IRC Logs for #crux Wednesday, 2014-12-03

koria ports(8) DRIVER.00:00
korilike the rsync and httpup ones.00:00
koriI'm currently using httpup for my git repo and I don't think that's needed...00:00
*** leo-unglaub has quit IRC00:01
diverseat the very beginning I thought your original plan was to make a git driver for ports00:01
*** jdolan has joined #crux00:02
*** kahiru has quit IRC00:02
*** leo-unglaub has joined #crux00:02
diverseleo-unglaub: how did it go?00:03
leo-unglaubdiverse: no difference, still no wlan000:04
leo-unglaubbut i already feel backdored just by downloading that file *g*00:04
diverseleo-unglaub: hmm... you did make sure that you switched on the wifi device, right?00:05
leo-unglaubdiverse: yes00:05
leo-unglaubthe led is even glowing00:05
jaegercheck the dmesg output again, see if there are still firmware loading errors00:05
koridiverse: well, I used the wrong term00:06
leo-unglaubi still miss iwlwifi-6000g2a-exp.ucode00:06
koribut now I want to do both00:06
leo-unglaubbut there is no *exp* file on the tree00:06
jaegerin your paste there were 4 files it listed, you might want to grab all 400:06
leo-unglaubi have 3 of 400:06
leo-unglaubiwlwifi-6000g2a-exp.ucode does not exist00:06
jaegerI only have iwlwifi-6000g2a-5.ucode for what that's worth00:07
diversea git driver for ports would be pretty useful00:07
leo-unglaubi have an idea00:07
jaegerleo-unglaub: you can also unload and reload the module without rebooting to test00:07
*** leo-unglaub has quit IRC00:07
jaegeror not00:08
diversejaeger: mention that to him when he comes back00:08
*** leo-unglaub has joined #crux00:09
jaegeras a side note,
diverseleo-unglaub: you didn't have to reboot, you could have unload and reload the module00:10
diversejaeger: you made one?00:11
jaegerI did not00:11
leo-unglaubdiverse: no, i did build the iwlwifi into the kernel, not as a module ... :(00:11
jaegerIt's someone else's work and looks pretty simplistic but I linked it for reference00:11
diverseleo-unglaub: ah00:12
jaegerleo-unglaub: oh. I somehow overlooked that earlier. rebuild it as a module and I bet it works00:12
diversejaeger: cool, it looks like the work is already done for kori00:12
jaegerbecause it's being loaded when the bootloader loads the kernel, the filesystem and firmware aren't available yet00:12
leo-unglaubWTF, for real??00:13
diversethe question is, who is via?00:13
jaegerleo-unglaub: it's either that or you build the firmware image into the kernel itself00:13
jaegerI recommend just building iwlwifi as a module, easier00:14
diverseI agree with jaeger00:14
leo-unglaubhaha, that sounds like a windows solution *g*00:14
leo-unglaubbut okay *g*00:14
diversehow does make it a module make it windowzy?00:14
leo-unglaubdiverse: i mean the fact that there is no word about this bug in the kernelconfig docs *g*00:15
jaegerI doubt it's considered a bug00:15
jaegerthere's a specific kernel option for including firmware blobs00:15
leo-unglaubhmm, still .. someone could have dropped a line about that into the kernelconfig ... would have safed me 3 evenings *g*00:16
diversejaeger: could we ask via to add his git driver in officially?00:16
jaegera more robust one might be preferable00:17
diverseah I see00:17
jaegerwith --depth 1 and branch/tag checking, etc.00:17
korijaeger: oh awesome00:18
koriwell yes I'll work on that00:18
koriI've never done something similar before so I'm going to ask for some time00:18
koriwell, it's not like this is a priority anyway :P00:18
jaegernor is it a homework assignment :D00:19
diversekori: you can ask Romster to help you out with suggestions too, he is very good with git00:23
*** leo-unglaub has quit IRC00:23
korijaeger: well00:24
koriI guess00:24
*** leo-unglaub has joined #crux00:25
leo-unglaubyeah, and now there is wlan0 :)00:25
leo-unglaubthanks guys00:25
diverseyay, happy wifi'ing00:26
leo-unglaubyeah, now i only have to aircrack-ng the wifi from the guy next door and then i have wifi as well *h*00:27
diverseleo-unglaub: is the wifi n or g?00:27
leo-unglaubnone, it's offline *g*00:28
diverseso much for stealing wifi00:28
leo-unglaubwell, getting wlan0 there was just part 100:29
leo-unglaubnow i have to get that wpa_supplicant crap right00:29
diverseyeah, just be happy you have a working component00:29
