IRC Logs for #crux Monday, 2019-07-15

*** joacim has quit IRC00:17
*** joacim has joined #crux00:19
*** obarun has joined #crux00:25
*** t58 has quit IRC01:41
*** tilman_ has joined #crux02:22
*** _________mavric6 has quit IRC02:49
*** _________mavric6 has joined #crux02:50
*** weednix has quit IRC04:32
*** kori has quit IRC05:11
*** kori has joined #crux05:24
*** kori has quit IRC05:24
*** kori has joined #crux05:24
*** kori has quit IRC05:51
*** kori has joined #crux06:11
*** kori has quit IRC06:11
*** kori has joined #crux06:11
*** kori has quit IRC06:16
*** kori has joined #crux06:38
*** kori has quit IRC06:44
*** deep42thought has joined #crux06:46
deep42thoughtHi, I'm in the process of upgrading to crux 3.5 and some packages fail to build due to signature mismatches - one example being qt5 from opt. Did I miss some update of a key?06:47
*** kori has joined #crux06:49
*** tracer has quit IRC06:50
*** john_cephalopoda has quit IRC06:50
deep42thoughttek__: also it would be nice if you could update xlockmore - the 5.56 source is unavailable (superseeded by version 5.57)07:15
*** john_cephalopoda has joined #crux07:16
*** ryuo has joined #crux07:16
*** tracer has joined #crux07:16
TimB_it would be nice if he would reply at all <.<07:16
deep42thoughtnah, no need to reply - just need to fix ;-)07:16
TimB_yeah, he is not doing that either07:17
TimB_and :P07:19
deep42thoughtah, right, I wanted to check your libreoffice-from-source (but needed/wanted to update to 3.5 first) ... thanks for the (indirect) reminder07:20
TimB_you can build it with 3.4, but you need to adjust the python version where needed (e.g. py3boost)07:20
deep42thoughtbut since I want to update to 3.5 anyways, I thought, this is the better order ;-)07:21
*** jolupa has joined #crux07:21
TimB_I agree ^^07:21
TimB_deep42thought: It seems that qt5 has the wrong signature right now :) ping Romster07:23
*** kori has quit IRC07:26
*** slek has joined #crux07:38
deep42thoughtRomster: I'm also unable to compile xmlto - it fails with "xsl:import : unable to load"07:41
deep42thoughtso: besides qt5 and xdg-utils (depending on xmlto), I'm now on crux 3.5 \o/07:42
TimB_have you run revdep?07:42
deep42thoughtI have, and after recompiling, only the usual suspects stay around (plus some more)07:43
deep42thoughteagle, gexiv2, libreoffice, pidgin, qt5, rust-bin, texlive, ufraw07:43
deep42thoughterr, these were less the last time I checked ...07:43
TimB_those are the usual suspects for you?07:43
deep42thoughtare the usual suspects07:44
deep42thoughtpidgin links to which is in a funny place07:44
TimB_for texlive you can try my updated port :) tek only updates what he still is interested it in seems07:46
*** kori has joined #crux07:48
*** elderK has joined #crux07:48
elderKHey guys, I was wondering what a good way to "kick off a new install" is, without actually nuking everything.07:49
elderKI read the "chroot" page in the WIki. I was wondering, can we kick off "setup" from the install media, inside a chroot on a "normal running system"?07:49
elderKAlso, I'm appreciate advice on how to later move that "new system" in the chroot to a real system, waiting.07:50
elderKI don't know all that much about copying UNIX filesystems from one place to another - I understand there'd be complexity due to permissions, symlinks, device files, etc. So, advice would be great.07:50
elderKAlso - yet more questions: Any good way to keep multiple systems effectively in sync? Without say, serving them all over NFS or something like that?07:51
*** kori has quit IRC07:56
TimB_elderK: not sure if I understand you correctly. you can bootstrap prt-get and pkgutils from somewhere else and prepare a CRUX install with that07:56
