IRC Logs for #crux Wednesday, 2020-01-15

*** duskmourn has joined #crux00:05
duskmournEverytime I install Crux I get stuck at the Grub command line after rebooting00:22
duskmournnot sure what I keep doing wrong00:22
*** obarun has joined #crux00:29
duskmournI'm starting to feel like getting efi to work is a lot more trouble than it's worth00:30
][_R_][EFI is pretty pants00:34
*** xor29ah has quit IRC00:37
*** xor29ah has joined #crux00:40
*** onodera has quit IRC00:43
*** duskmourn has quit IRC01:03
jaegerduskmourn: if you get the grub shell then grub is installed, it's just missing its config file or the config file wasn't generated properly02:02
fysi got efi working the first time i dealt with it, it seems more complicated than it is02:16
jaegerI prefer it, it seems simpler and less error-prone to me02:20
fysyeah02:22
fysbuilding nim on my pi4... this should be fun02:33
fyslol02:33
fyshm, not too bad actually.02:45
*** duskmourn has joined #crux02:53
duskmourndid another install, now I just get a black screen with a blinking cursor at boot02:54
duskmournLilo didn't throw any errors after I ran the comment to fix my "broken partition table" so that shouldn't be an issue still02:54
*** duskmourn has quit IRC03:09
*** duskmourn has joined #crux03:11
*** v3nus has joined #crux03:13
fysJesus, cmake is taking longer to compile than nim.03:16
fyslol03:16
fysThat's not cool, man.03:16
duskmournat least it's not qt503:26
jaegerduskmourn: if you get the grub shell then grub is installed, it's just missing its config file or the config file wasn't generated properly03:27
*** v3nus has quit IRC03:30
duskmournI'm using Lilo03:30
mayfrostI am running Linux with EFI stub enabled and no other bootloader03:30
*** duskmourn has quit IRC03:30
jaegerresponding to your previus grub comments03:31
jaegerprevious03:31
mayfrosttotalling recommended03:31
fyshe dropped.03:31
fys:s03:31
jaegeroh well. If he wants he can read the log03:31
*** duskmourn has joined #crux03:31
fysi spent all night looking for something which was on my desk /right/ next to my keyboard.. oof.03:32
mayfrostduskmourn: try running linux efi stub with no bootloader03:33
duskmournif I did a efi setup I would try that03:33
duskmournhttps://files.catbox.moe/r80pa9.jpg03:34
duskmournnothing wrong here right?03:35
jaegerwhy install=/boot/boot.0802?03:36
mayfrostI would remove the lines with install and timeout03:37
mayfrostor replace install with "install=text"03:38
jaegerIs /dev/sda2 mounted at /boot? If so the image= line may need to change03:38
duskmournyeah it is03:38
jaegerthen image=/vmlinuz is probably what you want03:40
duskmournah03:40
jaegerunless you have /boot/boot/vmlinuz03:40
jaegerAlso, if your kernel is named "vmlinuz" that's probably why you got a grub shell instead of menu03:40
jaegergrub-mkconfig was looking for "vmlinuz-*" so something like "vmlinuz-4.19.94"03:41
duskmournright03:41
duskmournthis is Lilo not Grub. also "image=vmlinuz" cannot be found when I run Lilo03:43
jaegeryes, I understand this is lilo03:43
jaegerEARLIER, you said:03:43
jaeger18:23 < duskmourn> Everytime I install Crux I get stuck at the Grub command line after rebooting03:43
jaegerThe grub talk is ENTIRELY in reference to that03:44
duskmournah03:44
duskmournyeah that was a different install attempt03:44
jaegerwas just sharing info because knowledge is always good03:48
duskmournI appreciate it03:49
*** tilman_ has joined #crux04:16
*** ivandi has quit IRC04:31
duskmournchanging "boot" to just /devl04:35
*** duskmourn has quit IRC05:21
*** onodera has joined #crux07:50
*** frinnst has joined #crux07:53
*** frinnst has quit IRC07:53
*** frinnst has joined #crux07:53
*** onodera has quit IRC08:02
*** pedja has joined #crux08:16
*** Workster has quit IRC08:43
*** guido_rokepo has joined #crux09:04
*** guido_rokepo has quit IRC09:07
*** posixmeharder has joined #crux09:41
cruxbot[contrib.git/3.5]: nnn: 2.8.1 -> 2.910:17