leo-unglaub.... wlan and linux does really suck big balls00:29
diverse*ahem* wlan *on* linux, not and00:30
leo-unglaubfor me woth are true *g*00:32
leo-unglaubfuck ... when i spell that wrong .. i should go to bed ... i am already up for 30 hours ...00:33
diversegeez, yeah, get some sleep already00:33
*** leo-unglaub has quit IRC00:34
*** jdolan has quit IRC00:34
*** leo-unglaub has joined #crux00:35
leo-unglaubyeah, i am goint to bed00:35
leo-unglaubsee you later00:35
*** leo-unglaub has quit IRC00:36
diverseI never had much trouble with wifi once I got it working00:37
*** jdolan has joined #crux00:42
diverseteK_: have you worked on the prt-get C porting?00:43
*** Puffin has joined #crux00:59
*** tilman has quit IRC01:03
*** tilman has joined #crux01:05
*** xeirrr has joined #crux01:15
*** jdolan has quit IRC01:23
*** jdolan has joined #crux01:24
Romsterwhen you spell that wrong you've been around me for too long, lol.01:31
diverseah I see, he got your influence01:38
*** BitPuffin has joined #crux01:39
diverseI thought he was just in a bad mood from being sleep deprived01:40
Romstermaybe a bit of both01:45
Romsterhmm thunder01:45
*** BitPuffin has quit IRC02:25
*** Roosevelt has joined #crux02:28
Roosevelthi everyone!02:29
xeirrrRoosevelt, hi02:30
Roosevelthi xeirrr02:31
*** leetspete1 has joined #crux02:31
xeirrrRoosevelt: what's up?02:31
Rooseveltxeirrr, need to apply this diff to kwebkitpart how should this go in Pkgfile? This is what I got now...
xeirrrRoosevelt: you should patch -p1 -i "patchname"02:39
koriyou know02:39
koripatch -p1 < patch02:39
Rooseveltwhat about the diff? The link to the diff should be place where?02:41
xeirrruse diff -uN from-file to-file >to-file.patch to generate patch02:41
xeirrrthe diff should be named with .patch, then you can patch02:42
RooseveltOkay got it!02:43
*** Feksclaus has quit IRC02:43
diversepatching is fun02:44
Rooseveltbuilding kde is fun to02:45
Rooseveltmatter of fact kde built faster with prt-get than emerge...02:47
xeirrrRoosevelt: really? I never use kde in source-based distro02:48
diverseI wonder if because of emerge being written in python? Although I'm doubting that.02:48
diversethe packaging part should be faster, though I don't know if the building part would be faster02:50
xeirrrno idea. Maybe depends on your machines02:50
diverseif you did, say, -j2 on Gentoo instead of -j5 on Crux, then yeah it would be faster on Crux02:52
diversebut only because you enabled more jobs02:52
RooseveltWell I didn't do no benchmarks build On my amd fx 6100 with -j6 was build faster on crux02:56
*** jdolan has quit IRC03:04
*** jdolan has joined #crux03:06
*** Workster_ has joined #crux03:08
*** Workster has quit IRC03:08
*** Workster has joined #crux03:08
diverseRoosevelt: how many cores do you use?03:12
*** jdolan has quit IRC03:13
Rooseveltdiverse, I have 6 cores03:13
*** jdolan has joined #crux03:14
diverseRoosevelt: you would want to do -j7 then03:15
diverse-j`$ncores + 1` is the rule of thumb03:17
Rooseveltdiverse, adding one more will increase build time?03:17
diversejust a little03:17
diversewait, you mean decrease, right?03:17
Rooseveltno increase...will make build time faster?03:18
diversesigh, my interpreting skills03:18
diverseyes, faster03:18
Rooseveltdiverse, doing -j7 in docker is okay?03:19
diversenot sure, I imagine it would be fine?03:19
diverseRoosevelt: ask Romster03:20
*** SirRafiki has quit IRC03:20
RooseveltOkay will. Anyway docker is loads of fun.03:20
RooseveltHaven't used virtualbox since using it.03:21
diverseso for example, I have 8 vcores because of using a i7-4770k, so I build with -j903:21
Rooseveltwhat do you have for your CFLAGS?03:23
diverse"-O2 -march=core-avx-i -pipe" although I should probably get rid of "-pipe" since I'm building in a tmpfs03:24
nbinowhy not -march=native ?03:26
RooseveltI always wondered that myself?03:26
nbinoif is only to run on the host machine is any problem ?03:27
diverseI heard gcc is having problems with the core-avx2 option in 4.8, where it actually builds slower binaries, so I want to wait for that to mature before using native03:27
diverseso I'm building ivybridge binaries on a haswell because gcc is not up to snuff yet03:29
RooseveltI never bother with tmpfs since I assumed doing so, just takes away from ram?03:29
