*BSDThis forum is for the discussion of all BSD variants.
FreeBSD, OpenBSD, NetBSD, etc.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Distribution: LFS 9.0 Custom, Merged Usr, Linux 4.19.x
Posts: 616
Rep:
Quote:
Originally Posted by cynwulf
In the future you will probably find Linux in the same situation. People will wonder why it wasn't a microkernel or a hybrid and why he didn't just trash it in the early 00's and build from scratch instead of building onto the same base perpetually. MS Windows is also full of cruft and crud - and from what I've seen of software development, most big and unwieldy projects are the same, but sadly it takes huge amounts of time and money to replace them with something "better" - people just talking about it and calling "crap", "crap" doesn't get code written.
Why it isn't a microkernel is pretty obvious to me: They do not work efficiently on production hardware. Specifically, their IPC is terrible with virtual memory. I've argued several times that microkernels are good theory but not so good application on available hardware.
Yes, I cherry picked and you "summary rubbish"ed. Touche.
Lastly, my point is that someone is finally developing something better. But, what do we get? We get "x is the Unix way" and people arguing to keep bad software around. Calling it crap, when it clearly is crap, will hopefully keep things moving forward. Or, we can spend another 20 years running garbage software because people assume that because it is old, it must be "the UNIX way" and is therefore worth hanging on to. Again old != Unix Philosophy, currently works != Unix philosophy.
You know, the first time I ran Weston it crashed. You know what else... It told me right there on the screen what failed, there was a module named incorrectly in the config file. When it did finally come up, it felt like a desktop. No flaky input lag, no tearing, no filling empty blocks when resizing windows. (Note: I never understood the reason for the input lag until I watched that presentation, the ridiculous amount of IPC involved in just drawing an X window finally explains it.)
Why would I ever want to go back to generating config files and xserver lag that can last seconds? To want X to stay around is to be a masochist.
Why it isn't a microkernel is pretty obvious to me: They do not work efficiently on production hardware. Specifically, their IPC is terrible with virtual memory. I've argued several times that microkernels are good theory but not so good application on available hardware.
Which is why XNU and NT are hybrid kernels... but my point is that I also prefer monolithic kernels for the reasons stated and more, but "the future" may view things differently. History is the judge as they say.
Quote:
Originally Posted by Luridis
Why would I ever want to go back to generating config files and xserver lag that can last seconds? To want X to stay around is to be a masochist.
I agree, but I've never made the point that I want X to stay around...
I just don't agree with your stance that we've been running "garbage" for "20 years". I also don't think the "The Unix Haters Handbook" makes better or more valid points than Eric S Raymond makes in his blog. In fact it's mainly a collection of quotes, more like throwaway comments, from some very biased people.
Distribution: LFS 9.0 Custom, Merged Usr, Linux 4.19.x
Posts: 616
Rep:
Quote:
Originally Posted by cynwulf
I just don't agree with your stance that we've been running "garbage" for "20 years". I also don't think the "The Unix Haters Handbook" makes better or more valid points than Eric S Raymond makes in his blog. In fact it's mainly a collection of quotes, more like throwaway comments, from some very biased people.
Again, I don't think X has just become "dated". So, I don't see how people see monolithic kernels in the future is relevant. I think, as many others do, that X has has always been terrible design. This is not a new idea for me. When I first used it in... 96ish it gave me a WTF moment because it did not seem possible that the console worked so well and X was... well, a nightmare. When I investigated further, I understood it to be some carry over (clone) from CDE and commercial Unixes, one that was not well liked there either.
I agree, but I've never made the point that I want X to stay around...
Could X11 have been better maintained, actually by the maintainers? here is a beautiful quote:
Quote:
There was a video of Wayland makers ((ex-)X11 coders) declaring X11 dead because of it's ugly add-ons over the decades. But who has made them if not the X11 coders?
Now that's taken as reason for Wayland...
X11 may be dead but it was killed by its own maintainers!
What I find utterly hilarious is that you think what I am saying was motivated by some corporate agenda. Perhaps you should take your tinfoil hat and head back to Roswell.
I wasn't talking about you personally, but whatever. Think what you want.
The corporate trend of replacing broken things with new and shiny is what worries me, not some individual.
Just toss it and buy a new one, until the point where it's soldered together in a factory so you can't repair it. EU has written laws against it.
Wayland will eventually get replaced when something new comes along, and then you will see it could've been much easier and cheaper to just strip down X to a minimum and fix what's broken.
