IRC Logs for #crux-devel Thursday, 2017-03-02

*** chinarulezzz has quit IRC01:05
*** chinarulezzz has joined #crux-devel01:07
*** frinnst has quit IRC02:19
*** frinnst has joined #crux-devel02:51
*** frinnst has quit IRC02:51
*** frinnst has joined #crux-devel02:51
*** _________mavric6 has joined #crux-devel03:44
*** chinarulezzz has quit IRC03:48
*** chinarulezzz has joined #crux-devel03:50
*** chinarulezzz has quit IRC06:29
*** chinarulezzz has joined #crux-devel06:32
*** chinarulezzz has quit IRC08:13
*** chinarulezzz has joined #crux-devel08:14
*** tnut has joined #crux-devel08:49
tnutMorning08:50
Romsternow that we are using .signatures can we kill off .md5sum ?09:12
tnutHello, I introduce myself (for peoples who don't know me) I've start a fork of pkgutils 3 and half years ago (https://github.com/NuTyX/cards). I've been asking myself how do you handle signature.09:23
Romsterhi tnut should be able to figure it out from our pkgutils git tree09:26
tnutReading09:26
Romsterhttps://crux.nu/gitweb/?p=tools/pkgutils.git;a=summary09:27
tnutin fact I don't have a picture the way it works09:28
Romsterhttps://crux.nu/ports/crux-3.3/core/signify/Pkgfile09:29
Romsterwe use that to make foo.pub and foo.sec09:29
Romsterpkgmk has code in it to read from /etc/ports/09:29
tnutHas it something to do with a key pair generation ?09:29
Romsterto read the .signature out of each directory and compare that to the foo.pub file09:30
tnuta public one available somewhere and your own one09:30
Romsterwhen you do pkgmk -us it uses foo.sec the secret key to make the .signature09:30
tnutI see09:30
Romsteryes signify makes a key pair09:30
Romsterwe have keys for a REPO name.09:30
Romsteror well collection so a opt.sec and opt.pub09:31
tnutREPO = Collection ?09:31
Romsterand so on.09:31
Romstercorrect09:31
tnutok09:31
Romsteri keep looking at my empty coffee cup need a refill09:32
tnutmeans when a user do ports -u, the public key comes over as well otherwise I don't see how this can works09:32
Romsteri have added romster.pub to the base of my ports collection in /usr/ports/romster/romster.pub09:33
tnutlet's say ports -u xorg for ex09:33
Romsterso users can cp that to /etc/ports09:33
Romsterthe main ones are already in there.09:34
tnutvoilĂ 09:34
tnutACTION start to get the picture09:34
Romsterprt-get fsearch opt.pub09:34
RomsterFound in /usr/ports/core/ports:09:34
Romster  /etc/ports/opt.pub09:34
Romstercore/ports/ has the rsync and pub files.09:35
Romsterfor core opt xorg contrib compat-3209:35
tnutchecked yep09:35
Romsterman signify09:36
Romsterthat should be all you need to get started.09:36
Romsterbeen meaning to look at nuttyx sometime to see what's different.09:37
tnutso idea behind that is to ensure  that the pkgfile XX and all it's relevant files  have been produce by Person YY and has not been change by anybody09:38
Romstercorrect09:38
tnutRomster: yes It's quite a bit of differences09:39
Romsterit makes a sha256sum of Pkgfile patches ... and source files and then that is signed09:39
tnutpkgmk.man.in can be a good starting point to have an idea09:39
tnutI see09:39
tnutthat's the bigest issue with cards09:40
Romsteri'm a opt about 50% of contrib and 1 of mostly 2 packagers in compat-32 as well as my own ports.09:40
Romsterso i do quite a bit here.09:40
Romsterthere is like i think 5 of us maintaining the official ports09:40
tnutthat's already a lot09:41
Romsterwell md5 is broken and so is sha1 now.09:41
tnutyou no need to many people for that anyway09:41
tnutyou need trust peoples09:41
Romstercorrect09:41
tnutwe are two people, and honestly that ok09:42
Romsteri test my stuff in docker to make sure it all works and dependencies are correct09:42
Romsterit would be ok if i wasn't busy with work and sleep and other things.09:42
Romsteri work in IT09:43
tnut:)09:43
Romsteri get burned out i need to take a rest09:43
tnuta bit like me09:43
tnuta great help (if I may talk about it otherwise just say it) with cards is:09:44
Romsteri work on photocopiers iphones computers networking and sometimes even electronics.09:44
tnutA port will always be compile from base system09:44
Romsteri can't see why you can't talk about it09:44
tnutand if the user choose the binarie scemario (which is the default one) ports have to be done in a chroot09:45
tnutcards will also shows the level of dependencies, a great use when dealing with deps09:46
tnutYou are now having professionnal break Romster ?09:50
Romsteri just made a coffee09:54
Romsterso you made cards do binary09:54
tnutyes09:55
tnutadd some variables as well in pkgmk script09:55
tnutACTION will need to consider the key story09:57
Romsteri'll take a look at it one day09:57
tnut:) as you wish09:58
Romsterand if i go quiet it's because i got busy with dogs or kitchen or bathroom or something else on the computer etc. replies can be late but i do read.09:58
tnut:) I know same here09:59
Romsterso how goes things?10:06
tnutfine fine :D10:08
tnutACTION is at work10:08
Romsterah i am at home in my evening 9:10pm10:10
tnutOk you are in US10:20
tnutno sorry10:20
RomsterAUS10:21
tnutAustralia may be10:21
tnutright10:21
Romsterone of a few10:22
tnutI drop the md5sum story10:23
tnuteven if still there (in the code)10:23
tnutNow it's a chance to switch10:24
*** tnut has quit IRC10:32
*** tnut has joined #crux-devel11:29
tnutre :)12:21
*** chinarulezzz has quit IRC13:33
*** chinarulezzz has joined #crux-devel13:35
*** tnut has left #crux-devel ()14:08
*** erdic has joined #crux-devel14:54
*** erdic has quit IRC15:10
*** erdic has joined #crux-devel15:14
frinnstjaeger: I switched the source URL for core/dash to use your mirror, hope you don't mind15:58
frinnstthe upstream site is still down15:58
frinnstalso switched all gnu sources to use ftpmirror.gnu.org instead15:59
jaegerok15:59
frinnstasdf. the motherboard i chose is blocking my ryzen order. according to newegg it will be released next week16:02
frinnsthttps://www.newegg.com/Product/Product.aspx?Item=N82E1681313296416:02
jaegerdoh16:03
frinnstgives me another week to change my mind and cancel the order :>16:04
jaegerhehe, I suppose so16:04
frinnstphoronix had some benchmarks. the performance looks mighty nice16:05
crash_ryzen seems to perform very nice18:07
*** tnut has joined #crux-devel19:22
*** jaeger has quit IRC20:36
*** jaeger has joined #crux-devel20:41
jaegercrux.ninja is updated to 3.3 now, it should start updating the distfiles to 3.3 sources20:56
RomsterAMD Ryzen Processors support DDR4 2666/ 2400/ 2133 <- really i read 3200Mhz and some places say they support 3600Mhz ram21:03
Romsterthat board says 2400MHz others say 3200Mhz OC so i guess the over clocking is what can go up to 3600MHz21:17
*** tnut has quit IRC21:55
jaegerAnything over 2133 is technically overclocked22:06
*** s4nni has quit IRC22:19
*** chinarulezzz has quit IRC22:37
*** chinarulezzz has joined #crux-devel22:41

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