Post a Comment
There seem to be great parallels between this project and Haiku -- both of which seem to be making extremely impressive progress against the "official" closed-source versions, AmigaOS 4 and Zeta. All we need now is an open-source clone of RISC-OS
It's a great shame really that there needs to be such a duplication of effort. I'm sure that the coders of both AROS and Haiku would like to be able to contribute to the official continuations of their beloved OSes. Given the financial troubles that the developers of AmigaOS 4 and Zeta seem to be in, plus the fact that inbetween Windows and free Linux, there doesn't really seem to be much of a market for a paid-for proprietary OS, is there really much of a motivation for keeping them closed?
Edited 2006-09-13 16:29
"both of which seem to be making extremely impressive progress against the "official" closed-source versions, AmigaOS 4 and Zeta"
Okay... could it be that you have never seen an real OS4 installation?
"Given the financial troubles that the developers of AmigaOS 4"
Which financial troublies? Come on guy! Don't tell the people such lies!
OS4 and MOS has become in less than 5 years much more features and native software (look at AROS Archives: only 69 Files avaible - OS4 on os4depot.net has 1027!!) than AROS will ever have. With the JIT in OS4 and MOS u can run nearly any OS conform software written for AmigaOS - thats one of the important features AROS lacks.
Look at pages like www.amigaworld.net and www.morphzone.org - to see whats really going on in OS4 and MOS communitys!
And the best thing about AROS is: it is not binary compatible between other architectures. If you run AROS PPC and youre compiling any piece of software for it you can't run it on an x86 installation - WTF???
>>OS4 and MOS has become in less than 5 years much more
>>features and native software (look at AROS Archives:
>>only 69 Files avaible - OS4 on os4depot.net has
>>1027!!) than AROS will ever have.
You are dealing with absolutes, always bad. There where lots of people who doubted that Linux would ever have more apps then Irix or Solaris but look at where we are now. You shouldn't underestimate the motivation of the people working on AROS.
>>And the best thing about AROS is: it is not binary
>>compatible between other architectures. If you run AROS
>>PPC and youre compiling any piece of software for it
>>you can't run it on an x86 installation - WTF???
Dude do you know what you're asking for? What do you want? AROS VM? AROS universal binaries? Do you know how ugly those solutions are? No, I'm happy the way it works now.
"You are dealing with absolutes, always bad. There where lots of people who doubted that Linux would ever have more apps then Irix or Solaris but look at where we are now. You shouldn't underestimate the motivation of the people working on AROS."
Oh man, you can't compare Linux with AROS. Linux is just an Kernel and AROS is an whole OS! AROS is going in competition with AOS4 and MOS and not with Linux, Windows, Zeta and others.
Linux had the advantage that it comes with GNU OSS Software which makes it comparable to Sinix or other Unix derivates which were popular back in the mid 90's - but damn expensive!
"Dude do you know what you're asking for? What do you want? AROS VM? AROS universal binaries? Do you know how ugly those solutions are? No, I'm happy the way it works now."
Yeah i know. But this is something people expect from an OS which runs on more than one CPU Architecture!!
"Dude do you know what you're asking for? What do you want? AROS VM? AROS universal binaries? Do you know how ugly those solutions are? No, I'm happy the way it works now."
Yeah i know. But this is something people expect from an OS which runs on more than one CPU Architecture!!
Could Linux PPC execute executables of Linux i386? Is Linux an OS that runs on more than one CPU Architecture?
Could Windows64 execute executables of Windows i386? Is Windows an OS that runs on more than one CPU Architecture?
@cHaOs667
>>OS4 and MOS has become in less than 5 years much more features and native software (look at AROS Archives: only 69 Files avaible - OS4 on os4depot.net has 1027!!) than AROS will ever have.
And why does this please you? It seems that you perceive AROS to be some kind of threat, but I can't for the life of me figure out why. Even Hyperion are fine with AROS.
I can only think it's the closed-source mentality I mentioned. When you're listing OS4Depot's 1027 apps, it's worth remembering what it doesn't have. A potential developer can put AROS on an old machine and contribute to the pool of Amiga apps - that's OS 4, MorphOS AND AROS. I can't see how a small run of Troika boards will do the same.
>>And the best thing about AROS is: it is not binary compatible between other architectures. If you run AROS PPC and youre compiling any piece of software for it you can't run it on an x86 installation - WTF???
Are we still comparing it to the unavailable-PowerPC-systems-only AmigaOS4? Am I missing some clever irony?
