IRC Logs for #crux-devel Wednesday, 2019-03-27

*** nthwyatt has joined #crux-devel03:55
*** nthwyatt has quit IRC03:57
*** darfo has joined #crux-devel03:58
*** deus_ex has joined #crux-devel09:10
frinnstjaeger does the vmware console launcher thingy work for you in 3.5?10:52
frinnstnot sure what the log *should* look like when it works but I suspect its not very pretty. anyways this is what I see now:10:55
frinnst2019-03-27T11:53:14.166+01:00| vmrc| I125: HostinfoLsb: Error: no distro file found10:55
frinnst2019-03-27T11:53:14.174+01:00| vmrc| I125: HALLoadLibrary: Could not dlopen libhal.so.1: libhal.so.1: cannot open shared object file: No such file or directory.10:56
frinnst2019-03-27T11:53:14.174+01:00| vmrc| I125: HALLoadLibrary: Could not dlopen libhal.so.0: libhal.so.0: cannot open shared object file: No such file or directory.10:56
frinnst2019-03-27T11:53:14.199+01:00| vmrc| I125: UUID: Unable to open /sys/firmware/efi/systab: Permission denied10:56
frinnst2019-03-27T11:53:14.199+01:00| vmrc| I125: UUID: Unable to open /dev/mem: Permission denied10:56
pedjahal? really, vmware?12:00
pedjalss thingy is /etc/os-release, I guess?12:01
pedjalsb*12:01
pedjaah. lsb-release12:04
*** Workster has quit IRC12:29
*** Workster has joined #crux-devel12:30
ryuojaeger: poke. i have a question to bounce off you. pkgrm's core logic does a second pass through the database to avoid removing files that are referenced by other packages.13:04
ryuohowever, most files are not shared between packages.13:04
ryuothe only ones i believe can be, given the design of pkgutils and OS limitations, are directories.13:05
ryuoit would seem to me that this check is utterly pointless as it can be handled during the actual removal phase.13:05
ryuojaeger: your thoughts?13:05
ryuoon second thought, there's one situation where it still matters.13:29
ryuoempty directories.13:29
ryuothe current logic preserves them if they're empty yet still a part of another package's file list.13:29
ryuoand there's no way to know whether they should be kept or not from the file system tree13:31
ryuohm. think i'm going to need some internal state tracking...13:43
ryuomostly as a sanity check against improper usage patterns.13:43
jaegerfrinnst: I haven't used it, I think. you mean the VMRC?13:50
jaegerryuo: sounds like you answered yourself :)13:52
ryuojaeger: indeed, but i'm re-evaluating what the original author did.13:53
ryuoi was left with the impression that they don't understand POSIX well...13:53
ryuoi'll probably end up creating a subroutine to build a hash table from the file list.13:53
ryuootherwise looking up files per package is going to be slow.13:54
ryuoin a few weeks i should be done with pkginfo's functionality13:55
ryuoand possibly pkgrm13:55
ryuopkgadd is the one i'm saving for last.13:55
ryuoit's the most complicated one to do correctly.13:55
frinnstyeah14:03
frinnstsilly qt5 fails to build and its fucking impossible to read out any errors. make -j1 weeee14:03
ryuojaeger: how essential is it to retain the exact same error formatting?15:03
ryuoafaik, pkgmk doesn't depend on the error message structure...15:04
ryuoi'm checking prt-get now.15:04
ryuointeresting.15:05
ryuoseems neither does prt-get15:05
ryuoif there's no opposition, i'd appreciate some freedom in mapping errors to text... i'm having to map errors from more than just the library.15:07
ryuosystem errors, lib archive errors, ...15:07
ryuoseems it should be doable... pkgutils writes all errors to cerr. it shouldn't break any software if i change the error text.15:12
ryuowhich is stderr15:12
ryuoi'm trying to provide more information about errors.15:13
*** onodera has joined #crux-devel16:21
*** onodera has quit IRC16:39
juejaeger: thx for the new ISO, I'll try test2 tomorrow17:39
juejaeger, frinnst: looks like your update went smooth?17:40
jaegersounds good, thanks17:44
jaegerI think it went smoothly but if you see anything wrong let me know17:44
jaegerryuo: personally I don't care about the error messages, though I was going to retain them as-is for my first iteration, then clean them up later17:44

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!