nbinodiverse: sorry the off-topic, how can I override the globals folders of pkgmk03:30
diversenbino: in /etc/pkgmk.conf03:31
nbinosorry my english, I want to specify diferent folders /dist/work/packages when building some packages03:31
diverseRoosevelt: depends, how much ram do you have?03:31
diversenbino: it's in the config, you want to edit the PKGMK_WORK_DIR var03:32
nbinofor example, pkgmk --packages ~/different_from_etc_config03:32
diverseRoosevelt: yeah I'd probably save the ram and not use tmpfs03:32
*** mavrick61 has quit IRC03:34
RooseveltI would probably need somewhere around 16GB to become effefient03:34
*** mavrick61 has joined #crux03:35
diversenbino: I don't think that option exists, might want to request that feature03:36
*** SirRafiki has joined #crux03:36
diversenbino: it's a good idea for a feature, none the less.03:36
nbino:) I'm shy to ask,03:37
SirRafikihey bino :D03:37
nbinohey hey :D03:37
diverseRoosevelt: I currently have 32GB of ram, so I don't worry about using up all my memory, so tmpfs ahoy!03:37
diversenbino: don't be shy, ask one of the developers03:38
nbinoyes, I will see on git wich one is, I was thinking you were one :)03:39
diversenope, I'm a crux user03:39
diversejaeger: is there a way to override the pkgmk's work directory path temporarily via arguments?03:43
diverseif not, can we add one?03:44
WorksterRoosevelt, 4 cores -j4 the number of cores plus 1 is only an advantage over tcp like distcc etc.03:46
Workstertoo many jobs will slow it down also03:47
Worksterdocker same jobs as number of cpu cores03:47
Worksternot all compiles use all the jobs effectively03:47
diverseah I see03:47
RooseveltOkay I will change my docker to -j4 so I don't slow my computer down.03:49
jaegerdiverse: temporarily? what for?03:49
diversejaeger: during the one instance of using pkgmk03:50
diversenbino would like to specify a workdir only one time without having to edit the config file03:50
jaegerI get what you want, just wondering why03:51
jaegerseems odd03:51
diverseask nbino03:51
diversejaeger: there is probably a reason for it, I wouldn't say it's odd03:52
*** jdolan has quit IRC03:52
*** jdolan has joined #crux03:53
nbinojaeger: I was thinking to play with core packages, where I just copy the /usr/ports/core to ~/test . then in each I wold call pkgmk --packages ~/new_core, this way I can then pkgadd in a loop inside newcore03:53
nbinothere is other ways to do it I'm sure,03:54
diverseoh I see, you want to recursively build all ports in a directory03:56
diverseand then you would recursively add all the packages in said directory?03:57
nbinoyes, pkgadd to diferent target also03:58
diversejaeger: ignore what I said about workdirs earlier03:59
nbinoso I can try updates in a chroot, then inside that chroot rebuild everything for second time to be sure it replicates :)03:59
nbinoits odd ?03:59
nbinoI have to read more :)04:00
diverseno, it's different than what I thought earlier04:00
diversea misinterpretation04:00
jaegerchroots are easy but yeah, different thing04:00
nbinoI'm bad speaker, realy sorry04:00
jaegeryou can pkgadd -r into a chroot04:00
jaegerthen chroot into it and work like you work normally04:01
nbinobut in first pkgmk, then i have to search in /usr/ports/packages for the ones,04:01
nbinoI can change /etc/pkgmk.conf temporary04:02
jaegerchange PKGMK_PACKAGE_DIR in /etc/pkgmk.conf04:02
jaegerfor the first build if you want04:02
jaegeror just find . -name "*.pkg.tar.*", easy04:02
diversemany ways to skin a catfish04:03
nbino:D yes, and thanks ;)04:03
nbinois any one using gitolite in crux ?04:04
nbinoI was thinking it was a "C" thing, but for the fast reading I made I think is perl script,04:04
nbinodocumentation says to ln to bin, but where to put the sources, in usr/share04:05
*** jdolan has quit IRC04:06
*** jdolan has joined #crux04:06
*** SirRafiki has left #crux ("Saindo")04:33
WorksterRoosevelt, if you want more speed compile in tmpfs and or use a different cpu scheduler BFQ if your keen on patching your kernel05:04
Worksternbino, pkgmk -r recursively compile05:04
Worksternbino> is any one using gitolite in crux < yes me05:05
*** GuY` has quit IRC05:06
Worksternbino, but i have installed that like ages ago and don't remember exactly, make a git user and install to /home/git/05:06
