What you describe, we already have in AmigaOS, 3.5 and 3.9 introduced a whole lot of incompatibility, OS4 is even worse in that regard. Features of both Workbench and shell has come, vanished, reappeared etc. And all this within "official AmigaOS
What, where? Excuse me, but since when are there incompatibilities? I don't know much about workbench problems since I'm only using it and it works fine - but as a matter of fact, I maintain the shell. Yes, I do. Thus, speak up open, which incompatibilities are there? I haven't received a bug report for years, and there's a straight line of shells from 3.1 to 3.9 and the various boing bags. Hopefully all compatible to each outher. If not, report a bug. Yes, please do.
The problem is that quality is not allowed to win,
Quality is not allowed to win if people make stupid patches and somebody has to clean up behind them. It is not that I don't care about my software. Just that there wasn't much request for caring lately. I believe I released a couple of old and older stuff lately, and updated it. So please - I beg your pardon?
Again and again the "official developers" have messed things up and created lots of trouble. Trouble that could have been avoided if development could have happened in the open so that _all_ developers and users could have a saying in what what solution we want, in what direction we want to see the OS advance.
Thus, once again, if you have a case where you believe that I've messed something up and it requires fixing, say so openly and I'll fix it. As soon as I find some time. But please don't tell me that I leave things unmaintained. Thank you.
Noone stops anyone from forking the Linux kernel and it happens all the time, distributions typically has a whole lot of their own kernel patches that they maintain, the different architectures developers have sets of patches they maintain as they slowly get them merged into main.
So, apparently, you have not been working on it, right? Yes, you can patch up your own kernel. But there's no benefit. You need to move it into the official branch - and that's the important part where you have to discuss and argue, and will often find yourself in a position that you do not quite understand the sources well enough to produce something that is acceptable. Yet, in Amigaland, this type of crap is thrown untested at people, actually *without* giving the maintainer of the program even a chance to check or argue.
Those professionals you write about used to be a bunch of happy amateurs too at some point, they were allowed to learn and advance and Linux evolved into what it is today.
Funny thing. Guess I did the same, but I learned on AmigaOs. Yet, I never did what I see just here - patch up some outside program without actually knowing what it does, and release that to end users. Strange, isn't it?
As a matter of fact, you *don't need* the Os sources to write good programs. You need the interface. Actually, it's one of the core disciplines of software development to write against well-defined interfaces, or to define such interfaces. Yet, such key knowledge is apparently ignored in Amiga-land. Instead, the average Amiga hacker goes "bang-bang" on the hardware and "hack hack" onto the Os. So what do you think will happen?
Yes, when most developers ignore such requests, have bouncing email addresses etc. In real world it is quite common to create patches and distribute them, because sometimes getting proper solutions upstream takes time. Why should this be different on Amiga?
Because the audience you get is different. The skilled software engineer has long left the platform.
Yes, and why is that a problem? The freedom to tweak and mess around as much as you like is _exactly_ what should happen.
No, should not. The problem is that such programs are released into the wild, there causing problems with *my* stuff, and I have the trouble cleaning up behind the lines. The amount of trouble tools like MCP caused and the amount of wasted hours because somebody didn't know better but still hacked the Os are enormous. There's an interface, program against it. If you don't like it, or need something else, talk to the person responsible for it. There are still ways to negotiate cooperation. Just because sources are closed does not mean they are locked away. They are only locked away from the fools, for the better of the rest of the users.
And noone would be forced to use his code, anyone can chose what to use themselves.
That's not quite that simple. You do not install such a hack and then say "it works or it does not". It will likely cause disruption elsewhere, with programs that seem to work just fine before. And believe me, in the end its not the author of the patch that gets the bug report. Such nonsense creates a lot of work *for other people* and trouble for the users that are completely unable to resolve the problems. Linux is a paradise for programmers, but its probably hell for ordinary users. Not that I'm not using it and loving it, but writing a program on Linux that still works in two years from now is almost impossible. You have to continuously update the sources because some of the smart developers again considered to change an interface. That's ok for Linux. It's not ok if you want stable products. I'm not a big fan of Microsoft, but there's a reason why they are successful. Their stuff just works. Guess what? It's what most people need: An environment that solves their problems. Not that creates new ones because patch A doesn't work with program B.
With code out in the open, you don't even have to discuss with him, anyone can look at the code and there will be concensus about what makes sense and what doesn't. Also people have different needs, for some uses one solution makes a heck lot more sense than "the official" - this is why there in "Linux land" exists so much diversity.
For the better or for the worse. Or how does one say? "You never know how the print command is called today". Yes, works ok for me, I know by now enough about the details. Try to teach your granny, and you'll learn what's wrong with it.
That's exactly the type of problem you get without a good project management, or somebody who overlooks the big picture. How many inconsistent user interfaces do we have in Linux? I stopped counting... Again, fine with me - but the hell for the average user.
In Amiga, we even had a user-interface style guide. Completely outdated by now of course, but at least somebody took the time to write down how an Amiga application should look like, which AREXX commands it should support and so on.
What do you mean "no legal constraints"?
Pretty simple. I can handout the mmu.lib to whomever I want - that's my source. I cannot hand out the latest layers.library, simply because I don't own it. Yet, I can maintain it.
Because of legal nonsense and the hostile attitude among certain developers, the Amiga community was never even given the chance. Luckily at least some people realized this so AROS was born.
Again, I don't have a problem with AROS, but I *personally* don't really care much. Problem is, time of AmgiaOs is really over, for various reasons, the whole construction of the Os is just upside down, and there is no market for it in first place. Yet, if you want to contribute your time, you are surely welcome. I personally don't see the vision behind all that, i.e. what the purpose of the project is about. But anyhow, who am I to judge...