsterwine 32bit' stuff.09:43
jueisn't that additional stuff, like lib*-32?09:44
*** lasso has joined #crux09:44
Romsterhttp://morpheus.net/gitweb/?p=jaeger/core-multilib.git;a=shortlog;h=refs/heads/2.709:44
Romsterhttp://morpheus.net/gitweb/?p=jaeger/opt-multilib.git;a=shortlog;h=refs/heads/2.709:44
Romsterpretty much09:44
jueTBH, if not it's a no-go for me09:44
Romsterdependent ports in 32bit for mostly wine currently though i got a few others that need 32bit but not many.09:45
jueRomster: all I can see is additional stuff, and some modification to pkgmk09:46
Romstertake it or leave it i'll keep running multilib which ever way you decide. and there is no rush on it.09:46
Romsteryeah to pkgmk.conf to allow 32bit options.09:46
jueRomster: no, that's a _very_ important point09:47
Romsteroh and pkgmk for the .32bit file09:47
Romsteri forgot about that part that needs to be reviewed.09:47
Romsterand merged if given the ok into the core pkgutils if prure64 gets the multilib gcc.09:48
juesure, that's not the problem09:48
Romsterif yoru asking me it's already turning into a too big of a headache to explain, jsut look over the changes and decide for yourself.09:48
juewell, there's the chances I'd expected, everything else is additional stuff the normal user will not see/need09:51
jues/chances/changes/09:51
*** jdolan_ has joined #crux09:51
jaegerRomster: only if it became official, which seems unlikely. no reason to spend the resources on crux.nu when I can already host it myself09:52
*** jdolan has quit IRC09:53
jueIMO in any case we don't need overlays but additional repos ;)09:53
jaegerI'm not sure I understand the difference in this particular case, would you elaborate?09:55
jueor is there's any reason why e.g. glib-32 has to reside in the same repo as glib?09:55
Romsterwell i was saying overlay as in a port treethat's on top of the prtdir /usr/ports/.... in prt-get.conf respect.09:55
Romstersorry if i confused you with that term.09:55
jaegertake pkgutils as an example; pkgmk and pkgmk.conf have changes for multilib, even though they're small... so I have a core-multilib/pkgutils09:55
jaegerIt's not overlaid onto core in the sense that it ends up copied into core/, it's in its own repo09:55
jaegerbut it's definitely used in place of core/pkgutils09:56
juejaeger: no, I'd use the same pkgutils for everything09:56
Romsterhttp://romster.dyndns.org/linux/ports/crux/multilib-romster/ other bits nogagplz and myself are working on not cleaned up alot yet.09:56
jaegerjue: it would have to support some other way to distinguish, then, such as the cruxppc .footprint and .footprint64 or whatever they're called09:57
Romsterif multilib gcc glibc-32 got in core then pkgutils would need them changes merged in to. but i think that's essentially all. that needs changing.09:57
jaegersome ports are perfectly fine without changes but for example, mesa3d09:57
jaegerwhen built 64-bit it leaves some things out09:58
jaegermultilib OR pure09:58
jaegerwith that said I tried to make my changes minimal09:58
jaegerif the file ".32bit" exists it switches to 32-bit building09:58
jaegerthe default is 64-bit09:59
frinnstjaeger: cloneable path for your repos?10:08
juejaeger: sorry, was AFK10:08
thrice`Romster, lots of ugly files in there :-)  .swp files, backups, etc :>10:09
Romsteryeah i know must clean it up more it's not done yet10:09
Romstergetting tired and there is still more todo10:10
juejaeger: your repos are no longer available via git?10:10
Romsterfrinnst, http://morpheus.net/gitweb/?p=jaeger/core-multilib.git;a=tree;f=ports;h=c406d7b0a42f3330eaeac65befa9af4a9153d8d3;hb=refs/heads/2.710:11
jaegerfrinnst: git://morpheus.net/jaeger/{core,opt,xorg}-multilib.git10:11
jaegerjue: ^10:11
juejaeger: that's what I have but got an error10:12
jaegeroh, I had to restart git-daemon, try now10:12
jueahh, it's working again now ;)10:12
jaegersorry, no idea what happened10:12