*** jdolan has quit IRC05:18
*** jdolan has joined #crux05:18
*** xeirrr has quit IRC05:20
nbinoWorkster: I just finish make a port,05:23
Worksteri just used git to grab it and installed it from that05:23
Worksteri'd look at that port though and want to red mine in a docker image sometime. if it's hosted to see it.05:24
Worksteri'm guessing you'll do like me and use a git hook to send to gitolite and generate the REPO/html05:25
*** GuY` has joined #crux05:26
nbinoWorkster: I don't have yet setup my ports, I will try to do it asap05:27
nbinoI don't understand what you mean about the hook05:28
*** GuY` has quit IRC05:30
Worksteri'll show you when i'm home in about an hours time05:42
*** sudobaal has joined #crux05:43
Worksteri use it to copy repogenerate ...05:43
Worksteron the server05:43
Worksterso when i git push on my desktop it's got everything regenerated05:44
*** Lilie34TY has quit IRC05:59
*** GuY` has joined #crux06:02
*** _root_ has quit IRC06:02
RooseveltHey need with setting up this patch here how should I write webcpp.path in the Pkgfile?06:02
Worksterpatch -p1 -i $SRC/whatever.patch06:04
Worksterand add whatever.patch to source=()06:05
Workstermaybe do patch -d ../$name-$version -p1 -i $SRC/whatever.patch06:07
Worksterif you want to avoid adding an extra cd to the source to just apply the patch06:07
WorksterRoosevelt, ^06:08
RooseveltI'm trying it out now to see if that works06:08
Worksterit will if the patch applies cleanly06:11
RooseveltI tried it in the /src/ directory and didn't had to no problems06:12
RooseveltThanks Workster patch -d worked for me.06:20
*** xeirrr has joined #crux06:30
xeirrrHi,how to tell ports to sync my local repo?06:31
tilmancreate /etc/ports/myrepo.rsync06:31
tilman(or myrepo.httpup of course) ;)06:32
xeirrrYeah, I did. How to set host and collection? I tried twice, but failed06:34
xeirrrJust a local dir06:35
tilmanxeirrr: i'd do this another way06:35
tilmantilman@brimstone [~] > grep prtdir /etc/prt-get.conf06:35
tilmanprtdir /home/tilman/devel/crux/ports06:35
tilmantilman@brimstone [~] >06:36
tilmani totally faked that shell session, but you get my meaning06:36
tilmanto use the rsync driver, you really have to have an rsync daemon running somewhere06:37
tilmanso it doesn't apply in your case06:38
xeirrrI must ensure it works before push it into ftp :)06:38
Romsterwho uses ftp anymore06:54
Romsterespecially for port syncing06:56
xeirrrcollege :)07:00
Romsterthey should know better07:00
Romstersamba share07:00
Romsterand http07:01
*** GuY` has quit IRC07:02
*** GuY` has joined #crux07:03
xeirrrRomster: Colleges in my country still use ftp to host stuff, like comics, animes..07:04
Romster[crux-reports] Potentially broken source download07:04
Romsterits like everything broke so the net died at when it did those checks07:04
Romsterwell lets hope they don't try to password protect any of it07:05
*** Pingax_ has quit IRC07:28
*** GuY` has quit IRC07:33
*** Pingax_ has joined #crux07:33
*** Pingax_ has quit IRC07:33
*** Pingax_ has joined #crux07:33
*** vlnx has quit IRC07:59
*** GuY` has joined #crux07:59
*** vlnx has joined #crux08:03
*** _root_ has joined #crux09:32
*** mhi^ has joined #crux10:04
teK_diverse: no I havent =)10:08
teK_I tried to make use of protobuf (to be precise it's inofficial C bindings) without success :)10:09
mhi^I like the subtle Devuan reference. :)10:24
teK_hehe, yeah10:24
*** Feksclaus has joined #crux10:44
cruxbot[core.git/3.1]: libtool: update to 2.4.411:01
diverseI guess the libtool update didn't turn out to be so bad11:23
Romsterwe'll see spins up another docker container and makes world11:28
*** _root_ has quit IRC12:02
*** _root_ has joined #crux12:02
_root_any package you give could suggest me to port?12:02
*** phant0mas has joined #crux12:04
_root_*/give/guys (an example of auto-correct script)12:04
dxlr8r will Crux ever become rolling release?12:14
diversedxlr8r: no12:32
diversehaving it fully rolling will kill the point of Crux12:33
*** u7knv9h has joined #crux12:34
*** phant0mas has quit IRC12:35
*** phant0mas has joined #crux12:36
teK_I disagree12:37
teK__root_: it would be nice to have a current kde but beware taking this task lightly it's a crapload of work12:38
