[SOLVED] Is 2019 Still Too Soon For Intelligent Assessment of SystemD?

Originally Posted by philanc Regarding ksh, I thought that it was compatible with the Bourne shell (ie. for example an ash script would run ok in ksh

No, it isn't fully upwards compatible and unlike bash (which isn't either) it doesn't disable its "advanced" features when called as "/bin/sh", like bash does.
I always compare it a bit like C++ vs C, or python 3 vs python 2:
ksh started with most of Bourne shell's features and then tried to straighten the ugly things IN it.
But the later Posix shell is more of a real Bourne shell upgrade.

Originally Posted by hazel I'm currently wrestling with BLFS-9.0 (sysV version) and finding it quite dispiriting. I can't quickly put up a functional but minimalist system as I could with earlier versions because of all the new dependencies that have been introduced. They are necessary, apparently, to allow higher level software to run without systemd. This is definitely the last LFS that I do and I don't even know if I shall finish it. It just isn't fun any more.

I used to do LFS, but not BLFS. I always thought the point was to have a nice base system to build on, why would I want the rest dictated to me? (Moreover, I did not follow the LFS procedure exactly either, for example if I knew of a better version of something to use, or wanted to use more recent kernel headers I did).

You can build whatever you want, and not build whatever you don't want after the base system. Sure, if you want to use Plasma, you'll have to build 87 metric shit tonnes of dependencies. In other cases, many dependencies will be optional. If they aren't there, they won't be dependencies. For example you can build XFCE without linkage to all that gnome/freedesktop junk, especially if you only build the parts of it you want.

For example, the first thing I used to do was build image libraries, freetype, X etc., then a simple, (mostly, with as much --disabled as I could) dependency free fluxbox to work in (and fall back on while working on the system) while I got the rest of what I wanted done. While we'd want more than that, it's possible to have a simple, easy to maintain system. You can simplify maintenance by compartmentalizing some things too, when building it yourself. I still do that in Slackware and Manjaro, for example some things go to /usr/local and other complex things I set up their own --prefix (PATH and/or ldconfig etc.) for example /opt/trinity and /opt/firefox so they can be instantly "rm -r" 'ed

Sounds like too much work? I agree. That's why I don't do LFS anymore. Nowadays I'd rather have a more comprehensive distro like Slackware to build on and replace things as desired. :-)

With x86_64 the need to compile and optimize EVERYTHING yourself isn't as great, because even with generic x86_64 you still get instructions like SSE2 whereas on 32 bit, generic i386 or even i686 was a drag on newer processors. That's what I've observed, anyway.

Originally Posted by enorbet ??? Why would I do that? I truly do not grasp why people think uncomfortable things should fall into the "he who shall not be named" category like cancer, aids, teenage pregnancy and drug abuse and propose sticking our heads in the sand. Free Speech is the single most important freedom there is and I don't think it is at all wise to minimize and diffuse it by ignore-ance.

Because this has nothing to do with free speech and avoiding an uncomfortable topic. It's about positivity, productivity and responsibility. This thread has not been particularly positive, there's been a lot of bickering and we've now had a warning from Jeremy; it's not been particularly productive so what's left is responsibility - responsibility to the LQ community.

One of the things that makes this place great is productive conversation, respect for all members and when there is sense of usefulness and achievement in discussion. That is what open source is about - it's about the cause rather than the individual - group effort, group discovery and group advancement in the name of overall betterment, an interest to which this thread has not positively contributed. It was a notable effort to open the topic but any conversation on systemd, no matter how it's dressed up, will, for the foreseeable future, turn into an unwinnable argument that constantly chases its tail [I say unwinnable, actually the systemd side wins by default]. Practicing free speech should also be about respecting the argument and those involved and knowing when to stop talking.

Originally Posted by Lysander666 Because this has nothing to do with free speech and avoiding an uncomfortable topic. It's about positivity, productivity and responsibility. This thread has not been particularly positive, there's been a lot of bickering and we've now had a warning from Jeremy; it's not been particularly productive so what's left is responsibility - responsibility to the LQ community.

I somehow totally missed any warning from any mod let alone jeremy. I will search for his take and defer to his guidance.

Originally Posted by Lysander666 One of the things that makes this place great is productive conversation, respect for all members and when there is sense of usefulness and achievement in discussion. That is what open source is about - it's about the cause rather than the individual - group effort, group discovery and group advancement in the name of overall betterment, an interest to which this thread has not positively contributed. It was a notable effort to open the topic but any conversation on systemd, no matter how it's dressed up, will, for the foreseeable future, turn into an unwinnable argument that constantly chases its tail [I say unwinnable, actually the systemd side wins by default]. Practicing free speech should also be about respecting the argument and those involved and knowing when to stop talking.

I don't understand this line of thinking, Lysander666. I didn't create this thread in hopes that anyone would win anything other than possibly a little enlightenment. Here's the whole story. I marked the thread "Solved" when it seemed any actual experience within the Slackware sub section had been exhausted and tried to make a similar and referenced thread in Linux-General. That was not allowed which I view as unfortunate but not a major problem. Oddly enough, since marking it "Solved" and now with just shy of 20K views, we have had a mix of a few obnoxious posts (who were warned) but also a few of actual experience that has me knowing a bit more than I did when I posted the starter. Additionally I think it is useful to know that despite there being a few who have not grown in the interim decade, I think everyone has to admit it is not the rampant flame fest it was just a few years ago.

