Path: csiph.com!v102.xanadu-bbs.net!xanadu-bbs.net!feeder.erje.net!us.feeder.erje.net!newsfeed.fsmpi.rwth-aachen.de!news-1.dfn.de!news.dfn.de!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: Tim Watts Newsgroups: comp.os.linux.misc Subject: Re: What distros *don't* use systemd ? Date: Mon, 08 Sep 2014 19:05:26 +0100 Lines: 67 Message-ID: <6o71eb-dgc.ln1@squidward.local.dionic.net> References: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Trace: individual.net ScZNAHIgotPfv3L7JnaGFgIA+OOpZ+nOhPPygm2KFN3w08ytc9 X-Orig-Path: squidward.local.dionic.net!not-for-mail Cancel-Lock: sha1:ec/NIt1Hw+XX9EkWzNWNwe7LbhU= User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.2.0 In-Reply-To: Xref: csiph.com comp.os.linux.misc:12038 On 08/09/14 18:41, Dan Espen wrote: > What I'm seeing is a bunch of complainers that can't seem to > demonstrate that they have better ideas. My "better idea" is to leave sysvinit alone :) Making things worse and going away from "unix is lots of small components, each one doing one job well" is not progress! >> My opinion regarding systemd is not so much "if it ain't broke don't >> fix it" as "it works pretty well, so don't replace it with something >> complicated, heavy, with binary formats and where the lead devs are >> rude to the users." > > I'm going with identify the problem and then solve it they best > way you know how. Simplicity should certainly be a goal, both in > the software and the user interface. I find systemd quite simple > in the user interface. I have't looked at the code so I'm unable > to judge whether the software is overly complicated. > > But I don't see anyone here making the case that the code > is overly complicated. I think several people have stated the bloat and the "all things in one place" as a problem. I tend to agree it is not a good thing. > In fact one person just claimed that > the old init system was overly complicated. sysvinit /sbin/init is pretty simple. (With apologies for granda egg sucking - this is just to illustrate my point) It runs some runs in /etc/inittab and one of those rules (on Debian) runs a script /etc/init.d/rc on runlevel changes. You could argue that the scripts are complicated, though the debian ones seem clean. This percolates through ultimately to the /etc/rcN.d/(S|K)NNthing symlinks. /sbin/init of course is not tainted by this - it is pretty well defined. And simpler embedded hardware may have a much stripped down series of scripts from /etc/init.d/rc or whatever. It's massively flexible, and leaves the complicated stuff in easily hackable scripts. >> I would be interested in a side discussion of what's wrong with >> sysvinit and how it could be improved incrementally... > > Fine. Bear in mind, that's what the systemd developers did. > Looks to me like they identified some fundamental issues > and came up with something not incremental. > Well, if anyone's game to throw up something about the sysvinit way that pisses them off, I'll have a go at saying how I might address it. I don't have to agree with you that it's a problem - I'll just try and address some hypothetical issues for fun...