TimB_but moving the chroot? you normaly mount something on mount to use it as a future root fs, you don't need to move it afterwards?07:57
elderKTimB_: I was more thinking along the lines of setting up a /mnt/newsys and using the setup program from the media to install the core packages I need into that, then chrooting into that, and doing the rest.07:57
TimB_and what do you want to keep in sync? ports?07:57
deep42thoughtelderK: The question is: what do you plan to do with a system in a chroot?07:57
TimB_elderK: you can do that I think07:57
elderKdeep42thought: It's only to bring up a "clean install" while still being able to do all the things I'd normally do.07:58
deep42thoughtso you want to minimize downtime07:58
elderKIdeally, I could also copy that "new system" to another laptop, too, and use it there.07:58
elderKI'm okay with recompiling the kernel for that new system, say. They're not that much different, anyway.07:59
elderKBut I'd like to basically bring the system up completely, before I'd copy it. That way, I could avoid the recompiles, etc.07:59
deep42thoughtyes, you can do the recompiles in the chroot07:59
deep42thoughtso you're talking about an upgrade, not a fresh install, right?08:00
elderKNo, fresh install.08:00
elderKJust in a chroot - then I reboot and use that "new install" for real :P08:00
TimB_that should work08:00
deep42thoughtok, this should work out: you install to a chroot, update everything there and finally wipe your old system and move the content of the croot in08:00
elderKdeep42thought: How would that handle symlinks and stuff?08:01
elderKCould I use something like rsync to do the copy?08:01
*** kori has joined #crux08:01
*** kori has quit IRC08:01
*** kori has joined #crux08:01
deep42thoughtassuming, the file systems are identical (both ext4 or whatever), `rsync -a` should be sufficient08:01
deep42thoughtyou may need to regenerate /etc/fstab and such08:01
deep42thought(and kernel/initramdisk)08:02
elderKWhat do you use for initrd for Crux? I generally just boot via EFISTUB.08:02
deep42thoughtI have no initrd08:02
deep42thoughtI just stuff everything into the kernel08:03
elderKI do too, after a short time of having something "seemingly optional" as a mdoule.08:03
elderKIf it's used a lot, I build it in. Otherwise, I leave it a module.08:03
elderKAlso stuff like VirtualBox.08:03
*** kori has quit IRC08:07
elderKOut of curiosity, does anyone here even "clone" their systems? Clonezilla / Ghost style?08:07
TimB_not me08:08
elderKI've got to find a way to recalibrate my battery, too. Only charging to 78% UNK nowadays.08:11
elderKThanks for the advice guyS :)08:15
elderKSince I have two laptops, I should learn to use ccache + distcc08:16
elderKAnd prt-cache, is it?08:16
elderKor some way of reusing the packages that wind up built.08:16
TimB_ccache helps when you have to compile a package over and over again for some reason (packaging maybe)08:19
TimB_distcc distributes the build to machines on the network08:20
deep42thoughtshared /usr/ports/packages is probably what you are looking fore08:20
TimB_prt-cache just builds a db for all the ports you give prt-get to work with and performs lookups and stuff a bit faster08:20
elderKThanks :) You'll probably see me tomorrow with more questions. Time to shower and get ready for bed - work tomorrow.08:22
elderKHave a good night, guys :)08:22
deep42thoughtgood night!08:22
*** john_cephalopoda has quit IRC08:26
*** ryuo has quit IRC08:26
*** tracer has quit IRC08:26
*** kori has joined #crux08:28
*** kori has quit IRC08:28
*** kori has joined #crux08:28
TimB_good night elderK08:30
