@bison

Quote:

An unprovoked "linux sucks" post. That's not very original.

I'm a AmigaOS fan not Linux fan, that's no surprise is it.

Quote:

Linux itself I have no complaints with, other than one: it's too big. But I think it's human nature to think that everything has to get bigger to improve. It was probably inevitable.

It's not that its big that is the problem, just too many ideas, and one to direct where its going, its like orchestra playing without a conductor, every one sort of playing out of tune.

I was a big RedHat fan, but when it went corporate, they sort of killed it for me.

Last edited by NutsAboutAmiga on 09-Jan-2016 at 11:12 AM.
Last edited by NutsAboutAmiga on 08-Jan-2016 at 06:16 PM.
Last edited by NutsAboutAmiga on 08-Jan-2016 at 06:10 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@NutsAboutAmiga

Quote:

I'm a AmigaOS fan not Linux fan, that's no surprise is it.

No, not with a name like NutsAboutAmiga.

I'm a not-partisan computer user. I like Linux, *BSD, AmigaOS and AROS. I'm kind of neutral on OS X, and I dislike Windows. My dislike of Windows is based on its many shortcomings, and Microsoft's questionable business practices.

Quote:

It's not that its big that is the problem, just too many ideas, and one to direct where its going, its like orchestra playing without a conductor, every one sort of playing out of tune.

That's something Linux inherited from Unix. Everyone has their own ideas about the best way to do things, and the only thing that will stop that is working for a company and having them tell you what to do. But I think that's an even worse problem.

Linux may suck, but it doesn't suck as bad as Windows. It's hard for software to get big and not suck, and it's hard for software designed by committee to not suck. Window is big *and* designed by committee, so it really sucks.

AmigaOS is small and designed by a few knowledgeable people, so it had a better start, and back when it was still relevant it didn't suck.

_________________
"Unix is supposed to fix that." -- Jay Miner

@bison

Quote:

There was a vibrant Linux community in the mid to late 90s that started to fade in the early 2000s. I'm not sure why, other than I think maybe too many people started using it and the community was diluted with so-called "normal" users.

I will give you two hints about this:

1. Windows 2000/XP

2. Freetards and Licensing talibans

About 1. Linux soon became a hope for everybody of us that were simply pissed off by Windows 9X/ME bugs, shortcomings and high prices. We saw Linux getting better and better, addressing all stability and reliability issues, while Windows just crashed. Linux was more robust. More secure. More stable, although more difficult to understand and to use. But once you got Tux friend, nothing could stop you.

Then MS Windows 2000 came out, NT based kernels replaced once forever the mess that domestic Windows used before, and Windows turned again into a more stable, robust, usable operating system, making all the efforts to learn using Linux absolutely pointless for the most.

About 2. While "normal users" turned using XP, and Linux started gaining attention from software houses and common people, flame wars arose between people dividing themselves into either being pro or against the use of QT (for instance), and you could not even enter a newsgroup asking informations about a program, that immediately after you would have been badly addressed by someone because "it's not free software, you suc*er!". Sorry, but - for instance - I had lot better things to do than spend my free time reading dozens of flames, always reiterated, about this or that License.

Licensing can't be more important than the need and the features of a program. Not in a productive, real world. I could simply not give to a printing service images edited with GIMP, even if it was free and wonderful for some people, because of its lack of real CYMK separation. And this is just an example of what I could use for my work at the time.

Then time lasted, and I can't just see why I should use Linux for anything else than 1) compiling AROS and Icaros when needed; 2) run a secondary computer I don't need a Windows license for.

@paolone

Quote:

Linux soon became a hope for everybody of us that were simply pissed off by Windows 9X/ME bugs, shortcomings and high prices. We saw Linux getting better and better, addressing all stability and reliability issues, while Windows just crashed. Linux was more robust. More secure. More stable, although more difficult to understand and to use.

This describes my experience almost exactly. I was using Windows 95 OSR 2.1 and playing a lot of Quake 2 over a dial-up connection. I would lock up once or twice a night and have to reboot, and after a few weeks the file system was so corrupted that the system wouldn't work reliably and I would have to reinstall Windows. I did this so often that I had the keystrokes memorized and could reinstall without reading the prompts.

I installed Caldera OpenLInux 1.1 in January 1998 for the express purpose of porting the Doom source release from Linux back to MS/DOS. I never did finish the port -- DOSDoom was already out by the time I got started, although I didn't know that at the time. But I got far enough along the Linux learning curve to realize that it's better than Windows and decided to stay with it.

Quote:

Then MS Windows 2000 came out, NT based kernels replaced once forever the mess that domestic Windows used before, and Windows turned again into a more stable, robust, usable operating system, making all the efforts to learn using Linux absolutely pointless for the most.

This is almost exactly *opposite* my experience.

I spend 8-10 hours a day C programming on Linux, so Windows is very nearly useless to me. I can use it to check my email and surf the Internet, but that's about it. @NutsAboutAmiga is right that there is a certain amount of disharmony between different components of a Linux system, but I have not found this to be a significant problem.