But there's no profit to be made on "crap" as you put it, so why not build another "crap" every year? Like apple does for example.
Distribution: LFS 9.0 Custom, Merged Usr, Linux 4.19.x
Posts: 616
Rep:
Quote:
Originally Posted by elcore
I wasn't talking about you personally, but whatever. Think what you want.
The corporate trend of replacing broken things with new and shiny is what worries me, not some individual.
Just toss it and buy a new one, until the point where it's soldered together in a factory so you can't repair it. EU has written laws against it.
Wayland will eventually get replaced when something new comes along, and then you will see it could've been much easier and cheaper to just strip down X to a minimum and fix what's broken.
But there's no profit to be made on "crap" as you put it, so why not build another "crap" every year? Like apple does for example.
If X can be stripped down and fixed so easily I suggest you setup a presentation at one of the OS/FS events. At the very least you should setup a blog about it. Otherwise, you're just another one of the internet "peanut gallery" talking b.s. and armchair quarterbacking on an issue he doesn't know anything about. If you have good arguments for fixing and continuing X, people will hear them, including me. If you can't do that then I will rightfully consider your opinion to be based on nostalgia or some other nonsense that doesn't apply to the actual problem at hand.
If X can be stripped down and fixed so easily I suggest you setup a presentation at one of the OS/FS events. At the very least you should setup a blog about it. Otherwise, you're just another one of the internet "peanut gallery" talking b.s. and armchair quarterbacking on an issue he doesn't know anything about. If you have good arguments for fixing and continuing X, people will hear them, including me. If you can't do that then I will rightfully consider your opinion to be based on nostalgia or some other nonsense that doesn't apply to the actual problem at hand.
X works, sometimes the end result depends on GPU driver and hardware, but it works fine for my use case; monitoring servers.
I've got nothing against alternative solutions, there always must be something to fill the blank when things break, I get that.
What you don't get, is that both packagers and sysadmins like well tested, feature complete things with a stable and modular interface.
Evidently, they're not going to toss a working solution until the alternative is stable enough.
Personally, I think part of the problem is the failure of many people to understand what X was designed to be and what it is used for. Taking my use-case for example:
- I use a tiling window manager. No resource-heavy DE that requires hardware acceleration and 2GB of RAM.
- I do not use my computer to play games. (Which is not an option anyway when using OpenBSD.)
- I use the lightest viable browsers. Not huge monstrosities like Firefox.
The result is X has always worked perfectly well for me with both BSD and Linux. The X window system was designed a long time ago for simple graphical needs. The problem is that it has been expanded and modified to accomodate what the new generation of Linux users want; a free Windows replacement. (And what the businesses that have been persuaded to use it in their workplaces want.) X is being used for a job it was not designed for, so a replacement is needed, but that will take time and money. Until then, one workaround is to have a simple and lean system that does not task the graphical capbilities of X, which, in my opinion, is something BSD and Linux users should want anyway.
Distribution: LFS 9.0 Custom, Merged Usr, Linux 4.19.x
Posts: 616
Rep:
Quote:
Originally Posted by elcore
X works, sometimes the end result depends on GPU driver and hardware, but it works fine for my use case; monitoring servers.
I've got nothing against alternative solutions, there always must be something to fill the blank when things break, I get that.
What you don't get, is that both packagers and sysadmins like well tested, feature complete things with a stable and modular interface.
Evidently, they're not going to toss a working solution until the alternative is stable enough.
They (Sysadmins) also don't like shoddy, problematic code bases that users and support personnel constantly complain about. I know this because I am a Systems Management Admin and I've been through it enough with Java. So, I get things just fine and you're just making assumptions again.
Assumptions, opinions, complaints and thinly veiled resistance to change... Since that's all I continue to see here, I'll unsubscribe. That way y'all can go on and on about how good X is, how easy it is to fix, how it's "network transparent" (not for some years now), how much the world still needs motif, and how this really bad IPC implementation (X11) is an exemplar of the Unix Philosophy and is the One True Way®. Meanwhile, the rest of the world will move on while you are all stilling here scoffing at True Type Fonts, composite effects and contextual menus like a bunch of old curmudgeons.
When I got my first Unix, it had 16k of RAM. We didn't even have no electricity and I had to run in a squirrel cage attached to a generator! Uphill, both ways!
So, go ahead commence with the reciprocal back patting. Oh, and PM your address. I think I might have some 8" floppy disks out in the garage that you might find useful. Hell, I might have a couple of mouldy reels or even a drum!
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.