diverseteK_: okay, why?12:39
diverseI thought the point of regular release was to ensure a stable core toolchain for building the entire system without breakages12:41
teK_I don't see any point of crux violated. It's tradition and it's easier/more stable to do upgrades offline12:41
*** hhhhhhhh has joined #crux12:42
diverseteK_: I think you're confused by the term "rolling"12:42
*** sglob_gemschl has joined #crux12:48
diverseteK_: rolling (like Arch) means never having to use a livecd to go offline to do OS upgrades, in which breakages can and do happen while running.12:51
_root_I want pkgmk to do all the jobs and create a package but not to install it!12:51
diverse_root_: then just pkgmk it, but don't pkgadd it12:52
_root_diverse: pkgmk -i?12:52
*** xeirrr has quit IRC12:52
diverse_root_: no -i either12:52
*** hhhhhhhh has quit IRC12:52
diversejust "pkgmk"12:53
_root_diverse: so no switch at all; anything like dry run ?12:53
*** mhi^ has left #crux ()12:54
*** xeirrr has joined #crux13:01
xeirrrHi, I've ported fcitx(chinese input framework) to Crux Linux, lol >
diversexeirrr: "For x86_64 by default, i686 maybe have carefully see Pkgfile. :)" -> "Default for x86_64, could work for i686, see Pkgfile :)"13:06
diverseshorter and concise13:07
xeirrrdiverse: Thank you. my grammar is not good ;P13:07
diverseIt's all good13:08
xeirrrThis is my first time using igt13:08
diversexeirrr: I just wanted to make it easier to read ;)13:08
xeirrrright, I'll change right now ;)  btw, file totally mess up ;(13:09
diversexeirrr: want me to grammer check?13:09
xeirrrThanks, that wuld be much helpful ;)13:10
tvaalenYou should also not keep the source and the built package in the port directory you publish.13:11
xeirrrtvaalen: thank you, but Chinese guys are always lazy, they want prebuilt packages. For source codes, I'll remove ;)13:12
teK_diverse: that was what I was talking about13:14
diverseteK_: alright, well there are advantages and disadvantages to the way arch does it.13:20
*** BitPuffin has joined #crux13:20
diversexeirrr: I'm editing it my vim editor, you really need to newline a lot of stuff13:20
diverseI assume the "$number>" means the step right?13:21
xeirrrdiverse: I'll do.13:21
diversexeirrr: don't worry I'm doing it for you13:21
xeirrrdiverse: I'll shorten readme.MD13:21
diversedoing that too13:21
diversexeirrr: does 0> mean step 0?13:22
xeirrrdiverse,you helped me a lot, man13:22
xeirrrdiverse: change as you like. ) means from groud up13:22
teK_diverse: I simply need to reserve more time to boot the ISO on my root server vs. running 3.0 3.1 && revdep && fix_things()13:23
*** henesy has quit IRC13:24
*** jdolan has quit IRC13:28
diverseteK_: alright, I think I done as much as I can, I'll give you a gist paste13:39
diversexeirrr: ^13:39
diversesorry teK_13:39
xeirrrdiverse, ;)13:39
diverseI removed a lot of cruft13:45
cruxbot[contrib.git/3.1]: [notify] thunderbird: updated to 31.3.0. Fixes various security issues:13:45
diversedangit github still linewraps it13:48
*** leetspete1 has quit IRC13:49
diversexeirrr: view the raw13:49
xeirrrdiverse, Thanks,man.13:49
diverseyou will see what I did exactly in the raw13:49
xeirrrdiverse, can i just copy and psate? would it be messed up again.13:49
diversexeirrr: for now yeah13:50
xeirrrOk,I'll do13:50
diversexeirrr: do you like what you see?13:50
xeirrrdiverse, yeah, much more tidy13:51
xeirrrand ordered13:51
xeirrrdiverse, someone told me This is packaging , not port. so we need to change first sentense13:52
diversesure go for it13:54
xeirrrdiverse, Thank you again :)13:55
diversexeirrr: actually hold on I need to fix it again, I realized the github line wrapping screwed me up originally.13:56
xeirrrokay, I am still editing :)13:57
diversexeirrr: okay, what other changes you made besides the first sentence?14:03
xeirrrdiverse, nothing ;)14:03
diversealright, stay tuned for my latest fix14:05
xeirrrdiverse, yes!14:06
diversefor the first sentence I have: "This project is for the fcitx packages to Crux Linux x86_64." is this fine?14:06
xeirrrdiverse, yes!14:07
jaegers/to/for/ perhaps14:07
xeirrrfor is better ;)14:08
*** jdolan has joined #crux14:11
xeirrrThanks all the guys ;)14:18
*** u7knv9h has quit IRC14:22
*** quasinoxen has quit IRC14:23
*** asie has joined #crux14:24
diversesigh, I'm really tired of dealing with githubs formatting...14:27