_________________
"Unix is supposed to fix that." -- Jay Miner

@bison: that's because you stick with a single distro, and live with it.

If you have to develop or test a product for several ("mainstream") distros, with their inconsistent changes in filesystem and/or settings, you'll change your mind.

@paolone: I quote every single word!

@paolone

Quote:

Licensing can't be more important than the need and the features of a program. Not in a productive, real world.

Oh the painful irony in that one.

And here we are, discussing the impact of decades old code being circulated within a market that has invested most of its time and energy squabbling over IP that is pretty much irrelevant.

Productive and real indeed.

@cdimauro

Quote:

that's because you stick with a single distro, and live with it.

Well, not a single distro, but mostly those in the Debina/Ubuntu/Linux Mint "family". I haven't used Fedora or Red Hat for years, but I have used Tiny Core, which is very different. Anything that uses X11 and bash seems familiar to me.

The only thing that I notice as being much different is linking, since different distros use different lib*.so names. Even Debian and Ubuntu use different names; the XCB libraries, for instance.

_________________
"Unix is supposed to fix that." -- Jay Miner

@bison

Quote:

bison wrote: @kolla Quote: No, I don't post there, but reading the comments I pretty much agree with the general consensus, also I am a "Linux community" person. I *used* to be a Linux community person, back in the late nineties, but that community seems to have dissipated, and I'm not sure why. LWN is about the only part that remains. Ubuntu, Linux Mint, etc. have their own sub-communities. Maybe it got too large for a single community.

If you look at it like that, yes. I tend to stick close to what goes on with the the kernel and embedded development, the distro war I have little or no interest in. Although, death to systemd! :)

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@bison

Quote:

Linux itself I have no complaints with, other than one: it's too big. But I think it's human nature to think that everything has to get bigger to improve. It was probably inevitable.

Exactly what is too big? Are you complaining about the amount of supported hardware? Are there too many features for you? Do you find it too complicated to filter out what you do not need when building a kernel?

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

@kolla

Quote:

Exactly what is too big?

You actually answered this yourself in your previous post -- systemd is my biggest complaint. It probably works well when it works, but if it stops working, I have no idea how to troubleshoot, and learning it well enough to effectively troubleshoot is probably over a hundred hours. I would be reduced to googling Stack Overflow in the hope of finding a solution, but that's hardly troubleshooting. And there's no upside -- it solves problems that I don't have.

I'm currently using Mint 17.3, which still uses Upstart, but 18 is switching to systemd. I also use LMDE2, which is based on Debian stable, and it is still using SysV init, but I don't know for how long. After that, I'm not sure what I will do, but I'm considering Gentoo or DragonFly BSD.

Still hungry

_________________
"Unix is supposed to fix that." -- Jay Miner

@bison

Well, nothing forces you to use systemd, there are plenty of alternatives.

About your considerations, funny enough, in my growing list of projects, I am considering a Gentoo/DFB (DragonFly BSD) effort, based around the existing Gentoo/FreeBSD :)

I have currently Gentoo/m68k (Linux) that I need updating, I am keeping an eye on FS-UAE on DFB, I have briefly started an attempt at getting AROS hosted on DFB, but it looks like I need to look more at FreeBSD too in this process. Also I want to do AROS hosted on ARM-BE8 (Linux, for example using nVidia Tegra), and see if 68k JIT works on that too.

What sucks in all this, is that I have a job I need to pay attention to, and ... sleep! I wish I had more time for all the toys (A600, A1200, CD32/SX32, A3000/CSPPC, CDTV, MIST, Minimig, Pegasos... it just goes on and on) and other fun projects I have :)

Last edited by kolla on 11-Jan-2016 at 08:42 AM.
Last edited by kolla on 10-Jan-2016 at 11:59 PM.
Last edited by kolla on 10-Jan-2016 at 11:58 PM.

_________________
B5D6A1D019D5D45BCC56F4782AC220D8B3E2A6CC

Quote:

ne_one wrote: @paolone Quote: Licensing can't be more important than the need and the features of a program. Not in a productive, real world. Oh the painful irony in that one. And here we are, discussing the impact of decades old code being circulated within a market that has invested most of its time and energy squabbling over IP that is pretty much irrelevant.

And that's a sad truth.

@bison

Quote:

systemd is my biggest complaint.

Well, that's not really a "linux" issue. There's lots of bad software around for basically every OS out there (AmigaOS included), that are better avoided.

However, the problem is, systemd is rather low level component, which means many popular higher level packages have ended up depending on it. Because of that, many linux distros have decided to switch to it.

So in short, it's not a linux (which is a kernel) problem, but a linux distro (which are operating system distributions, that use linux kernel) problem.

So, why is it so bad then? In my personal opinion:

-Firstly, what is it? Well, it started as an init system, but it keeps reaching out in every direction, absorbing new functionality from everywhere. Having many things handled by one component reduces choice (it's harder to switch some things to other programs), and is a security risk (there's lots of code). Going by the words of one Amiga developer, "Systemd is a monster !!!"

