Quote:
Quote:
Quote:
|
Quote:
Quote:
Of course you know that already, since according to your own words you have tested systemd for an extended period of time and it also has been pointed out more than once in many of the systemd threads you participated in. That makes me wonder why you bring that up here. Anyways, since the (alleged lack of) UNIX philosophy in systemd seems to be such a major point for many people, I wonder what those peoples opinion are on, for example, un-UNIXy things like btrfs or ZFS, which are far from doing only one thing and doing it well. Not long ago I came across a post on Phoronix which brings up the problem with those using the UNIX philosophy as arguments for anything: it is not well defined what "doing one thing" actually means. I will quote that post here: http://www.phoronix.com/forums/showt...809#post447809 Quote:
|
Hypothesis: the frequency of the word 'you' in a post is inversely correlated with constructive discussion. Evidence will be supplied below.
|
Quote:
|
I will continue to use Slackware even if systemd or other pieces of software are adopted, because I think that Slackware is more about building a sound and well thought system with the available components than about using an init system. I trust that the Slackware implementation of systemd will be the one that causes the least damage.
Over the years I have seen many things change in the Linux world; I remember when Linux experts laughed at Windows because it had graphic drivers in the kernel: how superior was X11 with its client/server design. Now Linux has graphics drivers in the kernel and users complain if they don't get the same performance as in Windows. X11 will probably be replaced, too: what happened to that clean design? Now init is becoming like svchost under Windows... seems like Windows made all the right choices years ahead of Linux :-) Just kidding: I'll stick with Slackware. |
Quote:
|
Quote:
Debian doesn't even require you to use systemd. It's only using it as a default (partly to allow some packages that do rely on it, though that's technically still a "bug"). I find it funny how people like ReaperX7 are even trying to convince people to use Gentoo instead, while it's offering the exact same approach. Also see: http://www.vitavonni.de/blog/201410/...g-systemd.html |
@Tobi: That quote is not so great, I see the point you're trying to make, but quoting silly troll posts from phoronix attacking UNIX philosophy at it's roots just to prove a point - dare I say, to try and justify your own move to systemd - is much the same as the kind of posts Reaper is alleged to have made.
UNIX philosophy still stands - it's not perfect, but it's a set of rough guidelines to follow. Thompson and Richie defined it and it's safe to say that they probably knew what they were talking about. One of the fundamentals which your poster trashes is that "power of a system comes more from the relationships among programs than from the programs themselves. Many UNIX programs do quite trivial things in isolation, but, combined with other programs, become general and useful tools" (from the UNIX Programming Environment".). This is referred to again in another paper by Kernighan and Pike: "a style of program design that makes programs easy to use and, more important, easy to combine with other programs". Again - to make my position clear - I don't think systemd project have to stick to the above, they can do what they like, but they should also expect criticism and for people that don't like it to avoid it. The goals of systemd, as clearly set out by the project now via mailing lists, blogs, articles, speeches and whatever other medium, are at odds with this and it's clear that. To look for excuses to prove that systemd is not being pushed and forced in wherever possible is to be exceptionally naive and ignore the principal developer's own posts and overbearing arrogance. Instead of posting hysterical rhetoric, rubbishing it and pulling it apart and looking for examples where it's not observed, your quoted poster should try to explain in detail why this approach is wrong or bad practice and what harm it has actually done for the last 30 odd years (in bringing about the *nix operating systems we use today). systemd needs to prove itself, not the UNIX philosophy. A good example of the above is something like grep or cat, which perform a role and can be adapted to other roles used in scripts be combined with other commands to do something useful. Those are just two examples. It's interesting to note in fact that Patrick Volkerding designed Slackware package management using this very approach (he is free to correct me if I'm wrong). Rather than building new binaries from scratch and reimplementing existing software all over again, he wrote scripts which simply made use of existing tools and can be modified and adapted by the user. UNIX philosophy in practice. |
I am only commenting because this thing has hit 5 pages now. Come on really there isn't going to be any exodus, over systemD if it even happens. Many of us have been using Slackware for 15 years. We are not going anywhere just because of a change like that, to do so would be giving up everything that is great about Slackware, which does not all dissapear just because systemD shows up even if I/we don't like it.
