IRC Logs for #crux Friday, 2014-12-12

*** thetornainbow has quit IRC00:03
*** jdolan has joined #crux00:08
*** jdolan has quit IRC00:35
*** jdolan has joined #crux00:35
*** jdolan has quit IRC00:43
*** jdolan has joined #crux00:43
*** thetornainbow has joined #crux00:50
*** tilman has quit IRC01:04
*** tilman has joined #crux01:05
*** thetornainbow has quit IRC01:17
*** thetornainbow has joined #crux01:17
*** fengshaun has quit IRC01:37
*** fengshaun has joined #crux01:38
*** sh4rm4 has joined #crux01:47
*** nilp has joined #crux01:48
*** doomicide has quit IRC02:11
*** jdolan has quit IRC02:52
*** jdolan has joined #crux03:05
*** SiFuh has quit IRC03:10
*** SiFuh_ has joined #crux03:10
*** Feksclaus has quit IRC03:12
*** darfo has joined #crux03:26
*** mavrick61 has quit IRC03:39
*** jdolan has quit IRC03:49
*** jdolan has joined #crux03:49
*** mavrick61 has joined #crux03:49
*** xeirrr has joined #crux03:55
Worksteris there a partitioning tool that doesn't assume i want GPT. sick of this crap. all the tools i've tried write a GPT label on a DOS partiton label disk04:26
diversewhat about fdisk?04:27
Worksterto resize NTFS partitons and move them04:27
diverseah04:28
Workstergparted tried it assumes i want GPT04:28
Worksterhirens disk wizard does the same04:28
diverseyeah parted is pro-GPT04:28
Worksterwhen there is a DOS label why the eff create a GPT label.04:28
Worksternope no warning don't ask just effing does it04:29
Worksterand brakes the system04:29
diversehttp://geekpeek.net/resize-filesystem-fdisk-resize2fs/04:30
diverseWorkster: did you get my project link?04:31
*** xeirrr has left #crux ("Leaving")04:36
jaegerparted can do both04:36
jaegergdisk (gptfdisk) tries to force gpt04:36
jaegerwith that said, why not just use gpt? do you have some reason you can't?04:39
Worksteryes but i havne't had time yet04:40
Worksteri used gparted jaeger it forced GPT without even my consent04:40
jaegerIt's never done that to me04:41
jaegergdisk does, though04:41
Worksterwell done it to me and pissed me off so i formatted installed windows fresh and customer will have to install programs again04:41
jaegerwhat's the problem with using gpt?04:41
diverseI just go with gpt anyway, although now I use it for my uefi, rather than booting into bios compatibility mode04:41
Worksterit broke booting windows04:42
joacimgparted sucka when i try to use it too04:42
diverseI don't recommend parted anymore04:43
joacimnow i just use parted and fdisk04:43
jaegerI use parted for everything, heh04:43
joacimntfsclone for what the name implies04:43
diverseI guess parted would be good for resizing although04:45
joacimi had great luck with minitool disk wizard04:45
*** thetornainbow has quit IRC04:46
joacimit is similar to gparted, but without the same quirks and limitations04:47
Worksteri'll give that a try next time joacim ta04:53
joacim*minitool partition wizard04:54
joacimlooks a bit shady tho.04:56
diversewhy does it look shady04:58
joacim"share on facebook" in the ui, "free download" on the website04:59
diverseoh04:59
joacimtypical windows-tier crap that makes me think it comes with a toolbar04:59
diversethat is quite the omen indeed05:01
joacimI tried to resize a fat32 partition with gparted, it failed, no matter what i tried. minitool partition wizard did it without any issues.05:01
diverseuse the best tool for the job05:02
joacimcan't be bothered with that. i just use the first thing i can think of.05:08
joacimit is something i use once every other year, so i can't be bothered with finding the best one05:09
*** jdolan has quit IRC05:09
*** thetornainbow has joined #crux05:13
*** jdolan has joined #crux05:21
diversesigh, s/best/better/05:26
joacimwhatever works05:28
