IRC Logs for #crux Thursday, 2011-08-18

*** rmascarenhas has quit IRC00:23
pitillogood morning00:58
frinnstnothing good about it! :)02:46
frinnstmm, 3 cups already02:46
*** Romster has quit IRC03:08
*** Romster has joined #crux03:09
laenCoffee coffee coffee <303:37
Romsterdinner dinner dinner then coffee coffee coffee :P03:38
Romstersorry about my ping outs earlier stormy weather03:38
Romsterlast one was me swapping a 100mbit for a 1gigabit lan card.03:39
*** SiFuh has quit IRC04:07
*** SiFuh has joined #crux04:09
luxhthink it is coffee time now again04:14
enteI agree with you on that04:42
* ente sips coffee04:42
laenMmmmm coffee. Getting some quick before the meeting starts.05:59
luxhslashbeast: time for morning coffee06:18
enten cups of coffee on the wall, n cups of coffee06:19
entetake one down, pass it around06:20
enten-1 cups of coffee on the wall :P06:20
slashbeastI like pepsi better over coffe.06:23
entewhere's the difference? :P06:24
* frinnst is an espresso enthusiast06:24
slashbeastwhen i wakeup I am like 'you all are going to die'06:24
slashbeastthen I drink pepsi06:24
enteand then you act like you're on brawndo and start kicking everyone's ass all the time?06:26
slashbeastnope, then I am like 'no war, make love'06:28
enteI need my two cups of coffee in the morning, otherwise I feel like I'm still asleep for the whole day06:33
enteam I an addict, or is this normal?06:33
entebesides, I don't really drink that much coffee, it's just that I feel utterly and completely numb without it/06:34
entelike someone hit me on the head with a frying pan, but without the pain, and I can't really wake up, and my head is filled with some sort of slime06:34
*** tnut has joined #crux07:26
cruxbot[opt-x86_64.git/2.7]: firefox-java-plugin:update to 1.6.0_2608:04
pedjamy first commit, something simple :)08:05
*** RyoS has quit IRC08:14
*** RyoS has joined #crux08:15
enteso you're an opt guy now? :P08:15
entefirefox-java-plugin sounds unimportant enough to be moved to contrib :P08:16
pedjaente: baby steps :)08:22
entewhy was there no announcement mail on the list that there is a new opt member?08:23
enteor are you not that new after all?08:23
pedjafrinnst kindly invited me to join, and I play only with opt-x86_64 repo :)08:27
pedjahmm, 2.6G of tmpfs is not enough for FF build..08:29
pedja'no space left on device'08:29
entesurprise, surprise ;)08:31
pedjaqt4 needs less space, iirc08:32
pedjabut someone here mentioned a while back that ff, gcc, qt4 build just fine with 2.5G on their machine, so I don't know...08:36
pedjaok, tmpfs is now 2.8G, let me see if it will fail again :)08:38
pedjaextra 200Mb did the trick.09:01
*** RyoS has quit IRC09:04
*** tnut has quit IRC09:10
*** tnut has joined #crux09:27
*** j^2 has joined #crux09:29
*** Modify has joined #crux10:16
Modifywhat do you say how mich Crux is alive, from 0 to 10? I mean, the latest version and the homepage is nearly an year old... it crux still actively developed? just curious10:18
thrice`  <-- shows activity in crux10:21
luxhstuff near perfection doesn't require much development10:22
Modifywow i didn't see that10:22
Modifynear perfection... are you seriously?10:22
thrice`usually releases are about once a year, when enough stuff (like the toolchain) has updated enough to justify it :-)10:22
*** tnut has quit IRC10:25
Modifyok, just another question: what would you say are the three main differences from Arch? :D10:33
thrice`it's source-based, and it seems to be a little simpler in design10:34
Modifysource based: do you have some automatic configuration options (like Gentoo or Freebsd), or do you edit the makefiles?10:36
thrice`not sure I follow?  simple build scripts for packages, with a semi-automated wrapper to do dependencies and such10:37
ModifyI mean configure options, USE flags in gentoo language10:37
thrice`ah, no.  usually packages have a minimal set of dependencies as the requirements10:38
