IRC Logs for #crux-devel Tuesday, 2016-11-08

teK__frinnst: push llvm update, also built mesa3d which went through fine. Please check if needed01:01
*** onodera has quit IRC01:15
teK__frinnst /  jaeger: first btrfs-progs need a bump to 4.8.2, then docker can be bumped to 1.12.3 with additional adjustment of the dynbinary install line02:06
teK__clang update: ETA tomorrow night02:06
*** mavrick61 has quit IRC03:38
*** mavrick61 has joined #crux-devel03:39
*** darfo has quit IRC05:16
*** darfo has joined #crux-devel05:16
*** Romster has quit IRC05:19
jaegerteK__: made a quick tweak to opt/llvm to remove config.h refs, hope that's ok05:19
*** Romster has joined #crux-devel05:19
frinnstack on btrfs07:40
frinnst2http://7c566eb6372fe998.paste.se/08:17
frinnst2im surprised that the ram sticks run so hot08:17
Romsteris it over clocked or the ram voltage to high?08:24
Romsterheat spreaders on ram?08:24
*** groovy2shoes has joined #crux-devel08:33
frinnst2no clock and yeah, spreaders09:00
frinnst2its the new crux server09:00
frinnst2ddr2 ecc buffered09:00
Romsteri guess it wont get much hotter than that09:32
Romsteralso i cna't decrypt that email as you used my romster.me key and not the more recent one i created that uses gmail. as i forwared email from romster.me to gmail.09:33
Romstergpg: key 5433B7BDD03BAE3109:34
frinnst2yeah forgot about that yesterday09:34
Romsteri figure it can't find the private key because it doesn't match my email.09:34
Romsteremail forwarding isn't as painless as i thought it was09:35
frinnst2you should be able to force that though09:35
frinnst2if you still have your private key that is09:35
Romsteri do09:36
Romsterbut i don't know how on thunderbird with enigmail addon09:36
Romsterhmm i can change the key in the email account area09:38
Romsterseems that is only for encrypting... guess i'll try expert mode and see if i can do anything then.09:45
frinnst2if your private key is added to the keychain you should be able to decrypt it. not too sure how exactly you do that in thunderbird though09:45
frinnst2but yeah, i agree. gpg and encryption in genreral can be a bit confusing :-)09:46
Romsteri have it in my key chain and enigmail picked up on that for me09:46
Romsterand i can pick what key i encrypt with09:47
Romsteri got 3 keys why can't i decrypt with either of them 3...09:48
frinnst2I have a separate key at work. let me try and send you an encrypted email from that09:48
frinnst2sent09:52
Romsterhttp://stackoverflow.com/questions/18335749/enigmail-gpg-error-gpg-decryption-failed-secret-key-not-available09:52
RomsterSounds stupid, but make sure you are not mixing up gpg keys and ssh keys. That's just the kind of brain fart that will have you tearing your hair out. That's why I keep my hair short :)09:52
RomsterError - no matching private/secret key found to decrypt message09:55
Romsteron all of them09:55
frinnst2can you send one to me?10:11
frinnst2sent you a clear text mail with my pub key attached10:12
frinnst2"hope you can read this"10:19
frinnst2yep10:19
Romstergpg: decryption failed: No secret key10:20
RomsterNote: The message is encrypted for the following User ID's / Keys: 0xA22D98121B616564 (Danny Rawlins (crux key) <monster.romster@gmail.com>), 0xA7CDA8BA68693DF9 (Fredrik Rinnestam <fredrik.rinnestam@valvea.se>)10:20
Romstermaybe Enigmail is broken on thunderbird it wont even ask me for my passphrase i set on my key10:21
Romsterhttp://askubuntu.com/questions/670908/enigmail-not-asking-for-pgp-passphrase-but-saying-no-key-available10:21
RomsterACTION shrugs10:21
Romstergpg-agent is running.10:22
*** onodera has joined #crux-devel10:23
Romsteri just installed pinentry-gtk-2 and nowi can decode it frinnst10:30
frinnst2cool10:31
Romsterbut now i ened to see if i did save my passphrase for that other key10:32
Romsteraha there we go finally..10:34
Romsterso do these go in /etc/ports/ ?10:35
frinnst2yep10:35
Romsterwell the pub ones is n the repo itself.10:35