*** sh4rm4 has joined #crux14:28
*** quasinoxen has joined #crux14:28
diverseI don't know if I made things worst or better14:28
diversebut be sure to compare14:29
diversexeirrr: copy from the raw14:30
diversenote to self: always copy from raw14:34
*** xeirrr has quit IRC14:35
*** xeirrr has joined #crux14:35
diversexeirrr: did you get my latest gist?14:36
xeirrrdiverse, yes and I just finished git push14:37
*** Feksclaus has quit IRC14:39
diversexeirrr: if there was only a way to fix the formatting on the readme in github14:40
xeirrrdiverse, I see READ me still have display issue, I'll fix it tomorrow. Here is already 22:39 now. I need a sleep ;)14:40
diversexeirrr: I know, I been fighting with github on that14:40
xeirrrdiverse, come on, man.14:40
xeirrrI need quit now14:41
xeirrrgood night14:41
diverseof course, have a good one14:41
diverseI don't mean to keep you up, I'm just saying it was a struggle14:41
*** xeirrr has quit IRC14:41
dxlr8rdiverse: yeah, taking the distro offline to update is the biggest complain I have against Crux14:45
*** chris2 has quit IRC14:49
asiei want to do an experimental crux fork to look into more lightweight software choices for the core14:53
asiethat is, musl, pkgconf, etc14:53
diversedxlr8r: the thing is, and I'm sure teK_ will disagree, but if Crux always upgrades the toolchain and at some point breaks, it would leave your system inoperable, and you would need to use a livecd to fix your system anyway. Because Crux is source based, you would have to tread more carefully than in arch.14:53
koriasie: I'd help ;)14:54
asiekori: I'd probably start with getting all of core/ working14:55
koriasie: well, the thing is14:55
koriI'm assuming you're looking into using dash as /bin/sh, right?14:56
asiekori: mksh14:56
asieand then bash as /bin/bash14:56
asiemirbsd's korn shell, i think it's the best one out there and might even be more compatible with things than dash14:56
asiewith more things*14:56
koriI'm using it as my interactive shell right now14:56
*** phant0mas has quit IRC15:01
*** phant0mas has joined #crux15:02
*** phant0mas has quit IRC15:02
*** chris2 has joined #crux15:15
*** Digit has quit IRC15:23
*** sglob_gemschl has quit IRC15:27
_root_hello folks;15:28
_root_what are you doing?15:28
EmoSpiceTrying to bootstrap my day :P15:28
_root_kori; what's up?15:28
_root_EmoSpice: tell me how if you figure that one out :)15:28
_root_EmoSpice: where was your ports15:29
EmoSpiceFallujah's new album is helping.15:29
EmoSpiceThose are all the ones I'm comfortable saying work.15:30
EmoSpiceThe rest of them are on my machine and may or may not actually work >_>15:30
*** phant0mas has joined #crux15:31
dxlr8rdiverse: true, but others do manage to do it, like gentoo15:31
*** BitPuffin has quit IRC15:32
EmoSpiceNeed to bump the version number on that weechat port.15:34
kori_root_: not much, just came back from lunch15:35
joacimEmoSpice: which port?15:38
EmoSpiceMy fork of your weechat port ;)15:38
joacimthere are so many...15:39
diverseI think I forked joacim's too15:39
EmoSpiceIt's hardly even a fork. I just cloned it locally and wanted to keep it around.15:39
joacimwasn't this many ports on the portdb when i forked mine from nogagplz15:39
diversewell, mine isn't hosted, just local15:40
EmoSpiceI'd like to get something like the AUR for Crux. A centralized hosted place for searching user made ports15:40
EmoSpicethat also has an api15:41
diverseEmoSpice: that's what contrib is15:41
diversenot sure about the api though15:41
EmoSpiceWell, that has more packages15:41
EmoSpiceContrib goes most of the way15:41
joacimaur seems different from contrib15:42
diversecontrib is user made ports15:42
EmoSpicejoacim: it is.15:42
diverseaur is like a database of itself15:42
EmoSpiceAUR is simply ONE spot to host the myriad of user made packages. There exist a ton of small scale ports hosts for Crux.15:43
diverseI know15:43
diversebest to get everyone to use contrib then15:44
EmoSpiceI'll have to read up on it. I'll shut up until i'm more familiar :)15:45
korijoacim: AUR is way different than contrib15:45
koricontrib is more like Arch's [community] repo15:45
teK_but no bad idea, imho15:45
koriteK_: eh15:45