pitilloModify: check the handbook if you have some time, there you'll find a very good explanation about CRUX10:40
ModifyI'm reading but I dont see compile options anywhere... e.g. if I want to build bash with or without completion support, what would I have to change?10:43
thrice`if you want to deviate from an official port in some way, the easiest thing to do is just create your own little repo, say in /usr/ports/mine/, copy it over, edit the port, and build10:44
Modifysounds complicated :D I generally like source based for being able to set my own compile options... e.g. in FreeBSD you can "make configure" to change some options10:46
Modifynothing like this in Crux, right?10:46
Modifyjust curious :P10:47
thrice`i've never found a text editor to be complicated, but meh.  I rarely ever edit ports from the defaults, I guess10:49
thrice`for your specific question, bash-completion has its own port :)10:51
ModifyI know, but if I knew I will not use it anyway, I'd like to compile bash without any support for this, making the binary a bit smaller10:52
pitilloModify: is this complicated? using ports as build templates, for bash
*** Rotwang has joined #crux10:52
thrice`Modify, typically that's the default for ports, pretty minimal10:52
rmullSo I've gotten some .footprint mismatches after some builds and I just ended up coping the port locally and editing the .footprint. Is there a better way to deal with this?10:53
Modifythrice: but these additional ports won't work then, would they? ^^10:53
thrice`I guess I don't follow :(10:54
pitillormull: you can update the footprint... or ignore them. Check prt-get man and you'll see those options. The mistmatch shouldn't be a problem if you see NEW files... if there are missing files may be there is something wrong10:54
jaegernot necessarily wrong but different from the maintainer's build10:54
Modifythrice: anyway, I think my question is answered10:55
pitillojaeger: right10:55
rmullpitillo: I was missing files, but it's probably because I don't have a lib that one of the ./configure scripts was looking for, so my copy was built without support for something or other10:56
rmullIf that makes sense10:56
pitillormull: may be the maintainer build the port outside a clean enviroment, it could be possible that he forgot to add a dependencie... there are some reasong to find those mistmatches10:57
pitillos/reasong/reasons :)10:57
rmulland the -uf update footprint argument will cause that port's .footprint file to be updated on my machine? what exactly happens when you specify -uf during a depinst?10:58
pitillormull: you can't use it with the depinst option if I'm right10:59
rmullI guess I'll have to experiment, I haven't tried the extended functionality of prt-get just yet10:59
rmullThanks for the pointers though10:59
pitillonp :)11:00
*** Modify has quit IRC11:01
*** RyoS has joined #crux11:40
frinnstyou can edit /etc/pkgmk.conf to ignore <NEW> footprints11:57
Rotwangis there any person around here that by chance is in possesion of "mastering cmake" pdf?12:02
tilmanis that the official cmake book?12:02
tilmanwritten by  these K-something dudes?12:02
tilmanbook fucking sucks12:02
tilmanworst book ever12:02
tilmanalmost useless12:03
tilmandont get it12:03
Rotwangtilman: unfortunately I couldn't find any good comprehensible tutorial12:03
tilmancmake sucks12:03
tilmanstay away12:03
tilmanbest advice i can give :P12:03
Rotwangtilman: why?12:03
tilmanwhy why12:03
RotwangIt looks very nice [;12:03
tilmancan't you just take my biased opinions for truth?12:03
Rotwangit looks less messy than autostuff12:04
tilmanbut it's also fucked up12:05
tilmanit can be really hard to build proper dependency chains12:05
tilmanalso, it's INSANELY hard to debug cmake weirdnesses12:05
thrice`tilman, doesn't xmms2 still use waf? :p12:05
tilmanRotwang: i can recommend make (for smallish stuff), scons and autotools12:05