TimB_stenur: i recently resortet 280gb with musicbrainz, worked ok, took some time, but i got there10:18
cruxbot[contrib.git/3.5]: pulseeffects: 4.7.0 -> 4.7.110:48
*** jue has joined #crux11:13
cruxbot[compat-32.git/3.5]: nss-32: 3.49 -> 3.49.111:22
stenurefi boot stub i should try too. efibootmgr and efivars i installed last week, but have not yet dared to try. It shows terrible things, what has refind done?? :-) But it works here (refind). maybe syslinux. But efi stub, if i would have more testicles, efibootmgr plus stub. Does this work from BTRFS subvolume???12:54
stenurTimB_: ? But not disc ripping, you used some native-to-musicbrainz GUI i presume?12:55
stenurThe query string built into the libdiscid of CRUX uses v1 protocol, and using it returns "https://blog.metabrainz.org/2018/02/01/web-service-ver-1-0-ws-1-will-be-removed-in-6-months/"12:56
stenur"That is way off maaan" (Cheech and Chong)12:57
*** v3nus has joined #crux13:00
stenurOr "way out man". Hm. Haven't watched for long time.13:09
stenurNaah, tried access via JSON protocol, but that is marked as "beta" so maybe i should have been more silent. But the one service i need (lookup via DiscID) does not yet support it, all other tested ones do. Mysterious. :) And quite a lot of disc do not have any Disc IDs, neither "old" FreeDB ones nor "new" MusicBrainz ones. I think i have to go for media id, but for that i have to check all CDs again!!13:11
stenurAnd also, take Pink Floyd's Dark Side of the Moon, the (manually entered?) sidebar notes are correct, but there are two disc entries that matter, i have Harvest, and that one has "wrong" TOC offsets and disc times. I cannot take the TOC entries of the CD and lookup the result!13:14
*** zimmer__Dl has joined #crux13:15
stenurThere is a correct entry in the "CD stub" thingie, but that database is unrelated to the main database, meaning there is no! way to get from one to the other. But only with the main database i can find the MusicBrainz ID that is actually used to lookup entries.13:15
stenur(That "CD stub" entry does not even has the artist field filled im.)13:17
*** zimmer_Dl has quit IRC13:17
stenurThat is real shit. Not that they get money from Microsoft. No. Not that they invent a SHA-1 based new ID that seems to have lesser duplicates than the CDDB one has/d. (But there are entries with dozens of duplicates still!)13:18
stenurThey have massive amounts of data, but FreeDB/CDDB ids they seem to have removed. I cannot do anything with them.  Is it four bytes per actual disc entry.13:20
stenurTo me it looks as of now (having started to deal with this yesterday evening) as if it is media ID centric, aka industry ID. And as if bugs in their frontend have made it into the database as a status quo - because what else does explain that my valid TOC entries cannot be used to match the very CD in the database?13:22
stenurBah. Beside that i will first hack together a small thing here to get rid of libcdio and cd-paranoia etc.  This is more than 30 MB, but it seems that on Linux you can do that (i.e., CD TOC, CD track read, hopefully CD-Text) in less than 30 KB.13:23
daftaupewhat are you trying to achieve stenur ?13:27
*** zimmer_Dl has joined #crux13:27
stenurI mean, the TOC is wrong for cd-paranoia, but it is right for cd-info. If i can do that in 30 KB i can even include it in my disc ripper perl script for optional on-the-fly compile/install, and we are still < 130KB (now 76KB, but i think MusicBrainz will require tremendous logic, HTTP Tiny, XML parser, redoing queries to finally find a result.)13:27