Romstercomputers for ya :D10:13
jaegerindeed10:13
Romsteri'll do more work on ports tomorrow10:13
thrice`I think debian and gentoo just provide a larger "32-bit-crap" package, no?10:14
jaegerdebian/ubuntu do10:15
jaegernot sure what gentoo does10:15
jaegerI *really* don't like that method, personally10:15
jaegernot minimal at all :)10:16
thrice`ah, yeah, they do too.  "emul-linux-x86-{xlibs,soundlibs,sdl,opengl,medialibs,...}10:16
jaegerah10:16
jaegerI used to be a fulltime gentoo user before using crux, ironically... been so long that I'm not very familiar with it anymore10:17
thrice`me neither, but it's interesting to see how others solve the problems :-)  debian has the multiarch stuff, which is neat (but, probably not achievable for source-based distros so easily)10:18
juejaeger: I cannot see a diff between libgmp from multilib and x86_64?10:18
jaegerjue: I don't think there is much difference, just hadn't had a chance to really dig into that yet10:19
jaegerjue: originally I believed I was going to have to build gmp{,-32}, mpfr{,-32}, and mpc{,-32} all for the multilib gcc but it didn't turn out that way10:19
jaegerwhich made sense once I learned more about it10:19
juejaeger: my best net is that everything without the -32 extension is the same?10:20
jues/net/bet/10:20
jaegernot always10:20
jaegerotherwise I would have no need for those duplicate ports in my -multilib repos10:20
jaeger(gmp may not be needed regardless but some are)10:21
jueI'm talking about core only10:21
jaegerwell, there are definitely some differences10:22
jaegergcc is a good example10:22
*** lasso has quit IRC10:22
jaegerusr/lib/gcc/i686-pc-linux-gnu vs usr/lib/gcc/x86_64-unknown-linux-gnu10:22
jaegerlib vs. lib3210:22
jaegeretc.10:22
jueyeah, gcc is obvious10:23
jaegerbinutils has small differences, ldscripts10:24
jaegerunzip can't use x86 asm, util-linux-ng includes setarch10:26
jaegersmall things10:26
jueisn't the same for x86_64?10:27
jaegernot sure, honestly, haven't dug through them yet10:28
juewell, I thought that you've started with our x86_64 repos and only changed stuff that needs to be changed?10:30
jaegerI mostly started with the 32-bit core. I wanted to diverge as little as possible from default10:31
*** horrorStruck has joined #crux10:32
jueok, should be easy to make diffs to see which ports are the same or can be merged together10:34
cruxbot[contrib.git/2.7]: dia: 0.96.1 -> 0.97.210:40
*** horrorStruck has quit IRC10:46
*** ardo has quit IRC11:02
entehttp://npaste.de/tvRr/tcl11:06
ente:)11:06
enteI think I like tcl's http lib11:06
juenice stuff to block ad banners -> http://pgl.yoyo.org/adservers/11:25
jaegerneat11:29
*** acrux|G4 has joined #crux11:30
entealso this: http://someonewhocares.org/hosts/11:34
*** joe9 has quit IRC11:38
*** joe9 has joined #crux11:41
*** Rotwang has joined #crux12:04
*** mike_k has joined #crux12:55
Rotwanghttp://d24w6bsrhbeh9d.cloudfront.net/photo/1323577_460s.jpg12:56
ente:)12:58
*** rmull has quit IRC13:07
*** j^2 has quit IRC13:07
*** rmull has joined #crux13:08
*** mike_k has quit IRC13:27
*** mike_k has joined #crux13:41
niklaswegood evening13:49
Rotwanghttp://d24w6bsrhbeh9d.cloudfront.net/photo/1237274_460s.jpg13:50
niklaswehmm I wondering if I should install crux on my macmini instead of osx.. just because Im bored13:50
*** rivia has joined #crux14:12
*** acrux|G4 has quit IRC14:47
*** rivia has left #crux15:12
enteRotwang: try aplus15:20
enteit's an APL dialect ;P15:20
*** Rotwang has quit IRC16:41
*** j^2 has joined #crux20:09
*** SiFuh has quit IRC20:18
*** SiFuh has joined #crux20:20
*** vaddi has quit IRC20:31
*** mike_k has quit IRC21:04
*** Dudde has quit IRC21:38
*** Dudde has joined #crux21:39
*** SiFuh has quit IRC23:45
*** SiFuh has joined #crux23:47
*** SiFuh has quit IRC23:53
*** SiFuh has joined #crux23:55

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