frinnst2all go in etc/ports10:36
frinnst2contrib.pub is not distributed with ports currently. I've added it to the root of contrib for now10:37
frinnst2we might want to do that with the rest of the repos too10:37
Romsterah10:38
Romsterok so the .pub and .sec are the public and private key pair. easy enough.10:39
frinnst2yep10:39
Romsteri'll need the other repos at some point but i can start on contrib.10:39
frinnst2and the .sec file should never ever ever be given out to anyone :)10:39
Romsteri realize that.10:39
Romsterall this does is secure the ports tree itself not the distfiles in it. that's left up to the packager to make sure it's not been tampered with.10:40
frinnst2https://crux.nu/Wiki/SignedPorts10:41
frinnst2yeah but the distfiles are signed as well10:41
Romsteris that the oage you deleted twice on the wiki :D10:41
frinnst2http://713706ab5af615f1.paste.se/10:41
frinnst2I JUST DELETED IT THE ONE TIME10:42
frinnst2:D10:42
Romsteri'm joking10:42
frinnst2me too :)10:42
frinnst2it was quite the epic fail10:42
Romsterhmm shouldn't we be not signing the footprint because of new files or not a issue?10:42
frinnst2I still owe tek a beer10:42
Romstersince we can still do ignore new.10:42
frinnst2we dont verify the footprint right now10:43
Romstershit happens there...10:43
frinnst2but the data is still there10:43
Romsteri see10:43
Romsterso we do ave sha256 for the hash i see.10:43
frinnst2everything is signed. obviously pre/post installscripts are critical10:43
frinnst2and stuff like rc-scripts10:44
Romstera duckduckgo search of that sha256 sum in the signify file leads me to this http://git.alpinelinux.org/cgit/aports/tree/main/gcc/APKBUILD#n63510:45
frinnst2From: Microsoft Azure Team [mailto:wapteams@microsoft.com]10:45
frinnst2Subject: If you want a datacenter to be future-proof, you need to think Open10:45
frinnst2Join EnterpriseDB and Microsoft for a webinar on Tuesday 15th November 2016 11:00-12:00 CET where we will talk through how to build a future-ready datacentre and its open source based solutions.10:45
Romsterso it's easy to see if a hash is bad or not.10:45
frinnst2wat?10:46
Romsterfiles are hashed and signed10:46
frinnst2no, those hashes are not generated with our private key10:46
frinnst2so they are not valid for us10:46
frinnst2actually, sorry10:47
frinnst2yes they are10:47
Romsterthe line 4 is a hash that is encrypted with secret key and added to the .signature10:48
frinnst2yeah10:48
Romsteri never really looked into it but it makes perfect sense now.10:49
Romsteri could writhe a tool to check them sha256 sums and more importantly i can use my distfiles/hash/ layout with that information and the real sha256 sum of the downloaded files i download and compare them.10:50
Romsterso i can do my own downloading and checking my mirrors files.10:50
Romsterinstead of relying on the .md5sum files that i do now.10:50
frinnst2sweet10:51
Romsterwhat i do is download make a directory for hash/md5sum/filename/ actual file here.10:51
Romsterand then i remove distfiles and hard link the hash back to the distfiles/ and anything missing is either corrupt or tampered with10:52
Romsteretc.10:52
Romsterglad i got this sorted so i had to makes sure i have pinentry-gtk10:53
Romsterfrinnst, hmm you could pass me over keys for opt xorg compat-32 sometime too please10:54
frinnst2yeah remind me in 6hrs :-)10:54
frinnst2or i might be able to get it done during lunch10:55
Romsterno rush just when you are not busy10:55
Romsterthank you for taking some timeout of work for helping me.10:56
frinnst2no worries10:56
juefrinnst2: guess you are running mesa 13 on 3.3?11:02
frinnst2not yet11:03
jueok, just started to play with it11:03
frinnst2iirc cmake is already on the iso?11:06
jueyes11:07
juewhy?11:08
jue13.0.0 build fine after I changed CONFIG_SHELL to bash, don't see a need for cmake11:10
juebesides that I've added --with-vulkan-drivers=intel,radeon and got the corresponding libs for that11:12