Romsterdeep42thought>, make sure you got docbook-xsl and docbook-xml and run post-install scripts and rejmerge08:33
deep42thoughtarrgh, damn post-install scripts, again08:33
Romsterfor xmlto08:33
deep42thoughtsry to have bothered you08:33
*** tracer has joined #crux08:35
*** john_cephalopoda has joined #crux08:36
*** ryuo has joined #crux08:37
cruxbot[opt.git/3.5]: qt5: fix signature08:37
Romsterand oops on qt508:39
Romsterit's fine deep42thought08:39
Romsteri hit that yesterday08:39
deep42thoughtI added an alias - let's see when this breaks :-)08:39
Romsteryou could just enable it in prt-get.conf08:39
deep42thoughtwhy didn't I see that option before?08:40
Romsterif you trust them of course08:40
deep42thoughtlet's pretend, I do trust them :-)08:41
Romsteri still got some broken things to fix08:41
deep42thoughtTimB_: your texlive fails to build, because it does not find gd09:05
*** kori has quit IRC09:07
Romsterlibgd is in opt deep42thought09:08
deep42thoughtit is installed09:08
Romsterprt-get deptree texlive09:08
deep42thoughtlooks all fine09:08
TimB_deep42thought: let me know what else you find?09:09
TimB_s/?// :P09:09
Romster has --disable-rpath09:09
deep42thoughtoh, I need to build all the deps from your repository?09:10
TimB_Romster: do we need that?09:10
deep42thoughtI thought, official deps would work, too09:10
TimB_I am not sure, might be snatched from slackware or blfs or something09:10
Romsternot a big deal TimB_ we run ports at the same path09:11
TimB_so rpath would affect stuff in /opt or /usr/local?09:11
Romstermore for binary distribution of where stuff looks09:11
Romsterpretty much09:11
TimB_ah, I see09:12
TimB_I could probably delete the duplicate09:12
Romsterno rpath makes it able to even run from PATH=/home/foo09:12
TimB_oh, I see09:12
Romsteryeah no biggie there09:12
Romsterit adds a layer of protection for say user injected libraries on the path09:13
TimB_understood, okay09:13
Romsterfor those sorts of critical things09:14
TimB_have you seen I sorted libindicator stuff out?09:14
TimB_need to merge that into the steam repo still09:14
Romsternot yet, i am poking at compat-32 currently, but i wanna try steam-native09:14
Romsterand look at merging into contrib/compat-3209:15
TimB_great news :) If you would give me a key I would keep them up2date or help with compat-32 in general09:15
TimB_but I still need to polish them out in a container a bit more09:15
TimB_it's not yet done09:16
TimB_maybe I'll release another snapshot of the build and promote it on the ML as well for "beta testers" :)09:16
TimB_I think it's safe to try it at this stage for sure09:16
TimB_ <- look at the list just for ubuntu indicators XD09:17
Romsteri regret clicking that link09:17
TimB_what, why?09:17
Romsterthat many parts09:17
TimB_they are rather small packages really09:17
Romsteri see09:18
TimB_and it's just 3 ports to be exact09:18
TimB_but they share files so you need to defiddle them...09:18
TimB_xfce and mate use only gtk3 parts09:18
TimB_probably other DEs as well09:18
Romsterno qt ones?09:18
TimB_uhm.. not sure if there is, I think there isn't09:18
TimB_those are ubuntus indicators09:18
TimB_I think they are gtk only09:19
Romsteri've never looked at them is that the way to have tray icons?09:20
TimB_for some stuff it seems09:20
Romsteri'm using tint2 here09:21
TimB_I am not sure if nm-applet works without it..09:21
TimB_would need to test if I get _any_ icon when I remove that all09:21
Romsterthat be interesting09:21
Romsteri got more ports to do dog on my arm asleep and i need food09:22
TimB_steam-native does depend on sharp bindings though -.- which is great09:22