diverseminitool may not be the best tool, but it is the better tool compared to gparted. I changed from "best" to make it not absolute.05:29
diversebecause time is effort anyway05:33
diversegeez, everytime I try Fira Sans, I always go back to Dejavu Sans05:48
diverseMozilla is trying to make a better font but I think they failed on the mono side05:49
diversemight look good on a word document though05:51
*** phant0mas has joined #crux06:18
Romsterfonts are all subjective06:32
Romsterits like the emacs vi debate.06:33
Romsteryou never win06:33
joacimVIn06:35
Romsterdidn't you mean VIm06:37
joacimwin -> VIn06:37
Romsterhmm and a bunch of linux ports that need attention06:37
*** deus_ex has joined #crux06:38
Romsterj^2, hi, i did see you enter earlier and chat but you been quiet since then.06:39
*** nilp has quit IRC06:44
_root_guys I ported rodent applications fully working it is nice.06:54
_root_in the new release it adds rodent-desk. it creates a very nice desktop and icon on desktop ability06:55
_root_Romster: joacim diverse jaeger: [HELLO]06:56
Romsterlo06:56
Romsteri left hel at work :D06:57
_root_Romster: that bad?06:57
Romsterfelt like it06:57
_root_Romster: one of those days.06:58
joacimhey06:58
Romsterit happens, now to unwind06:59
_root_joacim: hey back06:59
*** asie has joined #crux07:11
*** nilp has joined #crux07:12
*** asie has quit IRC07:23
diverse_root_: what's up?07:39
_root_diverse: not much; making ports pushing them. :)07:47
diversenice07:47
_root_diverse: the number is small ATM but I am trying to find out what apps are needed in crux and if I am able to port them.07:50
diverseit's nothing to worry about, just port what you need.07:51
diverseif you get stuck, ask here07:52
Romsterhttp://hvlinux.net/linux/ports/hvlinux/core/prt-get/prt-get-download.diff08:02
Romsteri forgot i had that patch08:02
Romsterlooking though my old stuff contemplating removing obsolete stuff. and yeah.08:04
*** chris2_ has joined #crux08:08
diversemaybe I could port prt-get next ;)08:11
Romsterhehe08:11
*** frinnst_ has joined #crux08:11
Romsterdiverse, i always wanted logging in pkgmk than in prt-get it's a pain in prt-get if you wanna use pkgmk directly.08:11
Romsteri haven't compiled metapkg yet but i glanced at the code earlier08:12
diversewell the "add" subcommand is incomplete atm, because I have no way to change entry paths08:13
Romsteroh08:14
*** frinnst has quit IRC08:14
*** chris2 has quit IRC08:14
diversebut test it anyway08:14
Romsteryeah i'll spin up docker and throw it in and test it out.08:15
Romsterbe cool to extend it where needed and refactor what needs doing to make it the ultimate package manager08:15
diverseas you can see, I already did some refactoring, by splitting pkgutil_t into Db and Pkg, and have them communicate with themselves instead having to do redudant work.08:17
diverseand added extra stuff08:18
Romsterhttps://gist.github.com/anonymous/b7234da3aa1a19c7fd68 my old pkgutils port i was using in 2010 on hvlinux. might get some ideas.08:19
Romsteryeah i like how you split that08:19
Romsteri had a mirror:// syntax i added that ages ago.08:25
Romsterand pigz support.08:25
Romsterused curl instead of wget hmm what else08:26
Romsterdetection if it downloaded a html file instead of an archive.08:28
Romster# if file is text and has archive extension it's wrong08:28
stdevelMorning08:30
Romstermorning stdevel08:30
stdevelDoes anybody know how much space it would need to install all the CRUX packages from core, opt, contrib?08:30
Romsterthe controversal sha256 sums08:30
stdevelI want to build a full CRUX-ARM mirror for RPi08:30
Romsterhmm and the variables for cflags cxxflags ldflags cc cxx per a port.08:32
Romsterstdevel, not a lot probably around 1GB or so if you included contrib opt xorg08:32