stenurdaftaupe: i have "ported" my discripper perl script to Linux, and had nothing but problems. And before i have heard that CDDB/FreeDB, what i use to get CD text (artist, song titles etc.), will be turned off by the end of March.13:28
stenurdaftaupe: So i need a different source. frinnst i think it was who pointed me to MusicBrainz. And that i tried yesterday evening.13:30
frinnstnope, never used it13:30
*** zimmer__Dl has quit IRC13:30
daftaupewhat did you use to run it under stenur ?13:31
*** obarun has quit IRC13:31
daftaupeoh they're turning CDDB/FreeDB off ?13:31
stenurThen romster?13:31
stenuryes.13:32
daftaupeand so they don't use the sames ids as CDDB/FreeDB if I understand you correctly13:35
stenurThey seem to have had a FreeDB interface until last year. Maybe their old version 1 format supported that ID, but the version 2 they have now does not.13:37
stenurI mean, that alone would not matter, since the MB-ID is calculated with the same data that CDDB/FreeDB-ID used.13:37
stenurBut it seems that this ID is not present for a lot of entries in the MB database. And those which are seem to be wrong, at least partial.13:38
stenurI presume it is because gathering CD TOC entries is sometimes bogus, for example with current cdparanoia and cd-paranoia they are wrong for what i've tried and could verify. They are right for cd-info.13:39
stenurSo i presume such false IDs made their way in the MB database, because my correct TOC data cannot be used to query some CDs there.13:40
*** v3nus has quit IRC13:41
stenurIt is pretty messy it seems. But it was no earlier but yesterday evening that i tried it first, i did not know about MB until some days ago. I used CDDB.pm of Perl for almost two decades i would say, and never cared for any background.13:42
stenurI have to go, a wonderful sunny day, we have southwest stream here in Germany/Europe, it is warm like never.  But winter will come, next week...13:43
daftaupehow could the toc be wrong for cdparanoia and not for cd-info ? Aren't they using the same data ?13:43
daftaupe(I'm a total noob concerning cds related stuff)13:44
*** v3nus_ has joined #crux13:49
pedjaerr, freedb gateway, stenur? gone, apparently https://blog.metabrainz.org/2018/09/18/freedb-gateway-end-of-life-notice-march-18-2019/13:54
pedjaapropos MB relationship with MS https://blog.metabrainz.org/2019/12/03/thank-you-microsoft/13:57
pedjatl;dr they donate Azure credits and sponsor annual team meeting13:58
pedjaseems fair, since they use the resource for their bing thing14:02
pedjaone day when I get my desktop back, I might spin up postgres db with MB, what.cd and freedb data14:05
cruxbot[opt.git/3.5]: cups-filters: update to 1.26.214:16
cruxbot[core.git/3.5]: libcap: update to 2.3114:21
BurnZeZ01:56:14 <Romster> BurnZeZ, https://crux.ster.zone/packages/3.5/qtwebengine%235.13.2-1.pkg.tar.xz15:03
BurnZeZThanks15:03
BurnZeZDownloading it now15:03
BurnZeZThis is why I wish the build system used namespaces with a built-up root15:04
BurnZeZSo that you build the thing in an environment where only the explicitly given dependencies exist15:04
cruxbot[xorg.git/3.5]: xorg-server: update to 1.20.715:11
pedjapre build setup can probably be automated, and podman and, I think, lxc can run rootless15:29
pedjaso you have a dedicated builder user, with minimal privileges15:29
pedjadocker images are, after all, a glorified chroot, as people say :)15:31
*** weednix has joined #crux15:34
*** zimmer__Dl has joined #crux16:05
*** zimmer_Dl has quit IRC16:08
BurnZeZRomster: Seems it crashes16:43
*** duskmourn has joined #crux16:44
duskmournis there supposed to actually be stuff in /use/ports after an install?16:44