TimB_alright, I think I'll prepare a merge to the steam git and try to focus on some uni stuff after that. :)09:23
Romstermust move dog and cook food but crux09:23
*** kori has joined #crux09:28
*** kori has quit IRC09:28
*** kori has joined #crux09:28
TimB_steam-native repo is updated09:43
TimB_ is updated as well09:44
TimB_also: signed ports now :)09:44
TimB_ for easier access09:46
TimB_and posted to the ML to maybe get some feedback09:59
*** elderK has quit IRC10:28
TimB_Romster: seems like nm-applet and redshift don'10:45
TimB_t need indicators to show a tray icon10:45
TimB_works fine for me10:46
TimB_claws also shows without them10:46
TimB_mate-panel works with them as well without them...10:47
*** jolupa has quit IRC10:52
*** t58 has joined #crux10:55
cruxbot[compat-32.git/3.5]: xorg-libxt-32: 1.1.5 -> 1.2.011:50
cruxbot[compat-32.git/3.5]: xorg-libpciaccess-32: 0.14 -> 0.1511:50
cruxbot[compat-32.git/3.5]: util-linux-32: 2.33.2 -> 2.3411:50
cruxbot[compat-32.git/3.5]: nettle-32: 3.4.1 -> 3.5.111:50
cruxbot[compat-32.git/3.5]: mesa3d-32: 19.1.0 -> 19.1.211:50
cruxbot[compat-32.git/3.5]: libinput-32: 1.13.2 -> 1.13.411:50
cruxbot[compat-32.git/3.5]: libdrm-32: 2.4.98 -> 2.4.9911:50
cruxbot[compat-32.git/3.5]: harfbuzz-32: 2.5.1 -> 2.5.311:50
cruxbot[compat-32.git/3.5]: freetype-32: 2.10.0 -> 2.10.111:50
TimB_sudo ports -u compat-32 D:11:51
TimB_my cpu was already thinking compiling time was over =)11:52
Romstersorry :D11:52
Romsteri've beena  bit behind11:52
Romsterno one seems to notice11:52
TimB_did you get the patch I prepared for your romster repo? ^^11:54
Romsteri got the paste yeah11:54
Romsteri haven't got to that yet11:55
Romsteri did touch a few other ports though but i need to sync them to my server11:55
TimB_alright, just so it doesn't get lost. I only went through stuff that I had installed and needed a bump or something11:57
Romsterthose are quite fast to compile except for mesa3d-3211:57
Romsteryeah i need to go though the entire repo at some point11:57
TimB_next month I have some more time, maybe we want to pickup what we talked about some weeks/months ago11:58
Romsterbeen thinking of that at the back of my mind11:58
TimB_right now, I am quiet happy with what I got. lo seems done, steam is almost done. :)11:59
Romsteri was gonan attmept that but you've done all the hard work for me.11:59
TimB_ck4up is set up for most of the stuff (what's not on github, anyway)11:59
TimB_Romster: but I was not working hard on CRUX for the past years, so that's okay if you ask me12:00
Romstertrue i guess so, i've been here since crux 2.3/2.3 might of even been 2.1, i found some old burned cd-r's with texter of 2.2 on one of them. think i seen a 2.1 here12:02
TimB_I don't have any of that ancient stuff flying around =D12:03
TimB_I was amazed to find myself in the irc logs that are linked in \topic12:04
Romsteri should toss those old ones out12:04
TimB_10 years back12:04
Romstertimcowchip was it? or was it another name12:04
TimB_no ^^12:05
TimB_RyoS I think it was12:05
Romsterah right you were almost with ryuo12:05
TimB_time to get ready for work12:15
*** slek has quit IRC12:53
*** slek has joined #crux12:55
*** deep42thought has left #crux ()12:58
*** nthwyatt has joined #crux12:59
*** jue has joined #crux13:00
*** darfo has quit IRC13:02
*** darfo has joined #crux13:17
*** nthwyatt has quit IRC13:19
*** nthwyatt has joined #crux13:19
*** darfo has quit IRC13:21
*** darfo has joined #crux13:21
*** nthwyatt has quit IRC13:24
*** deep42thought has joined #crux13:28
deep42thoughtgexiv2 complains about a missing exiv2/xmp.hpp - and exiv2 indeed comes without that file13:28
deep42thoughtam I missing install-scripts again?13:29