I sincerely do not see a problem.

Originally Posted by enorbet Additionally I think it is useful to know that despite there being a few who have not grown in the interim decade, I think everyone has to admit it is not the rampant flame fest it was just a few years ago. I sincerely do not see a problem.

In that case it's a matter of perspective. Maybe this thread is less heated than similar threads years ago about systemd. I have not been in the Linux world for that long, and I certainly came here after most of the fiery arguments on the topic had abated. I'm not having a go at you, but I am addressing the usefulness of discussing the topic. Though I suppose, one could argue, that's the water this new discussion came to test.

Originally Posted by Lysander666 In that case it's a matter of perspective. Maybe this thread is less heated than similar threads years ago about systemd. I have not been in the Linux world for that long, and I certainly came here after most of the fiery arguments on the topic had abated. I'm not having a go at you, but I am addressing the usefulness of discussing the topic. Though I suppose, one could argue, that's the water this new discussion came to test.

Thank you for your usual and welcome solid character and tolerance. I have since seen jeremy's post and need to think on this overnight. It may be that the likelihood of anything new is so low that this thread should be closed but OTOH if it is a dead subject it should just die out and become necro. To be perfectly forthcoming a decade ago I was among the somewhat rabid detractors of systemd. It isn't that I've grown to like it now, it's just that trying it out it was nearly invisible on my Desktop PCs. It was/is no big deal as far as I experienced. I keep wondering why there has not been either a mass exodus or some potent story from some admin somewhere about how systemd saved his bacon, or at the very least from some enterprise stating how much time and money they have saved since it's deployment. Oddly to me, none of these has occurred. Maybe the "other shoe" has yet to drop.

>Intelligent Assessment

I'd rather be a bumbling troglodyte than adopt SystemD. Not even going to rationalize it too much.
It's just a good intention that leads directly to hell to me. Not worth it. I have no trouble with the setup that is currently in effect.
Onga Bonga Zugg Zogg Grunttttt. Me dumb, but at least me no have more Poetterware. Geist still mad about loud noise from magic box due to evil word in pulse instruction file.
Geist no liked that. UGHH ZOGG BONGA AUGHH SHRIEK OOOOK OOK OOGAHHH
Poettering BAD much hurt Geist. OGGGGHHH.

Originally Posted by TheRealGrogan You can build whatever you want, and not build whatever you don't want after the base system. Sure, if you want to use Plasma, you'll have to build 87 metric shit tonnes of dependencies. In other cases, many dependencies will be optional. If they aren't there, they won't be dependencies.

It's not as simple as that, not for me anyway. You can ignore "optional" but if you don't use the "recommended" dependencies, something breaks further down the line, unless you have a lot more experience in tweaking config parameters than I have. What I used to like about LFS was that it gave clear instructions on HOW to build (which I feel I need; I'm not a hacker by temperament) but still left me a completely free choice of WHAT to build. I feel I don't have that choice any more, or at any rate wresting it back is too much of an effort. And, as this is a systemd thread, let me say that I attribute the problem at least partly to the degree of software integration that systemd has encouraged.

Originally Posted by ehartman As far as I know (but correct me, more knowledgable people, if I'm wrong) csh (and especially tcsh) is still the default shell in the *BSD's (csh originated in the BSD flavor of Unix, just like ex/vi, by the way). I'm actually using both, whichever is more appropiate to the problem I'm writing a script for. For instance, my "backup" script uses tcsh, but to restore something FROM such a backup bash is used (as it's using "$@", which is a bash'ism). And I think there are still quite some people using ksh, which is NOT a Bourne-shell replacement (although bash did borrow some enhancements from it).

FreeBSD, uses tcsh(1) as root shell. (/bin/csh is a symlink to /bin/tcsh).

sh(1) (ash derived) is also available in the base system.

OpenBSD uses ksh(1) (which is actually pdksh) and sh(1) is a duplicate of ksh(1) which has a different man page - and from the sh(1) man page it is explained why:

This version of sh is actually ksh in disguise. As such, it also supports the features described in ksh(1). This manual page describes only the parts relevant to a POSIX compliant sh. If portability is a concern, use only those features described in this page.

csh(1) is also available in the base system.

Other shells such as zsh, bash or dash are available in ports.

Originally Posted by hazel It's not as simple as that, not for me anyway. You can ignore "optional" but if you don't use the "recommended" dependencies, something breaks further down the line, unless you have a lot more experience in tweaking config parameters than I have. What I used to like about LFS was that it gave clear instructions on HOW to build (which I feel I need; I'm not a hacker by temperament) but still left me a completely free choice of WHAT to build. I feel I don't have that choice any more, or at any rate wresting it back is too much of an effort. And, as this is a systemd thread, let me say that I attribute the problem at least partly to the degree of software integration that systemd has encouraged.

I did have prior knowledge of building everything from source, before I undertook a LFS build. I was already used to doing a lot of things my own way, and I was highly familiar with Slackware's methods when in doubt about something.

I haven't done a LFS in probably 5'ish years, but I do agree that things have gotten more complex. For my own workstations, my needs are still fairly simple though.

this is what i think of systemd