stdevelAh cool08:32
Romsterbuilt packages that is source would take more.08:33
stdevelYeah that's what I meant - I want to build binary packages for everything in core opt contrib xorg and generate a repository08:33
joacimhmm. might need more than 1GB. texlive and such is pretty big08:34
stdevelI remember that there are some very huge packages like glibc and gcc08:34
stdevelI had to adjust the inode sizing for my SD card so that the glibc source code was able to unpack :D08:34
Romsterdiverse, yeah ^ pretty much what i had been doing.08:35
Romsteri'd still would love a SMP compressor. for .tar.xz haven't bothered to look for one.08:36
Romsterwhen i was using pigz instead of gunsip it made a hell of a difference on large packages.08:36
Romsterstdevel, well you may want to collaborate with prologic and me on that we are looking at a build setup in docker to build packages. although we are targeting x86_6408:37
Romsterman it would be cool if curl was used on pkgutils as per alncos patch and add libmetalink so meta4 files work. would solve all the downloading issues.08:40
Romsterprobably not kiss enough for crux though.08:40
Romstercurl supports metalink version 3 and metalink version 4 now,., if built with libmetalink08:41
stdevelThat sounds interesting, Romster08:42
frinnstglibc and gcc aren't that big08:42
Romsterno core is like 200MB or so with some opt ports.08:43
Romsteron the ISO08:43
stdevelDo you have a environment/toolkit ready for that? I was just thinking about to write something in Python to automate all this and include it in my Icinga monitoring :)08:43
frinnstqt, texlive, firefox, libreoffice and so on are quite big though08:43
Romsterbut when you include contrib and all of opt then yes08:43
Romsteruh texlive might be over 1GB alone iirc08:44
stdevelCompiling Firefox took 3h on my i7 quadcore - and needed 3 GB of memory :D08:44
Romsteronly 3gb :P08:44
stdevelThat will be a pain in the a.. on the Rpi :D08:44
Romsterwhy would you build all that on a Rpi anyways?08:45
stdevelI don't know how to set-up a cross-compiler for this08:46
stdevel:D08:46
Romsterpitillo, and sepen look after crux ppc or was it crux arm i can't remember now but last i looked its in hiatus mode08:46
Romsterah08:46
Romstercrosstool-ng08:46
Romsterin my repo romster/08:46
stdevelAh okay08:47
stdevelNeed to have a look at this08:47
joacim3 hours? heh. takes about 20 minutes on my i5. about an hour, maybe two on my c2d08:47
Romsteror go play with linux fro scratch crosstool from scratch08:47
stdevelWould be much better as I have a server equipped with a Intel Xeon :)08:47
Romsterfro/from08:47
Romsteryou can't run it but you can compile on another arch08:47
stdeveljoacim: Compiled on natively or inside a VM? I did it in a VM which had 2 vCPUs08:47
Romsterunless you vbox it perhaps.08:47
joacimoh. natively. compiled using all 2-4 cores available on my c2d and i508:48
stdevelOkay08:48
pitillocheck differences between cross built binaries and native ones :)08:49
Romsternot the same pitillo ?08:49
stdevelYeah I should do some tests and diff the packages08:49
nwewhich program can I use to burn iso to cd-r ?08:50
pitillojust for curiosity, take a look to them08:50
stdevelIf diff/md5sum matches everything is allright :)08:50
Romsterif it was me i'd crosstool binutils gcc and needed libraries in it's own prefix.08:50
pitilloit won't be equals stdevel, you'll see libs hardcoded in binaries from your host08:50
Romsteror just run a VM and do it all in there.08:50
Romsterif not on native hardware08:50
nwegrowisofs can I use to burn iso-> DVD08:50
stdevelOkay08:51
Romsternwe, i use wodim08:51
stdevelLooks like this will be an interesting project for my holidays ;)08:51