*** nthwyatt has joined #crux13:33
*** darfo has quit IRC13:36
*** tracer has quit IRC13:37
*** tracer has joined #crux13:37
*** darfo has joined #crux13:41
*** nthwyatt has quit IRC13:43
deep42thoughtlooks, like the port is just outdated - a recent version compiles fine (with help from arch linux)13:52
*** deep42thought has quit IRC14:08
cruxbot[opt.git/3.5]: wine: 4.11 -> 4.12.114:15
*** obarun has quit IRC14:29
cruxbot[opt.git/3.5]: fuse3: update to 3.6.214:33
cruxbot[opt.git/3.5]: dovecot: update to 2.3.714:33
cruxbot[opt.git/3.5]: msmtp: update to 1.8.514:40
cruxbot[opt.git/3.5]: meson: 0.50.1 -> 0.51.114:41
*** irclogger_do has quit IRC14:44
*** irclogger_do has joined #crux14:44
*** obarun has joined #crux14:52
*** tracer has quit IRC15:38
*** deep42thought has joined #crux16:23
deep42thoughtHi, with crux 3.5 my boot process takes a lot longer - it looks like it's waiting for "crng" - at least, there is after a long break "[  204.718826] random: crng init done" in dmesg and then the daemons get started. Might this be related to the upgrade to 3.5 or is this something else?16:25
TimB_deep42thought: I haven't hit that yet, no idea16:26
TimB_gexiv2 is also in flyspray, but tek__ is not responding o:)16:27
*** jolupa has joined #crux16:27
deep42thoughtTimB_: I upgraded gexiv2 manually16:27
deep42thoughtand put it into my ports collection16:27
TimB_yeah, I understood that16:27
TimB_but if it were fixed upstream you wouldn't have to do that :)16:28
TimB_now who knows how much people fork gexiv until it's fixed?16:28
deep42thoughtyes :-)16:28
deep42thoughtit's all on for others to find and copy ;-)16:28
TimB_ Richard Dunbar posted that some days ago16:29
TimB_that's what I was refering to16:29
stenurNew kernel RNG policy; i have posted a program to -devel:
TimB_why am I not registered to devel yet..?16:32
stenurShould be placed even earlier indeed, before udev loading (if possible)16:32
deep42thoughtI'm not sure if /etc/rc is executed already when it hangs16:33
deep42thoughtoh, well, yes _that_ one is being executed :-)16:33
deep42thoughtwhere do I get /sbin/entropy-saver ?16:35
stenurYep.  It is attached to the mail, or you can download it from my toolbox.git; i do not have a port for it, hmm. Should be part of core/ :)16:36
stenurWait, the URL ...16:36
*** kyaaa has quit IRC16:37
deep42thoughtgot it :-)16:37
stenurThat should also be it.16:38
stenurThat is, i do not know whether newer Linux tweaked their RNG init again.16:40
deep42thoughtthe funny thing is, that I did not actually update my kernel16:40
stenurFor 4.19 it is definetely an improvement.16:40
deep42thoughtI was at 5.2 and merely upgraded to 5.2.116:40
stenurMaybe they blacklisted the AMD CPUs because of that CPU bug?16:41
deep42thoughtI have an intel :-/16:41
stenurThe RAND init hardware support at least16:41
stenurWell, then, i do not know, sorry.16:41
deep42thoughtlet's see if it does the job :-)16:42
*** deep42thought has quit IRC16:42
*** deep42thought has joined #crux16:44
deep42thoughtit comes too late16:44
deep42thoughte.g. it comes after "crng init done" (before which the boot process hangs)16:45
deep42thoughtbut since I now know, that I "simply" have to hit some keys, I can always do _that_ ;-)16:45
stenurYour boot hangs in the kernel???16:46
*** kyaaa has joined #crux16:46
deep42thoughtit hangs before "[   14.536360] random: crng init done"16:47
deep42thoughtand after "[    4.239874] sd 6:0:0:3: [sdg] Attached SCSI removable disk" (probably unrelated)16:47
deep42thought(these are dmesg messages)16:47
deep42thoughtalso interesting: right the next line in dmesg is "[   14.537201] random: 7 urandom warning(s) missed due to ratelimiting"16:48
