#qi-hardware IRC log for Sunday, 2017-04-23

DocScrutinizer05wpwrak: it's not a mere technicality as soon as you want to replace libsystemd0 by a dummy04:34
DocScrutinizer05technically lsusb may depend on udev, but not on larger systemd-* src04:35
DocScrutinizer05but according to Pali it's no longer possible to build lsusb for maemo, thanks to missing systemd04:38
DocScrutinizer05this sounds trivial but there was more to it than just "when it depends on XY, then it needs XY"04:39
DocScrutinizer05with the systemd-cabal douchebags I have a hard time trying "never attribute to malice..."04:41
infobotit has been said that nosystemd is https://devuan.org http://without-systemd.org/wiki/index.php/Arguments_against_systemd http://blog.darknedgy.net/technology/2015/10/11/0/04:43
DocScrutinizer05mth: the question really isn't which is better, usurpd init or sysvinit. The problem is: with poetterware you have no choice anymore04:45
DocScrutinizer05the choice you got is basically devuan and that took over 2 years to reach 1st release candidate. Up to anybody's guess how long it would take a single admin to clean out systemd cancer from their system and then even maintain it04:47
DocScrutinizer05yes, you can replace PID1 by another init process. Still you need systemd running as a parocess for all the attached subsystems like udev, cron, ntp, whatnot04:49
DocScrutinizer05and systemd cabal not only feature creeping, they also forcefeed-introduce not-so-subtle semantic changes. (systemd-)Cron suddenly works like anacron, running missed events on resume from suspend, sth that cron never did by itself. login now terminating *all* processes the user started, on logout (nohup needs a patch to tell systemd not to kill it). Resolver now silently uses when there's no working config, NTP also changed 04:54
DocScrutinizer05in a weird way. and so on and on04:54
whitequarkDocScrutinizer05: you don't need systemd running for systemd-udevd to run12:26
whitequarkthat's what I do on my debian system right now...12:26
wpwrakincredible. gnat attacked at mid-day. those beasts know no limits.16:58
whitequarkyeah, the FOSS Ada tooling is bad17:14
Action: DocScrutinizer05 missed that joke20:45
DocScrutinizer05whitequark: thanks for clarification about udev. good to know20:46
whitequarkDocScrutinizer05: gnat is the gnu ada compiler20:47
whitequarkand it's bad20:47
DocScrutinizer05aah :-)20:47
Action: DocScrutinizer05 wonders if Ada is worth having a learning session and evaluating it. Heard high security projects like e.g. deep space missions and nuclear power plants use Ada a lot, dor some of its properties20:50
DocScrutinizer05then otoh all devels seem to hate it20:51
whitequarkDocScrutinizer05: Ada isn't half bad but it has a LOT of ceremony21:16
whitequarkand there aren't any good FOSS compilers (or really any except gnat)21:16
whitequarkso it's not really usable in the broader FOSS scene21:16
DocScrutinizer05I see. Thanks21:17
DocScrutinizer05Yeah, I guess Ada compilers need heavy and expensive certification21:17
DocScrutinizer05so this would explain why not so much FOSS support for them21:18
DocScrutinizer05any commercial enterprise wouldn't use and thus not improve/contribute to gnat21:19
whitequarkno idea21:21
whitequarkI think Ada is just very niche and maybe nearly dead21:21
whitequarkbarely anyone outside of DoD uses it21:21
--- Mon Apr 24 201700:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!