EmoSpicekori: Thanks for the clarification15:46
koriit actually is...15:46
korithe AUR is a mess15:46
korithe current decentralized format is actually really nice15:46
koriEmoSpice: also, does the portsdb page not suffice?15:46
EmoSpiceWell, would it harm someone if there was a centralized thing sitting on top of the decentralized thing?15:46
diversekori: it allows for overlaying that's for sure15:46
*** hhhhhhhh has joined #crux15:46
koriEmoSpice: the portsdb is that centralized thing you're looking for15:47
EmoSpicekind of15:47
koridiverse: that too15:47
asieEmoSpice: AUR is worse simply because sometimes different people want different versions of one package15:48
EmoSpiceMy end goal is actually something like aurora:
EmoSpiceasie: Sure.15:48
koriEmoSpice: have you used pacaur15:51
EmoSpiceI haven't.15:51
EmoSpicemostly because I wrote aurora and just got so used to using it.15:51
kori┐('~'; )┌15:51
korioh well15:51
koriI like the current system a lot, the only thing I'd like to see changed I'm already working on15:52
korito add a git driver for ports so you don't need httpup anymore15:52
diversekori: the logo? :P15:52
koridiverse: that too15:52
korispeaking of which, I need that high quality logo15:52
EmoSpiceI'm only looking for a tool to help FIND the packages in portsdb. Maybe I'll build a tool to scrape portsdb for me :)15:52
korithat's good15:53
EmoSpiceIs the source for portsdb hosted anywhere? I'd be interested in reading it over to see if there's the possibility of hacking in some simple api15:55
EmoSpiceNot that it'd go anywhere, I suspect. but I'd be interested in reading it over15:55
*** joacim has quit IRC15:55
*** BitPuffin has joined #crux15:56
*** joacim has joined #crux15:56
korialright, time to test the git driver16:02
diversebtw, I got cargo to f'ing building16:06
diversethat was unnecessarily painful, now that I got it, it's smooth sailing16:07
diversetons of libraries at my fingertips16:07
korialright, done modifying via's git driver16:14
korikori.git (goes in /etc/ports)
korithe driver
koriPORTS_DIR is defined in /usr/bin/ports16:15
kori result16:16
korijaeger, jue, tilman ^16:16
_root_hello jaeger16:17
korinow, to get individual ports? I have no clue16:17
korithat may be a problem.16:17
tvaalenWhat is the behaviour of the other drivers if you have manually edited some drivers?16:19
tvaalenIf they just overwrite maybe that should be done for the git driver as well.16:19
korican you reword that?16:20
korithe other drivers just overwrite16:20
tvaalenIn that case a git pull might be troublesome.16:20
tvaalenCould lead to merge conflicts and things?16:20
tvaalenMight be better with a git fetch && git reset --hard origin/master16:21
korior we could replicate the behavior via git pull -f16:21
tvaalenNo, that would probably lead to a conflict if you would have for example bumped the version manually and then run the driver and the version is bumped there as well.16:28
koriwell, I honestly think maintaining custom ports in the repo is a bad idea anyway16:30
koriI keep a /usr/ports/custom/ dir for modified ports16:30
*** c0ck4m0u53 has joined #crux16:31
tvaalenAnd getting everything wiped would be a good lesson learned for people who do that. :P16:31
koriI've had my ports wiped :<16:32
_root_I need a life-time  pastebin service. do you know any?16:33
koritvaalen: well, I replaced git pull with your git fetch && git reset --hard origin/master16:33
_root_except pastebin16:33
tvaalenSame here, some time, I think.16:33
tvaalenActually, maybe the drivers as a default should just barf instead and point out the modified ports and ask for them to be moved, "or -f to overwrite".16:34
tvaalenOn the other hand, that would be painful to implement for other drivers than the git one.16:35
tvaalenSo maybe not.16:35
*** hbekel has joined #crux16:37
*** pnutzh4x0r has left #crux ("WeeChat 1.0")16:42
*** kahiru has joined #crux16:44
*** asie has quit IRC16:48
*** quasinoxen has quit IRC17:17
*** lnds has joined #crux17:18
*** hbekel has quit IRC17:43
*** kori has quit IRC18:47
*** kori has joined #crux18:51
*** kori has quit IRC18:51
*** kori has joined #crux18:51
*** sh4rm4 has quit IRC18:58
*** kori has quit IRC19:08
*** kori has joined #crux19:10
*** kori has quit IRC19:10
*** kori has joined #crux19:10
*** leetspete1 has joined #crux19:22
diversekori: wait, you lost your ports?19:22