pitilloARM VMs are pretty slower than native builds08:51
frinnstnwe: wodim08:51
frinnst(cdrecord port)08:51
Romsternwe, but i do use growisofs to make the DVD iso first.08:52
stdevelOn the other hand it would no problem to have a RPi running 24/h. It's not very power consuming ^^08:52
frinnsts/port/fork/g08:52
pitillowe used cross only when started... curretnly ARM devices are powerfull... and with some of them.. distcc and you can cut down times08:52
nwethanks guys then I will use wodim08:52
Romsteri'm still using cdrkit08:52
frinnstyeah wodim is part of cdrkit08:52
Romsterthats a dead project i should look at a alternative.08:52
Romsterbut lazy too busy08:53
Romsterand it works as is08:53
frinnstindeed it is. but the alternative is cdrecord and schilling is insane08:53
Romsteryeah do not like that alternative either.08:53
Romsterwhy did cdrkit die :/08:53
Romsteroh yeah distcc <308:54
Romsterpitillo, so you distcc a lot too08:54
Romsterfrinnst, if you ever find something else for optical burning please package it08:56
Romsterthat is not a insane dev08:56
joacimWe don't like insane devs? How about some TempleOS08:58
*** titanmech has joined #crux09:08
*** Feigrim has quit IRC09:13
diverseRomster: most of what you did in your refactor has already been refactored in my way and Rust defaults to immutable types by default09:14
diverses/types/values/09:14
diverseI'm pretty tired, gonna get some sleep.09:14
stdevelbrb09:16
*** Feigrim has joined #crux09:21
jueRomster: still waiting for your openssl enable-ec_nistp_64_gcc_128 ticket, shall I do it for you?10:01
Romsteri'll do it now jue10:11
Romsteri have a bunch of things to do this weekend.10:11
Romsterjust been too busy10:11
jueRomster: many thanks10:15
*** SiFuh_ has quit IRC10:18
Romsterhttp://crux.nu/bugs/index.php?do=details&task_id=109610:18
Romsterdone10:18
*** SiFuh has joined #crux10:19
*** nilp has quit IRC10:27
pitilloRomster: not really, I'm patient, but hlavery who has some devices is making distcc and minimizing builds times a lot (I did in the past with two efikamx smarttops)10:29
Romsteryeah distcc makes a huge difference.10:30
*** leo-unglaub has joined #crux10:31
leo-unglaubgood morning!!!!10:31
pitilloyes it does10:31
leo-unglaubisn't live awesome?10:31
pitillomorning leo-unglaub10:32
leo-unglaubfrinnst: can you please upgrade perl from 5.18 to 5.20.1 ?11:01
*** hhhhhhhh has joined #crux11:08
jueleo-unglaub: sorry, no, we will do this with the next crux release11:10
leo-unglaubjue: okay, may i ask why?11:11
leo-unglaubincompatible changes?11:11
juethat's not the reason, but we must rebuild every p5-* port because of the changed directory layout11:13
juethink at /usr/lib/perl5/site_perl/5.18/ ...11:14
leo-unglaubah, i seee11:14
leo-unglaubwell, thats a perl fuckup in my opinion, why don't they just use perl5/ ?11:15
juethat's the same as with other interpreters like python, ruby etc.11:16
leo-unglaubhehe, understood11:16
leo-unglaubthen i try to work around that perl5.18 bug that currently crashes my program11:17
jueit has the advantage that you can install several versions parallel11:18
leo-unglaubyeah, but having multiple versions of the same thing on one computer usualy just means that the codebase is to fucked up to be ported to the next release11:18
leo-unglaubi am not a fan of that11:18
jueme neither, you've reported that bug upstream already?11:19
leo-unglaubyes, and the first response at 4 at night was that i should try it with the new perl version *g*11:19
leo-unglaubthen i asked if they mean perl611:20
jue:)11:20
leo-unglaubthen an entire flame war started *g*11:20
joacimgood job, troll11:23