stenurNormal, but urandom does not hang (i have three from udev, but i could move to in /etc/rc to get rid of that)16:49
stenur(I could move above udev init that is.)16:49
deep42thoughtI don't know _who_ waits for "crng init done"16:49
deep42thoughtI'll try again with moved-up "load" clause ... let's see ...16:52
*** deep42thought has quit IRC16:52
*** deep42thought has joined #crux16:54
deep42thoughtyup, when moving to the top of /etc/rc, it does it's job as intended :-)16:54
stenurGood. I really do not understand why they changed that. I suddenly had hangs of half on our on my VM last year.16:57
stenurOn FreeBSD or OpenBSD that would have been pointed out loudly, but in Linux with only a git log, or not even that, i was totally stunned.16:58
stenurIn fact i fed in data like grazy into urandom first .. but it did not change anything..  Weird.16:59
stenur(as in, changing the boot script to feed on much more.)16:59
deep42thoughtI think, writing to /dev/urandom does on increase the entropy counter (for quite some time)17:01
stenurI did not know that ... not even initial seeding!17:01
*** jolupa has quit IRC17:17
deep42thoughtoops, meant to say "... does not increase ..."17:33
pedjahaveged ftw17:48
deep42thoughtthe problem is, that daemons are not started yet at that point17:49
deep42thoughtso haveged will not help17:50
stenurhaveged counteracts that random promise that Linux RNG now seems to try to give, no? It produces thousands of random bits a second...17:50
stenurI have it on the server VM though, there mailman is running and that slurps random bits like grazy, even without work. wtf17:51
stenurThat is.. i have no idea; rather more than good enough for anything, in practice, anyway.17:52
stenurIt's just that it is grazy that the good stuff is thrown away / not used, so that people use that haveged "hot air". Weird..17:53
deep42thoughtI once read a pretty good article about haveged - my conlusion from it was: entropy may be only missing in the beginning (after boot) - "entropy" afterwards makes hardly any sense: either your pseudo-rng is good and you can't guess the internal state from the random output, thus you do not _need_ more entropy - or your pseudo-rng is bad and you're screwed anyways17:56
pedja <--thought of an actual cryptographer about haveged17:58
deep42thoughtwhat does dieharder say to pure mersenne-twister output?18:00
deep42thoughtI guess, it does classify it as good randomness, too18:00
pedjaif the German BIS says it's OK, that's good enough for me :)18:02
deep42thoughtthat's fine for me, too18:03
deep42thoughtI just wanted to say, that the sole output of dieharder is hardly convincing ;-)18:03
pedjawell, there is this
stenurAnyway, even after reading all that (again), not me, no.  Only because the VM needs it.18:18
pedja(from anti-cheat thingie)18:35
stenurTotally zero??18:44
*** t58_ has joined #crux19:03
*** deep42thought has left #crux ()19:04
*** t58 has quit IRC19:05
*** zimmer_Dl has joined #crux20:17
*** sammi`_ has joined #crux21:18
*** sammi` has quit IRC21:20
*** nthwyatt has joined #crux21:47
*** darfo has quit IRC21:50
*** nthwyatt has quit IRC21:52
*** darfo has joined #crux21:53
*** slek has quit IRC21:57
*** nthwyatt has joined #crux22:05
*** darfo has quit IRC22:07
*** nthwyatt has quit IRC22:08
*** darfo has joined #crux22:09
*** jakelyz has joined #crux22:48
*** jakelyz has quit IRC22:52
cruxbot[compat-32.git/3.5]: harfbuzz-32: disable graphite222:53
*** jakelyz has joined #crux22:53
*** john_cephalopoda has quit IRC23:05
*** john_cephalopoda has joined #crux23:19

Generated by 2.14.0 by Marius Gedminas - find it at!