juedunno if we need something else ;)11:13
Romsterevery llvm update breaks ABI for mesa3d11:55
juewell, new llvm works with mesa 12 and 1312:00
jueand nothing seems broken after the mesa update to 1312:01
frinnst2/usr/ports/work/gcc-6.2.0-1-work/src/build/x86_64-pc-linux-gnu/32/libstdc++-v3/include/tr1/utility:106:1: fatal error: can't write PCH file: No space left on device12:40
frinnst2asdf12:40
frinnst2is radeon correect for vulkan, jue?13:29
frinnst2its not the right driver for amdgpu iirc13:29
frinnst2i'll take a look after work13:29
teK__thanks jaeger fsck, I missed that. Will looki into it tonight14:05
juefrinnst2: AFAIK, vulkan is not a graphic driver but a new 3-D API like DirectX14:17
frinnst2yeah I know14:17
frinnst2but iirc its called something else for amdgpu14:17
frinnst2but the configure script accepts "radeon"?14:17
jueyeah, and we get a new lib /usr/lib/libvulkan_radeon.so14:18
frinnst2*downloading now*14:18
frinnst2RADV or something14:21
frinnst2The RADV driver will not work with the Radeon DRM driver nor will it be ported to any pre-GCN (HD 7000 series) hardware.14:22
frinnst2http://www.phoronix.com/scan.php?page=article&item=radv-hits-mesa&num=114:22
*** onodera has quit IRC14:22
frinnst2my brain is too fried to think14:25
frinnst2or i might have dreamt it14:26
*** onodera has joined #crux-devel14:55
frinnstRomster: sent16:05
frinnstactually, think i sent to your old key16:11
frinnstthere we go16:16
pedjateK__, is ti possible to explicitly disable the build of bindings for llvm?18:07
pedjaI can't find anything in cmake -L output18:07
pedjait would be cleaner that way, imho18:08
pedjathere was a configure switch for that, but it never made it to cmake.18:11
frinnstcmake is horrible to use18:19
frinnstor maybe im just stupid18:19
pedjanah, cmake is PITA ;)18:22
pedjasince there is no way to disable bindings, I found a patch that at least puts them in the right place.18:23
pedjaI'll open the bug on FS when I finish it.18:24
Romsteri end up looking manually in CMakeLists.txt21:22
Romsterand look for HAS or PROVIDES or ENABLE_ or what ever it is i am looking oor if there is an option to enable or disable.21:23
Romsterwhich reminds me i need to finish up on conky too21:23
Romsterfrinnst, i can decode from contact@romster.me and would prefer to use that in future, i thought it was to do with not matching my email but it was because i had no pinentry-gtk221:29
Romsterand cheers i got them now21:29
RomsterteK__, opt/libesmtp has 404 for source missing21:36
Romstersepen if you are reading this you need to fix missing patch for opt/imlib2 also jdk needs a bump to match jre and flash-player-plugin also needs a bump.21:46
Romsteri'll file bug reports later.21:46
RomsterURL transformed to HTTPS due to an HSTS policy21:46
Romster--2016-11-09 19:40:34--  https://www.wireshark.org/download/src/all-versions/wireshark-2.2.1.tar.bz221:46
Romsteralan you need to fix up missing sources for hunspell-es it's been broken for ages now21:47
*** heroux has quit IRC22:07
*** heroux has joined #crux-devel22:08
pedjaso I patched llvm to install ocaml bindings properly, but now ocamlfind can't find them.fun.22:18
pedjaI emailed ocaml-findlib developer, that might be a bug.We'll see.22:19
pedjaRomster, have you tried using Claws Mail with GnuPG?If you still have issues with Thunderbird, that is.22:27
Worksterpedja, it works now but i have tried to use claws mail in the past but i ended up going back to thunderbird, i don't remember why now.22:34
pedjaok22:35
pedjaI've never used the gnupg portion of it, but it should just work[tm], if Internet is to be believed :)22:36
Workstermy issue was i didn't have the gtk version of pinentry so thunderbird would not ask for my passphrase22:49
teK__frinnst: cmake is fine, if you know about cmake -LAH23:16
teK__also @pedja -^ maybe23:17
pedjaI always used just the -L switch.hmm.23:20
teK__dunno if AH changes things23:21

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