*** crash_ has joined #crux11:31
*** crash_ has quit IRC11:31
*** crash_ has joined #crux11:31
*** crash_ has quit IRC11:31
*** crash_ has joined #crux11:32
*** Feigrim has quit IRC11:45
*** doomicide has joined #crux11:48
*** Feigrim has joined #crux11:53
*** _root_ has quit IRC11:54
*** xeirrr has joined #crux11:54
*** jdolan has quit IRC12:04
*** Feksclaus has joined #crux12:04
*** crash___ has joined #crux12:06
*** crash_ has quit IRC12:08
*** jdolan has joined #crux12:24
*** jdolan has quit IRC12:29
*** titanmech has quit IRC12:30
*** BitPuffin has joined #crux12:41
*** xeirrr has quit IRC12:42
*** jdolan has joined #crux12:57
*** jdolan has quit IRC13:04
*** BitPuffin has quit IRC13:04
*** BitPuffin has joined #crux13:05
*** jdolan has joined #crux13:06
*** crash___ has quit IRC13:14
*** u7knv9h has joined #crux13:21
u7knv9hhello, i have a ""problem"" with acpi and can't figure how to solve this: i'm configuring the events and the handler for the volup/voldown buttons, but i can't keep them pressed to lower/raise the vol (to keep lowering/raising the volume i need to repeatedly press the buttons, which is annoying)13:25
u7knv9hanyone can help me?13:25
*** jdolan_ has joined #crux13:29
*** jdolan has quit IRC13:29
EmoSpiceu7knv9h: I don't know of a solution to the problem (as Xorg seems to respond to the keypress/keyrelease, not the keypress duration), but I'd be interested in hearing when / if you find a solution13:29
EmoSpicewell, ACPI not Xorg13:30
EmoSpiceYou might have success if you were to code up a small shell script that tested the key state every second and had it fire the volup/down logic if the button is still down.13:32
u7knv9huh, maybe i'll do this13:37
u7knv9hbut with xorg i used to bind keys with sxhkd and i had no issues13:37
EmoSpiceinteresting13:37
joacimwhat does acpi_listen say when you hold the key?13:38
u7knv9hit writes the button name once13:40
joacimif there is a different event for button release?13:40
u7knv9hnope13:41
u7knv9hthe same goes with every button i configured with acpi13:41
u7knv9hbrightness up/down have no issues since they worked oob13:42
EmoSpiceI suspect that sxhkd simply polls for which keys are pressed after a small interval13:42
EmoSpicewhich acpi doesn't do13:42
joacimhttp://f.cl.ly/items/2q3d3l2z1C271D0k2D2C/Screen%20Recording%202014-12-12%20at%2002.42%20pm.gif13:43
joacimmaybe you should buy a mac!13:43
EmoSpiceThat's totally the fix!13:43
joacimyeah!13:43
u7knv9hmaybe it's relevant, but i use only one event for everything, which points to a handler with every button action (someone in this channel showed me this, maybe jaeger but i really don't remember)13:43
EmoSpiceu7knv9h: That's a pretty standard approach13:44
EmoSpiceand I don't really think it's the cause of the problem13:44
u7knv9huh, i thought the standard was separate events and scripts for every event13:45
u7knv9hanyway i'll buy a mac, to fix all this problems!13:45
EmoSpiceI know that at least Arch uses the one you're using.13:45
EmoSpiceBecause of...science...definitely not because I'm cheating on Crux with Arch at work >_>13:46
u7knv9hoh well, i thought that arch used separate scripts after seeing this13:46
u7knv9hhttps://wiki.archlinux.org/index.php/acpid13:46
u7knv9hy-you're cheating on crux f-for systemd?!13:47
EmoSpice"By default, all ACPI events are passed through the /etc/acpi/handler.sh script. This is due to the ruleset outlined in /etc/acpi/events/anything:"13:47
u7knv9hyou're right lol13:47
EmoSpicenot FOR systemd. It's an unfortunate consequence13:47
u7knv9hi should have read with more attention13:47
EmoSpiceI'd switch to https://github.com/hut/minirc if I felt I could get away with the downtime13:48
EmoSpiceu7knv9h: http://stackoverflow.com/questions/3649874/how-to-get-keyboard-state-in-linux13:50
*** doomicide has quit IRC13:54
u7knv9huh13:54
