IRC Logs for #crux-devel Tuesday, 2015-02-24

*** leo-unglaub has quit IRC02:11
*** mavrick61 has quit IRC03:26
*** mavrick61 has joined #crux-devel03:27
*** deus_ex has quit IRC05:04
frinnstchecking for clang... clang07:44
frinnstchecking for C compiler default output file name... configure: error: C compiler cannot create executables07:44
Romsterdid you rebuild clang after llvm?07:44
frinnstthat was clang failing07:45
frinnst=======> ERROR: Building '/usr/ports/pkg/clang#3.5.1-1.pkg.tar.gz' failed.07:46
frinnstbetter using a proper compiler to build clang ;-)07:46
frinnstremoving clang and rebuilding it seems to do the trick07:48
frinnstclang doesnt use the system llvm port ?07:48
Romsteroh it used system clang to build itself07:51
Romsterchecking for C compiler default output file name... configure: error: C compiler cannot create executables08:15
Romsterwell thats dumb if it can't build from itself08:15
frinnstit probably can, but llvm broke it08:16
Romsteryeah i realize that but if we didn't have gcc on our system how would you update clang08:17
frinnstseems the footprint for clang hasnt been updated?08:28
Romsteri guess this is why splitting clang and llvm wasn't wise. teK_ you will need to force clang to compile with gcc.08:28
frinnstpkgadd: listed file(s) already installed (use -f to ignore and overwrite)08:29
Romsterpkgrm clang08:29
Romsterchecking for C compiler default output file name... configure: error: C compiler cannot create executables08:29
Romstertries in a container08:30
frinnstwell, splitting them was more or less a requirement since llvm is needed for mesa08:31
frinnstand forcing another compiler on all mesa systems is asking for trouble08:31
Romsteronly to half the time of building llvm08:31
Romsteroh right my ccache issue of dangling symlinks after removing clang -_-08:32
Romsterthat fixed it08:33
Romstercompiler-rt libraries what do these do?08:35
Romsterand clang-tools-extra08:35
*** Romster has quit IRC08:54
frinnstteK_: clang footprint needs to be updated; clang installs usr/bin/FileCheck usr/bin/count usr/bin/not that are already installed by llvm09:05
*** Romster has joined #crux-devel09:17
RomsterteK_, clang footprint is not updated and there is 3 conflicting files09:45
Romsterusr/bin/FileCheck usr/bin/count usr/bin/not09:46
frinnstI just said that09:46
Romsteroh i was off09:46
frinnstoh right, sorry09:46
Romsterthem 3 are from llvm so rm $PKG/... them 309:47
*** Romster has quit IRC10:14
*** Romster has joined #crux-devel10:17
jueprologic: where does your experience with runit comes from? Have you tried Void Linux or something else?11:31
prologicwhat no I never claimed any experience with runit :)11:33
prologicRomster has tried it and used it successfully not i11:33
prologicI have however researched both runit and finit and a couple others whose names I can't remember now :)11:34
prologicrunit's drawback is all the symlink it uses all over the place11:34
prologicfinit seems more fitting to crux ihmo - single file config11:34
Romsteri've used runit and i've read of others using runit in #crux in the past.11:36
Romsterbut runits documentation is lacking.11:36
Romsterfinit might be better.11:37
Romsternot tried finit yet.11:37
Romsteryes runit used symlinks to enable services.11:38
*** Feksclaus has joined #crux-devel11:46
*** mavrick61 has quit IRC15:40
*** Feksclaus has quit IRC15:46 is down too, not just crux.nu16:19
frinnstpoor charlie16:19
jueFYI, Steffen Nurpmeso is the (very active) maintainer of s-nail aka nail which we use as mailx16:19
frinnsthavent looked at my mails yet - 92 unread16:22
jueTBH, I never could imagine that such a simple thing like start-stop-daemon would create such a fuss16:22
juewrt to needed features: in the meantime I've converted most of my opt-daemons to ssd and used almost anything it provides16:25
jaegerjue: no doubt, was a big surprise16:25
juewhat I absolutely not understand is the aversion against it because it comes from debian ???16:27
jaegerI think there's some hyperbole involved there along the lines of "well, the next step must be systemd"16:40
frinnstyeah, fucking insane18:11
*** leo-unglaub has joined #crux-devel21:17
frinnstlooks like the troubles is due to the DC's internet provider22:04
frinnstanyways, thats my theory from reading local papers :)22:04
frinnstall sors of chaos in the city where is hosted22:04
frinnstthe city has pretty much lost its uplink22:08
Worksterjust tested and both are up22:16
prologicI don't necessarily think there's an avesion to ssd22:45
prologicit's just ihmo it's solving a part of a much bigger problem that we have22:45
prologicwhich is also why the discussions of an init alternative have spawned22:45
prologicssd only solves one very small part of the problem22:46
prologicand ihmo not very well22:46
prologicbut as I've also said, I'm over it :) if this is the way forward fine22:46
prologicbut like jaeger also mentioned we should probably start to consider an init alternative22:46
prologicwhich would make ssd redundant22:46
prologicI apologize if anything I said was terribly negative or fille dwith emtoion :)22:47
prologicI was only intially upset because of the poorly communciated change and the potential to break systems so easily if you weren't carefcul :)22:47

Generated by 2.11.0 by Marius Gedminas - find it at!