Don't throw the baby out with the bath water springs to mind. BSD is nice, but hardware and software limitations are just to much for most of us; at least to use exclusively. Slackware is already far enough out of the main stream that getting certain things done/working frequently requires real effort where it would be a simple package install on some other OS; but you can always get there. Slackware has so many wonderful things going for it (as we all know) that its worth the occasional headache (saves on so much other day to day frustrations). Patrick is the BFDL and we should trust him; his record might not be perfect but its far better than most. If there are alternatives to systemD likely to be both available and suitable on a useful time horizon I am sure he find them. If not I know he is not going to the one thing that really would cause folks to leave SL and that would shipping a obsolete or broken distro. If the choice is drop things people use like KDE or systemD, he will pick systemD. I can't really speak for him of course but I am fairly certain of that. Sometimes you just can't hold back the tide, systemD may prove inevitable or it may not. If you don't wan't systemD look for alternatives, test stuff out post your experiences, but don't go spreading a bunch of FUD about how people will abandon Slackware. They won't if for no other reason than for folks who really know and like Slackware there are few places to reasonably go and no places where systemD would be unlikely to follow. |
@cynwulf: I can assure you that I don't need this quote to justify my switch to systemd, that move was entirely caused by my research of that software and evaluating if and how it fits my needs, exactly how I have done with other tools.
The point remains that the UNIX philosophy is, while a valid philosophy, is only a raw recommendation and by far not the only valid one. For example, while Slackware's package management tools are indeed a representation of those principles, there would have been no problem at all if Mr. Volkerding would have written it in Python or C. Also, I wanted to point out that often times UNIX philosophy proponents are cherry picking the projects they want to test against said philosophy, which is why I also mentioned projects like ZFS and btrfs and asked about oppinions about that. |
Quote:
|
It's not a case of needing, it's what you're using to justify your position. If you like systemd, then by all means go ahead. As far as I'm concerned, I do question someone who previously said he was avoiding it and had gotten rid of his Debian systems to do so and then moved to gentoo and has now switched to it and is defending it. I don't have any problem with you changing distros, as I'm not a distro fanboi - that's entirely up to you, but quick shifts of opinion from avoiding something to being an active proponent, makes me doubt that person's opinions - however forcefully and thoroughly expressed. In a few months you could be "playing for a different team".
UNIX philosophy is no less valid because some people misquote/misapply it. That's much like the old "you use distro X and therefore are an idiot". There are people coming out with "sound bytes" and there are some admittedly stupid sites and stupid ideas out there ("forking" Debian for one), but the systemd PR machine is certainly no more mature. As you "gentoo folks" really should know by now. |
Quote:
|
Quote:
|
Quote:
This is the UNIX philosophy in action: Code:
cd /srv/local/music && ogg123 -d alsa -o dev:default -q -@ <( ls -d */* | shuf) The thing about traditional init systems and init-scripts is that they use the exact same shell to start the system that the users interact with. The result is that the same flexibility that the end user has is also available to the system administrator who can use it to do all sorts of things that the developers never even imagined he'd do, and he only needs to know the exact same shell language that an end user will use to achieve it all. systemd will probably work well for the modern pointy-clicky desktop world, whose consumers are blissfully unaware of what they're missing and quite happy to only use what they're given, and I'm not saying that service supervision and some of the other features systemd has aren't useful, but IMO systemd throws the baby out with the bath water and you'll never convince me it's a better system-glue than the shell and the increasingly maligned (by systemd proponents) "UNIX Philosophy". |
All times are GMT -5. The time now is 10:57 PM. |