@denisbloodnok @siina regardless, #SystemD does boot faster simply because it doesn't go strictly linear but parallelizes startups and does dependency resolution on services.

• Which is way more convenient than #SystemVinit.

Cuz original #SysVinit s slow af.

• The only reason it's not feeling slow anymore is because we don't rock 2,5" 5400rpm IDE HDDs in laptops anymore but even most "budget" SSDs can saturate SATA-6G during reads, if not exceed it even in worst-case scenarios.
  • And the only reason I didn't put systemd into @OS1337 as of now is the fact that I can't make it fit!

Now I see why some people shit on systemd. For simple things, it's needlessly complex haha. I mean I already did something like this last year where I tried out other inits and I'd say sysvinit is one of the most complicated inits I've ever used when I played around with it.

It's WAY older than me. I'm only 23 years old and that thing is like 30+ years old

#voidlinux #runit #systemd #archlinux #cachyos #void #linux #freesoftware #opensource #sysvinit

Now I see why some people shit on systemd. For simple things, it's needlessly complex haha. I mean I already did something like this last year where I tried out other inits and I'd say sysvinit is one of the most complicated inits I've ever used when I played around with it.

It's WAY older than me. I'm only 23 years old and that thing is like 30+ years old

#voidlinux #runit #systemd #archlinux #cachyos #void #linux #freesoftware #opensource #sysvinit

@Zenie @rl_dane

On this #FreeBSD laptop, I just use startx to run #openbox. I've done that on plenty of #Debian systems in the past....even in its lighter days with #sysvinit. It it helpful with multiple users, however.

As for #systemD, if it were just an init manager I'd have less of a problem with it. Creating "unit files" and where things are located are convoluted at best.

SystemD has its tentacles in other traditionally non-init things like /etc/resolv.conf and others. That's where my biggest problem lies.

@DrHyde @glitzersachen @rl_dane

I remember that and still viscerally recall the firestorm that broke out on the #Debian mailing lists.

I am not a fan of #SystemD. While I recognize that Linux's implementation #SysVInit needed replacing, SystemD was not the right solution.

Even if there is some kind of "correctness" to it, the design is not based on #empathy or nor did it improve upon the user experience. It is a case study in the failure of not using design thinking.

We went from:

service $SERVICE start

To:

systemctl start $SERVICE

Why couldn't it be:

start $SERVICE

Of course, the additional info when checking a service's status is valuable and certainly an improvement, HOW it was implement is my biggest criticism. Also, SystemD sprawl. Poettering's terrible attitude didn't help either.

#designthinking #userexperience #design #ux

@jasongorman nodds in agreement

I think "#architecture" and "#design" can be two entirely different things.

• I.e. the architecture of #macOS is a cobbled clusterfuck to a degree, but #Apple did nail the #design (until the #LiquidAss update) to be a good #UI & #UX, and they have a ton of #accessibility options integrated into it.

Similarly #SystemD is at worst a "necessary evil" as those that architected #SysVinit refused to modernize it.

• Waiting 5 minutes to bootup to a #Desktop may have been acceptable in the 2000s when everyone ran dogshit-slow 5400rpm HDDs in their laptops, but those get steamrolled even by cheap microSD cards these days!

#GnuPG is an example of horrendous design to the point that everyone I know either uses a replacement like #enc or a #GUI like #Kleopatra.

• Showing that said design and architecture are decoupled from each other.
• Also it's the prime showcase for everyone saying '#FLOSS can't do #UI!' when there are dozens of good distros that just work...