entewaf waf12:09
Rotwangtilman: we are using make for a huge component, and it is a complete mess12:10
Rotwangmake is great for small stuff of course12:10
RotwangI wonder how they managed to survive with make in pre autotools times12:11
Rotwangbtw. I'm coding some C for fun, and I can't find any good practices guide12:47
Rotwange.g. on howto not forget about free()ing stuff or error handling12:48
Rotwangany good resources appreciated12:48
*** ThePub_ has joined #crux12:56
*** ThePub has quit IRC12:56
*** ThePub_ is now known as ThePub12:56
rmullRotwang: If you're concerned about forgetting free() and memory leaks, look into learning how to use valgrind during debugging13:11
RotwangI was affraid someone is going to say that [;13:12
rmullIt's easy enough13:15
RotwangI've hoped that there are some good practices that I could apply actually before compilation13:15
rmullYou could practice by implementing a linked list and its associated operations13:33
tilmanRotwang: there's techniques, but for learning its best to discipline yourself =)13:34
Rotwangrmull: I think linked list isn't remedy for my memory management problems13:35
*** iToast has quit IRC13:50
*** Evil_Bob has joined #crux13:53
rmullRotwang: Why not? If you write some basic linked list ops, you can make sure (using valgrind) that you don't orphan any nodes, forget to free something, or get the syntax wrong. Seems like a good exercise to me. Though I'm not sure what it is you're looking for... Just a set of tips for how to not forget to use free()?13:55
Rotwangwhat linked list has to do with all of that?13:57
tilmanusually you'll allocate memory for the linked list nodes on the heap. using malloc ;)14:01
*** tnut has joined #crux14:01
Rotwangbut I said nothing about linked list14:04
tilmanyou asked "how to learn"14:04
tilmanhe said "here's an example practice assignment"14:04
tilmanyou said "wha? me no understand"14:04
Rotwangtilman: no14:05
RotwangI didn't ask "how to learn" to be specific14:05
RotwangI just asked about set of good practices to avoid problems with memory management and the like14:06
Rotwangit has nothing to do really with data structures14:06
Rotwangit is similar to me asking "how to drive a car" and someone replying "take that truck over there and learn"14:08
*** SiFuh has quit IRC14:09
*** SiFuh has joined #crux14:10
rmullThe data structure gives a (simple) purpose to the memory management - but I guess I don't understand the request, so I'll let someone else chip in14:15
Rotwangrmull: the point is I have to remember to do that separate action like free'ing something14:16
Rotwangin c++ it is solved with destructors, which are executed automatically14:17
Rotwangin c even if I create linked list, binary tree, or any other graph I still need to remember about freeing it by hand14:17
Rotwangok, I'm not sure if 'new' classes are deleted automaticaly, probably you have to delete them yourself14:19
tilmanand yes, new+delete same as malloc+free (mostly)14:23
Rotwangnow I'm starting to understand why memory management is an issue in c/c++14:25
*** RyoS has quit IRC14:25
*** RyoS has joined #crux14:25
ThePubjust write everything in Java and you cannot consume more memory than the VM allows ;p14:43
*** tnut has quit IRC14:44
*** Evil_Bob has quit IRC16:49
*** Rotwang has quit IRC17:02
*** cruxbot has quit IRC17:08
*** cruxbot has joined #crux17:10
ThePubeverything should be java.  just imagine the crux tools rewritten in the superior java language!17:34
enteoh my17:38
ThePubdoesn't it sound like a great idea? the packages could just be serialized java objects, yeah.18:07
*** ThePub has quit IRC18:30
*** Ovim|afk_nA has joined #crux18:37
*** Ovim-Obscurum has quit IRC18:40
*** Romster has quit IRC19:58
*** Romster has joined #crux20:01
*** Dudde has quit IRC21:26
*** Dudde has joined #crux21:27
*** ThePub has joined #crux22:13

Generated by 2.11.0 by Marius Gedminas - find it at!