koridiverse: huh? no19:26
koribut before I learnt to make a custom ports dir, I did lose changes I had made to Pkgfiles when they updated via ports -u19:26
diverseah, right19:27
*** c0ck4m0u53 has quit IRC19:30
diversekori: so, the git driver is ready for use?19:33
koridiverse: not yet19:34
koriI'm still testing it19:34
koridiverse: but here's how the .git files are going to work
koriwhat do you think?19:56
*** kori has quit IRC20:25
nbinoI was thinking prt-get search wold find packages in contrib, for exammple I have been making some packages that Romster have in portdb20:25
*** kori_ has joined #crux20:32
*** kori_ has quit IRC20:32
*** kori has joined #crux20:33
*** kori has quit IRC20:33
*** kori has joined #crux20:33
jaegerkori: just as a thought about branches. not tested much20:33
korijaeger: mmm20:35
korijaeger: is the differentiation between ROOT_DIR and ports(8)'s PORTS_DIR really needed?20:35
korialso, do you have a high quality version of the CRUX penguin?20:35
korihigh res20:35
korithat is20:35
jaegerI don't have any version of it, heh20:36
korido you know anyone that does?20:36
koriI want to make a flat logo for CRUX20:36
jaegeras far as ROOT_DIR vs. PORTS_DIR it's sourced from the .git or .httpup or whatever file20:36
jaegeryou could call it BUTTS if you felt like it20:36
jaegeras long as the driver uses that20:36
koriI don't define ROOT_DIR for the git driver and files20:37
koriI use PORTS_DIR/NAME20:37
koriand NAME is defined in the .git file20:37
jaegerthe variable names aren't the point, though, it was the branch bit20:38
koriaye I'm just asking since you're here now20:38
koribut yes the branch idea is nice.20:38
*** kori has quit IRC20:39
*** kori has joined #crux20:40
*** kori_ has joined #crux20:42
*** kori has quit IRC20:42
*** kori has joined #crux20:42
koriright as I reconnected, it started raining and I lost connection20:43
tvaalenYou lose the connection when it rains?20:44
koritvaalen: it's been happening.20:45
koriI'm really not sure why20:45
koribut it happens 100% of the time20:45
tilmanare you in a country with poor infrastructure?20:45
tilmanlike the usa?20:45
tilmani feel a bit bad now20:46
koritilman: I live in Brazil20:47
koriso yes.20:47
tvaalenSounds like a weird effect in any case.20:48
koriI use some form of wireless20:48
koriethernet to a router connected wirelessly to another router20:49
tvaalenSomewhere around the equator with those monster rains? :)20:50
korijaeger: might have to add a message to maintain consistency20:54
*** sudobaal has quit IRC20:58
*** Feksclaus has joined #crux21:00
*** lnds has quit IRC21:10
*** leo-unglaub has joined #crux21:21
*** quasinoxen has joined #crux21:24
*** BitPuffin has quit IRC21:38
*** jdolan has quit IRC21:49
*** hhhhhhhh_ has joined #crux21:55
*** Workster_ has joined #crux22:07
*** Workster has quit IRC22:07
*** Workster has joined #crux22:07
*** phant0mas has quit IRC22:21
*** u7knv9h has joined #crux22:29
u7knv9hyo, how do i extract my initramfs/initrd from vmlinuz?22:31
koricome on, dudes22:36
*** u7knv9h has quit IRC22:37
*** u7knv9h has joined #crux22:37
joacimdo it yourself =)22:37
korijoacim: about to22:38
korifrogatto is a damn good game22:38
koriI'll have to make a port for anura too22:39
koriWorkster: huh22:41
Worksteroh not the same name nevermind22:41
Worksteri saw aur22:41
Workstermissed the n22:41
korimm I'm gonna leave this for later22:44
korisounds like its going to be some work22:44
*** jdolan has joined #crux22:47
Worksternot unless you have to make the dependency ports first22:48
koriWorkster: I do22:50
koriwell, I have done this before22:50
koriwith neovim22:50
koribut neovim was a priority, frogatto isn't22:50
*** jdolan has quit IRC23:01
*** Roosevelt has quit IRC23:02
*** nilp has quit IRC23:02
*** nilp has joined #crux23:04
Workstereh none of the deps are exotic and in crux now.23:21
Workster use that as a guide when porting kori23:21
koriWorkster: I always do23:22
koriwell, unless the port is simple enough23:22
*** jdolan has joined #crux23:25
*** hhhhhhhh_ has quit IRC23:28
*** kahiru has quit IRC23:40
*** BitPuffin has joined #crux23:43

Generated by 2.11.0 by Marius Gedminas - find it at!