jaegerNot until you run 'ports -u'16:45
duskmournah16:45
BurnZeZlibpthread.so.0, libQt5WebEngineWidgets.so.5, and libQt5WebEngineCore.so.5 are the first in the stacktrace16:46
duskmournyeah I don't fully understand the package management yet16:46
jaegerJust play with it for a while and it'll make sense16:48
duskmournyeah I've used Slackware before and it's fairly similar to that so it shouldn't be too hard to figure out16:48
*** duskmourn has quit IRC17:07
*** duskmourn has joined #crux17:10
*** duskmourn has quit IRC17:14
*** zimmer_Dl has joined #crux17:18
*** zimmer__Dl has quit IRC17:21
*** pedja has quit IRC17:24
*** duskmourn has joined #crux17:30
duskmournman it seems like every port I try to install fails17:30
*** duskmourn2 has joined #crux17:32
*** onodera has joined #crux17:33
duskmourn2and yet when I check pkginfo -i it says the ports are actually installed17:33
jaegerprt-get is your best friend when installing ports, to catch dependencies17:34
jaegerbut if you're using that already and running into errors, paste specific ones and we can take a look17:34
jaegervia pastebin.com or similar17:34
*** duskmourn has quit IRC17:35
duskmourn2no I'm not using it, but I probably should17:36
jaegeryeah. In the future maybe pkgutils itself will understand deps but right now it doesn't but prt-get does17:38
duskmourn2yeah I'm running a prt-get sysup right now, outdated packages from the iso might be why everything keeps failing17:43
jaegerAlso not a bad idea to check if any new deps have been added since the ISO17:44
jaegerprt-get depends $(prt-get listinst) | grep "\[ \]"17:44
jaegeryou should also install prt-utils and get used to using revdep17:45
jaegerWhen libraries get upgraded and break ABI, revdep will help you find what needs to be recompiled17:45
duskmourn2alright17:45
*** pedja has joined #crux17:53
*** v3nus has joined #crux17:59
BurnZeZRomster: Also, you stripped the debug symbols18:43
BurnZeZOr whoever built it18:44
*** zimmer__Dl has joined #crux18:44
*** zimmer_Dl has quit IRC18:46
*** v3nus has quit IRC18:47
*** v3nus has joined #crux18:49
*** zimmer_Dl has joined #crux19:05
*** zimmer__Dl has quit IRC19:07
*** mayfrost has quit IRC19:26
*** v3nus has quit IRC19:34
TimB_stenur: picard19:40
*** obarun has joined #crux19:42
frinnstBurnZeZ: not stripping debug symbols quickly gets insanely messy19:42
frinnstyou need to rebuild with it enabled19:42
frinnsti remember when i enabled debug in pkgmk.conf for some reason. forgot to remove it when I had done what i was doing.. down the line i rebuilt everything. rootfs grew *quite* a bit19:43
pedjainteresting question "how much swap for a machine with 512gb of ram?"19:50
pedjaI am guessing the answer is "it depends" :)19:51
duskmourn2is there anyway to run prt-get with -j5?20:14
TimB_duskmourn2: edit /etc/pkgmk.conf20:14
TimB_there is JOBS and MAKEFLAGS which you want20:15
cruxbot[core.git/3.5]: sed: update to 4.820:19
cruxbot[core.git/3.5]: sysklogd: force serial build20:19
duskmourn2ah20:25
*** pedja has quit IRC20:40
*** SovietPony has quit IRC21:01
*** SovietPony has joined #crux21:01
stenurdaftaupe: dunno really. It may have something to do with a 2 second "pre-gap" that is on Audio-CDs. Can be seen on my cdio bug report https://savannah.gnu.org/bugs/?57550. The real problem is that the leadout info is way off. Why they read two different lead-out informations, i do not know.22:13
stenurpedja: yep, they have removed it.  Only their native thing remains.22:13
stenurTimB_: maybe their native GUI is bug-compatible :)  I have no idea :-)22:14
TimB_dunno :)22:26
*** duskmourn2 has quit IRC22:42
*** Workster has joined #crux23:31
*** SiFuh has quit IRC23:48

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