-For many users, it doesn't solve any problems, that weren't already solved by other programs. As an option this wouldn't matter, but people are more or less forced to switch (either to systemd, or another distro)

-It's announced as being strictly Linux-only project. I'd definitely have expected more open source folks being worried about this. Sure, it's open source and anyone can download the sources, port them to any other OS and release both binary & sources, But much of that would probably have to happen without any help from the core systemd developers. And after systemd is updated significantly, you will have to adapt your changes to the new codebase again.

@kolla

Quote:

Well, nothing forces you to use systemd, there are plenty of alternatives.

Yes, but... if systemd is the default init system, the alternatives aren't going to get much love and will probably be poorly supported. I don't mind swapping out high-level stuff such as the desktop environment, but I don't want to do that on system-level components -- I'd rather find a system that doesn't use systemd as the default so that whatever it does use is well supported.

Quote:

I am considering a Gentoo/DFB (DragonFly BSD) effort, based around the existing Gentoo/FreeBSD

That looks interesting -- I hadn't even heard of the project. I'd look into it, but I'm already up to my eyeballs in another project. As you say, things like sleep are important too.

I have this idea that a Workbench-like DE on Dragonfly would be a very fine thing.

Last edited by bison on 11-Jan-2016 at 02:57 PM.

_________________
"Unix is supposed to fix that." -- Jay Miner

@Jupp3

Yes, I agree with all that.

The image of Lennart wearing an Amiga shirt is unsettling.

I wonder if systemd isn't bigger than the whole of AmigaOS by itself?

Last edited by bison on 11-Jan-2016 at 02:53 PM.

_________________
"Unix is supposed to fix that." -- Jay Miner

@
I don't see systemd as the problem, rather the breadth of what it is trying to support shows in systemd as a symptom of the problem, lack of standards for hardware of the same type that do the same job but in various ways, many of which are only different simply to be different and require different relationships to function within the same environment.

Then there are the users. Bunch of lazy-ass people whose only contribution to computing is to whine. They will cry about anything and everything while demanding the latest and greatest bleeding edge of all things computer from file systems to graphics cards. They can't do anything more within a 64bit system than what they could do with a 32bit system but still demand that the 'buzz' be catered to, not realizing that the complexity must be managed. Got more memory on your card, manage it. Want your hard drives to be journaled, manage them. Want multitasking, again,,, manage it. Systemd and its related functions is not the bad guy, rather a symptom of the complexity necessary to satisfy the ever increasing demands of a pack of dumb asses.

Now, don't worry,,be happy.

_________________
Tinkering with computers.

@bison

Quote:

bison wrote: @NutsAboutAmiga Quote: I'm a AmigaOS fan not Linux fan, that's no surprise is it. No, not with a name like NutsAboutAmiga. I'm a not-partisan computer user. I like Linux, *BSD, AmigaOS and AROS. I'm kind of neutral on OS X, and I dislike Windows. My dislike of Windows is based on its many shortcomings, and Microsoft's questionable business practices. Quote: It's not that its big that is the problem, just too many ideas, and one to direct where its going, its like orchestra playing without a conductor, every one sort of playing out of tune. That's something Linux inherited from Unix. Everyone has their own ideas about the best way to do things, and the only thing that will stop that is working for a company and having them tell you what to do. But I think that's an even worse problem. Linux may suck, but it doesn't suck as bad as Windows. It's hard for software to get big and not suck, and it's hard for software designed by committee to not suck. Window is big *and* designed by committee, so it really sucks. AmigaOS is small and designed by a few knowledgeable people, so it had a better start, and back when it was still relevant it didn't suck.

As for Windows' shortcomings, Windows 10's DirectX12 has fixed it's graphics API overheads and multithreading issues i.e. removal of serialisation of multiple deferred context threads into single thread immediate context bottleneck.

Linux's Vulkan operational state is currently behind DirectX12. Linux;s OpenGL 4.x has similar limitations as DirectX11.

Last edited by Hammer on 11-Jan-2016 at 09:22 PM.

_________________
Ryzen 9 7900X, DDR5-6000 64 GB RAM, GeForce RTX 4080 16 GB
Amiga 1200 (Rev 1D1, KS 3.2, PiStorm32lite/RPi 4B 4GB/Emu68)
Amiga 500 (Rev 6A, KS 3.2, PiStorm/RPi 3a/Emu68)

@Hammer

My number one complaint with Windows 10 is that Microsoft is using it to collect user data. For that reason alone I will not use it.

Last edited by bison on 11-Jan-2016 at 10:55 PM.

_________________
"Unix is supposed to fix that." -- Jay Miner

@kolla

Quote:

What sucks in all this, is that I have a job I need to pay attention to, and ... sleep! I wish I had more time for all the toys...

Perhaps cloning should be the first priority? ;)

On a more serious note, how do you feel about the potential of DBSD moving forward as the basis of a NGX OS? It sounds like many of the building blocks are starting to come together.

@bison

Quote:

I have this idea that a Workbench-like DE on Dragonfly would be a very fine thing.

Very.

The delta to piece together a NG OS should be an order of magnitude less insane than back-filling a 30 -year-old foundation.