*** jdolan_ has quit IRC13:55
u7knv9hbut the strange thing is that 'long press' with brightness buttons works13:55
u7knv9hand with every button configured in the handler doesn't13:55
u7knv9h(the brightness ones worked oob)13:55
EmoSpiceAgain I suspect that acpid (which I suspect your brightness buttons are sidestepping) is only reporting keypresses.13:56
rmulljaeger: I think my dbus isn't working right - is that something you can help me with?14:04
rmullMy first question is about the .footprint - it puts configuration files in /usr/etc instead of /etc - is that correct?14:05
rmullI installed pulseaudio and bluez and they install configuration files into /usr/etc/dbus-1/system.d, then I restart the dbus init script and also do the killall -HUP dbus-daemon but I can't seem to send messages to org.bluez for example using dbus-send14:09
jaegerI'm no expert on dbus, I only use it because other things require it. Haven't tried to send it commands on my own. Yes, /usr/etc is correct14:09
rmullIs restarting the rc script sufficient to get it to load up the new config? Is there any way to "dump" the current set of org targets?14:14
rmullI saw the d-feet software that lets you visualize the dbus network but I don't want to get that deep into it14:15
jaegerNo idea on that but I would guess there's some easy way14:15
u7knv9hEmoSpice: so it's not possible at all?14:18
EmoSpiceu7knv9h: No. I suspect it is, but it's probably more effort than it's worth, if you can just do it from sxhkd14:19
rmulljaeger: Okay, thanks. I'll keep poking at it14:19
u7knv9hyea, i'll use sxhkd for now. if i'll find a way to do it directly with an acpi event i'll do it, but for now it's ok to have sxhkd14:21
u7knv9hother acpi buttons doesn't need to have 'long press' and works just fine14:22
joacimrmull: dbus usually picks up on that stuff on its own. had my own problems with avahi earlier today14:29
joacimit installed those system.d files to /etc, but can only find them when they're in /usr/etc14:30
*** DaViruz has quit IRC14:31
*** sh4rm4 has quit IRC14:31
*** jdolan has joined #crux14:32
*** DaViruz has joined #crux14:44
hhhhhhhhrandom thought of the day14:44
hhhhhhhhshould we add support to pkgmk for passing in user specified flags to configure, without having to make a new package?14:44
hhhhhhhhit would save me from having to keep a vim-lua package up to date when all i want is --enable-luainterp passed to configure14:45
prologicno14:46
prologicthe problem is too complex14:46
prologicihmo14:46
prologicthere are too many various ways to orchestrate packages from ports14:46
hhhhhhhhi disagree on it being too complicated14:47
*** u7knv9h has quit IRC14:47
hhhhhhhhwould need to add "./configure ... $shit" to every pkgfile14:47
prologicyou'd have to support configure, C/CPP flags, CMake, python distutils/setuputils14:47
prologicthe list goes on14:47
prologicyou're only thinking about ond kind of build tools :)14:48
prologicthere are dozens and dozens14:48
hhhhhhhhand prolly 10 lines to pkgmk to read from /usr/shit/vim.txt and drop it in14:48
prologicthe simplest solutions have always been to14:48
prologica) keep your own cloned port(s)14:48
prologicb) use revision control14:48
prologicI'm not even sure how Gentoo folk implement this14:51
prologicUSE env var as such14:51
prologicperhaps they only support USE flags for a limited no. of build systems14:51
prologicmaybe just autotools/configure14:51
hhhhhhhhlooks like their build scripts check for USE flags and do the right thing by han14:53
hhhhhhhhd14:53
hhhhhhhhwhich puts all the complexity on the packager14:53
prologicyay :)14:53
prologicmore work for us port maintainers14:53
hhhhhhhhi am suggesting something similar but simpler14:54
