Re: A dumb question

From: Francisco Gómez García <espectalll_at_kydara.com>
Date: Tue, 02 May 2017 09:30:44 +0000

May 2, 2017 12:08 AM, "Alex Efros" <powerman_at_powerman.name> wrote: > Hi! Hey! First of all, thanks for the quick, polite reply. Didn't expect to be written in days, actually - so that on itself is already a good-enough greeting. > That sounds like everyone has own preference and that's ok. If you > like complexity and issues - use systemd, that's what it's for. It > works too, more or less, so you've such an option. Most people on the init debate I've seen actually think that they're saving effort because Systemd is such a complex project that can carry heavy tasks and wisely assign system resources to every daemon, allowing an ease of use and a stable-enough system that would otherwise be impossible. I thought so before trying out Runit, especially since my previous OpenRC experience was not-so-good and systems like Debian make Systemd a dependency even for Xorg. Then again, most people prefer using Windows. > Most of djb's and djb-like software (with notable exception of qmail) > is small, simple, does just one thing, does it completely enough, > does it reliable enough… and thus it doesn't needs to be updated, > fixed or even improved. Because if you needs it to be improved then > usually that means you just need some other software, not that one. Of course, the epitomy of the UNIX philosophy. Yet not everyone seems to think that it is the right model for controlling their systems, and they'd rather have something that keeps on giving, either bugs or features, in a complete package. A matter of taste? > Runit may have some very minor issues, which no one is interested enough to fix, but in general it does it's work very well I have, in fact, bothered to take a quick look at the source code. It does have some XXX comments indicating assumptions, warnings and a few things that should be coded, which basically means that the program is not "finished" as in an error-free, 100% reliable-by-itself piece of software. Perhaps that could give enough work to keep the project releasing a few more versions, as "good enough" is still not "perfect". > and it doesn't matter how many years ago it was released if it still > does everything we needs today without any issues. Actually I find > this very cool, because it's just one thing less to bother about and > to re-learn every few years because it's buggy and/or changes too > fast under my fingers. Personally, I don't care at all about the ever-changing world of software making yet another change. But I do see the value for those who would rather have a defined set of tools. In this case, it's a matter of preferences as well. > I don't see anything wrong with such a question. People gets used to > software which wasn't designed well and/or try to do many tasks at > once and thus they doesn't expect anything good from software which > doesn't looks like it fix a couple of bugs every week - because they > know there are bugs in every software, and they know they'll need > active support to fix bugs they'll find… and thus without such > support using software isn't an option at all. Runit, and few other > notable software, is different - it just works, and works well. It > doesn't do everything you may need (just like all other similar > software), but everything it does it does well and it functionality > is more than enough in many use cases - no necessarily in your use > case, of course (take a look at s6 if you'll need more features). Exactly! I do, however, come from being shocked again and again of software becoming heavier, more centralized and not proportionally useful or stable enough. I certainly don't hate Systemd, but neither consider it "good design" or the ideal by any means. And so I'm glad Runit exists, I'm really impressed by it and so I don't want to abandon it because "it's dead". Good to know there's s6 as well, chances are I'll end up using it in the future. For now, though, runsv services do the job. Thank you for the help, Alex! Best regards! Francisco / @espectalll
Received on Tue May 02 2017 - 09:30:44 UTC

This archive was generated by hypermail 2.3.0 : Sun May 09 2021 - 19:44:19 UTC