Posts and replies

(and if you're not one of them, you'd do well to not get involved with the project.)

and, seriously, this is not salvageable, there's no salvaging a project managed by a nazi.

@denisbloodnok there are valid reasons to dislike systemd (even though, as a sysop, i tend to like its usability) – that's not the same as believing it was forced on the community by dastardly shenanigans.

@vekkq thank you for your valuable input.

@Netux @mawhrin

It depends on the reason. There are three kinds of people who oppose those these projects:

• People who don’t believe that the problems that they try to fix are real.
• People who oppose them because they believe they are part of a conspiracy by big Linux vendors to try to dominate the ecosystem.
• People who agree that the problems are real but think these projects, at best, solve a subset of the, while also introducing a load of new ones.

The last people are typically right. The middle people have some basis in fact (Red Hat and Canonical are not exactly keeping secret their desire to control the core platform) but rapidly descend into unhinged conspiracy theories. The first class are completely in denial.

@david_chisnall @mawhrin I'll still fight the idea that systemd fixed any real problem. Your laptop taking two minutes instead of thirty seconds to start is worth not having some binary build to start a new daemon and binary logs.

It may solve something else, but in twenty to thirty years I've yet to stumble across it.

@Netux

with the service managers like launchd or systemd, the boot time is purely a red herring; it's a nice effect, but not the reason they're designed the way they are. (some of these are: the complexity of the hardware, especially when dealing with hotplugging, and the need to dynamically respond to changing conditions; modern desktop or laptop computer, for many reasons, is a complex beast.)

i would not fixate on the on-disk log format either. i used to, a decade ago – but in all these years i've seen the maintainers make no attempts to remove the ability to forward the log data directly to syslog; and it's genuinely useful to have structured logs with more information available on hand.

(this is getting very advocacy-adjacent; @david_chisnall, let me know if i should drop you from further mentions.)

@mawhrin @david_chisnall hotplugging I can't speak to, I've never been anywhere it actual happens (unless you are talking about usb stuff on a user machine).

The other stuff always seemed to chug along fine. Mkdev is still used in the kernel, even if you don't invoke it yourself, init files were easy to write (if not easy to comprehend all the time)

It seems if it was that useful, at least one of the bsds would have ported it over.

@Netux @mawhrin

As I said, the problems that systemd addresses are real. Configuring a service that runs only when you’re on your home WiFi network, for example, is really annoying on FreeBSD because devd is not a source of events for a process manager, it’s a totally distinct thing.

Unfortunately, the way that it solves them is not great and the code quality is not impressive.

@david_chisnall @mawhrin do all of the xxd services require the scaffolding that systemd introduced, or are they all relativity independent but because they are all pottering related they all got the xxd name?

You have an example similar to what you posted above? I use my stuff as servers, same as work. Never contemplated having the laptop do anything trickier than picking the right wifi network or the ethernet. Even my work VPN does a network down-up, same as of I were using ifdown/up.

@cotton this is not surprising, considering everything else; if you have the link handy, i'll update the post.

@mawhrin@circumstances.run looks life you found something any way ^^ I think the comments I was looking for got deleted, and not archived on wayback.

anyways, Enrico is/was also a covid vaccine denier [lore.kernel.org]

@kris_of_pnictogen [all other things aside] seems like weigelt was not happy with the level of adoption of his patches by x.org, and the general criticism of the quality and usefulness of his code, and decided that he's being persecuted, and not for his lack of merit, but because the “big corps” wanted to suppress his attemps at what he felt was revival of x11 development.

my take is that there's probably a little that can be done with X11 architecture other than keep it on life support while getting wayland on par with features that still work better in X11 (accessibility being a primary concern).

@mawhrin I would love to see a better #X11 (I have some undue fondness for it, after years of college #ComputerScience classes that now seem useless because of trends) but yeah I think I'll be staying well away from this mess for now. We need soothing baby's first #programming stuff right now anyway. Like...maybe not even a PC. Wooden blocks perhaps or marbles. We have yet to map the extent of Kel's conceptual deficits fully, because o̦̒f͈͓ t͙ͮ̂hé͕̲͐̅͟ c̹̖̺̓̌̎͋̌u̢͓ͪb̧͍̐͗e͙͂̏̕͡s̡̙̥̓̑ͨ— (*at this point Akemi Homura teleports in and punches Chara in the solar plexus, knocking them flat and gasping for breath*)

HOMURA ~Madoka

...it's fine... ~Chara

@dzwiedziu rright. it's so nice that they tend to congregate.