prologicit's enough work keeping up with broken upstream sources and mirrors14:54
prologicand shitty ass services like bintray14:54
hhhhhhhhinstead of 50000 if blocks, just have a "user can insert garbage at this point" marker14:54
prologicactually that could work14:55
joacimi think this would complicate the pkgfiles14:56
prologicactually hold on :)14:56
prologicI think it could be quite simple14:56
prologicoptional14:56
prologicand complexity left up to the the maintainers14:56
hhhhhhhhyea. packagers could just add $flags to the end of the configure line14:56
prologicin fact14:56
joacimi like the idea, but i think it would be too inconsistent as I assume the best way to do it is to make all packagers write their pkgfiles with support for that kind of stuff14:56
prologicthere's nothing we need to do14:56
hhhhhhhhif they wanted to14:56
prologicit can be done already14:56
prologicwe just need a convention14:57
prologice.g: ${BUIDLOPTS}14:57
prologicwhich is up to the disgression of the maintainer as to how to use14:57
prologicBUIDLOPTS=--enable-lura prt-get depinst lua14:57
jaegerthe big difference is that gentoo's portage doesn't use footprints14:57
joacimIt would be a really inconsistent experience if some Pkgfiles supported $flags and some didnt14:57
jaegerit can't, considering the vast discrepancies that USE flags can cause14:57
prologicI agree joacim14:58
prologicbut still the idea is rather simple14:58
prologicpkgutils doesn't need to change14:58
prologicwe do :)14:58
hhhhhhhhsome pkgmk tweaks would help enforce some consistency though!14:58
joacimprt-get could help by exposing ports that supports buildopts14:58
joacimletting the users know when they can pass their own flags14:59
prologicahh yes14:59
prologictake the Homewbre route14:59
prologicthe way they do it is rather nice14:59
prologicand is agnostic of the actual build system used to build the software (I think)15:00
prologicbut basically yeah a) tells the user up front what they can add/disable/etc15:00
hhhhhhhhfreebsd does something like that too and i actually don't like it15:03
hhhhhhhhbecause it either makes the pkgfiles horrrrrrrrrrrrrrrrrrrrible15:03
hhhhhhhhor you have to write some other file listing all the options and their effects which will be horrible15:04
prologicyoud just use simple env vars ihmo15:04
prologicwith simple bash/sh boolean conditions15:04
prologicto keep the Pkgfile(s) clean15:04
hhhhhhhhbut with packages like vim and mutt you could end up with 30 lines of conditions15:04
prologicif [ $use_foo -eq 1 ]; then15:04
prologic...15:04
hhhhhhhhor, you don't cover anyhting and ppl have to fork their own packages anyway15:05
joacimif the packager doesn't list all the options the configure script supports, the users would have to unpack the source and run ./configure --help themselves15:05
prologicinded15:05
joacimso i think gentoo use flags might be easier and simpler for the user15:05
prologicit's the same as it is for HOmebrew Formulae15:05
prologicpackage use their best judgement to provide options15:05
prologicand user can use them15:05
prologicyou can't satisfy everyone :)15:06
hhhhhhhhwith the method i am suggesting, it's very possible for them to shoot themselves in the foot by passing in trash15:06
joacimkinda like #Nice to have:15:06
hhhhhhhhbut the maintainer doesn't have to think about it15:06
joacimA lot of Pkgfiles already have that #nice to have line15:06
prologicI'd still opt for the Homebrew style15:06
prologic--enable-feature15:06
prologicand list them15:06
*** kori has quit IRC15:09
*** kori has joined #crux15:13
*** u7knv9h has joined #crux15:15
j^2Hey Romster I'm still around :) just got like 20 channels open nowadays15:16
*** titanmech has joined #crux15:31
*** c0ck4m0u53 has joined #crux15:40
*** kori has quit IRC15:43
*** kori has joined #crux15:45
*** Workster has quit IRC16:16
*** mhe has quit IRC16:18
*** leo-unglaub has quit IRC17:00
rmullwould be kinda nice to handle optional deps better17:01
rmullsomething like prt-get options package could print out some optional packages that enhance the features of package17:01
rmullthen the user could just depinst the options that they want17:02
rmullI guess that kinda sucks17:02
rmullnever mind17:02
rmullwhen I used gentoo, I had very few system-wide USE flags, but I had tons of per-package USE flags17:03
rmullI micromanaged it17:03
*** phant0mas has quit IRC17:20
joacimyeah useflags seemed like a good idea, but they became a chore after a while. i spent way too much time just sorting through them17:23
u7knv9hmeh, i never liked use flags too much, but i liked the idea of not having something you don't want using them (-systemd, yo)17:30
jaegerI really like use flags. I was a fulltime gentoo user before crux17:30
u7knv9has long as i used -useflag -anotheruseflag i was fine, but i felt really uncomfortable using (+)useflag (+)anotheruseflag17:32
joacimI had to force myself to use as few as possible17:32
*** hhhhhhhh_ has joined #crux17:39
*** hhhhhhhh has quit IRC17:39
u7knv9hcrux's ports system is much more simple and cleaner17:40
u7knv9hmaybe a little bit harsher, but definitely more simple17:41
korirmull: have you ever used FreeBSD ports?18:08
korithose have a menu where you can disable or enable features18:08
koriits quite nicde18:08
korinice'18:08
*** BitPuffin has quit IRC18:25
*** avalon_ has joined #crux18:37
*** _root_ has joined #crux18:37
_root_Hello people18:39
u7knv9hhello dudo18:45
_root_u7knv9h hello back19:06
*** phant0mas has joined #crux19:09
u7knv9hif i wanted to replace openssl with libressl, what i'd have to do? just remove the `openssl` port and install teh `libressl` port?19:21
rmullkori: No, I have never tried freebsd19:34
*** c0ck4m0u53 has quit IRC19:35
rmullI would totally not object if someone added explicit optional dependency management to the prt utils though19:36
rmullUSE flags were, for me, convenient19:37
rmullJust not sure what the right way to do it is for crux19:37
u7knv9hPkgfile(s) have everything you need, adapt them to your needs19:41
u7knv9husing manual dependency handling can give you even more control on your system19:42
pitillou7knv9h: rebuild all ports depending on openssl and not sure if you'll need to tweak some to use libressl instead of openssl19:55
u7knv9hand, rebuilding them, could be there any flag to set in order to use libressl instead of openssl?19:57
pitillothat's the hard point... you need to review all ports linking with openssl and check compatibility... I never did anything similar19:58
u7knv9hoh well, i think i can wait hahah19:59
pitilloVM or container to make any kind of test... core ports install and check :)20:00
u7knv9hyea, i was thinking for a chroot... but compile times are an issue for me right now lol20:04
u7knv9hi'll wait the christmas holidays20:04
*** u7knv9h has quit IRC21:21
*** leo-unglaub has joined #crux21:51
*** phant0mas has quit IRC22:22
*** hhhhhhhh has joined #crux22:22
*** Feksclaus has quit IRC22:23
frinnstvacation bitches!22:26
frinnstmuhahah22:26
leo-unglaubfrinnst: gz22:26
leo-unglaubvacation or "let go" because you drop the entire webshop db?22:27
*** jdolan has quit IRC23:02
*** hhhhhhhh has quit IRC23:37
frinnstheh, vacation23:39
frinnstthe reindexing might be completed when i return to work :)23:39
*** hhhhhhhh_ has quit IRC23:39
diversefrinnst: nice23:40
prologicgopher://arrow.shortcircuit.net.au/1/crux23:41
*** jdolan has joined #crux23:47

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