AROS needs community support NOW. There's still a very closed-source "official version" mentality IMHO, which given everything that has happened, is very surprising.
Most Amiga developers don't consider AROS for their apps let alone use their very unique and niche skills to contribute to the OS itself.
Unless AROS gets support, I just don't see a future for anything Amiga-like.
Edited 2006-09-13 16:41
for a great OS. It's source-compatible with Amiga OS 3.x, making it easy to port any properly-writeen apps (stuff that banged directly on the m68k or the Amiga chipset won't, of course, compile as they rely on stuff simply not present in a PC) and it has a truly gorgeous GUI (it implements all the Amiga paradigms but in a nice, modern looking way) It's even starting to get apps. With a switch to self-hosting (cuurently the OS and all apps have to be compiled on Linux or *BSD) and the addition of a working USB stack, this will ROCKET to the forefront of Hobby OS development. Currently one of the three "next-gen" OSS OSes to watch (Haiku and Syllable being the others)
Actually, you can compile apps in AROS. I've developed Lunapaint under AROS Native on a dedicated PC. The compiler has some small lacks in the enviroment though ('make' and some other things) but it is perfectly OK to develop apps in aros.
You're right though. AROS can become a really big hit if more developers help to finish the standing tasks. I'm confident that in a year or so, AROS will look very interesting indeed. Key word: Simplicity! Not many OS's has this anymore. In the past, OS's revolutionized computing with combining simplicity with easy-of-use, why do they now one by one abandon that? It takes out the fun in my opinion, gets in your way! :-)
AROS keeps this simplicity, and that gives you more CPU/RAM for your apps :-)
How can you build a large project without make or some other similar program? Plus, if AROS dev environment is currently good enough to build and develop in, then why is AROS itself still being cross-compiled rather than self-compiled? If all that's needed is a port of gmake or something, I can't see things taking very long to bring AROS into the wonderful world of self-contained-ness. On the other hand, if the build in environment is otherwise flawed and buggy then why hasn't fixing that to remove *nix dependencies in development been a key focus?
Without make it's hard to jump in and port code that uses it for its build system. Making you own projects without it on the other hand is possible. It's not ideal without all the build tools though, but I've not run into problems.
As for self compiling AROS - I guess it could be done with allot of modifying of the AROS code, as it's right now dependent on the AROS build system on Linux. You can't compile AROS on Windows either, nor any other system. That's the choice the devs made in the beginning, and yes you could question it now. But feel free to help changing the situation ;-)
Regarding the dude that said Linux is just a kernel and AROS is a whole OS. I don't know how well he knows about AROS, but the point is that AROS isn't nessesarily a whole OS. Many Amiga people have problems with grasping this fact. AROS can have replacement parts just like any distro can in context with Linux. AROS is an open source project, so different elements of a would be OS solution can be exchanged for others, to deliver a wholly new OS solution. Furthermore, I expect that you very well will see several distros, and perhaps really soon. But they won't be exactly like Linux that's true. They will most probably remain faithful to the basic file structures of AROS and - the KERNEL. :-) Does it sound like Linux/BSD yet? That's right. Open Source projects gives the same *choice*. What it is or isn't is up to you.
As for some of the shortcommings, they could be remedied with some help, over time. But let's not forget that AROS doesn't try to be the end all be all OS. I think it's refreshing with an alternative operating system that isn't created to "fix" the flaws in "inferior predecessors". AROS isn't like that - it's not an experiment and exercise in being new and original - it's about the good old simplicity and easy of use that many OSes had/have, but aren't focusing on at the moment. Elegancy in simplicity anyone? Also, it's a 80's era OS philosophy that should issue an interesting challenge for OS developers in every camp. It's not unix! How many new Open Source OSes can you say that about!
having used the Amiga for many many years back in the day, it's great to see that this project keeps on trucking.
last I checked it out (which was quite a while ago), it still suffered from the lack of proper memory protection that plagued the Amiga OS back in the good ole' days (Guru Meditation anyone?).
memory protection adds quite a bit of overhead and thus will make it slower. but having user applications being able to drag the whole os down isn't pretty. so, does anyone know if this has been added or is in the works? the article (though very well written) didn't mention this, unless I somehow missed it (not entirely impossible).
quick answer is: "no..unfortunatly there's still no memory protection for all flavor of amigaOS systems (this including AROS, Morphos, AmigaOS4, Amithlon..).
This is still the big weak point of AmigaOS.
It is very sad that for sake of compatibility with now outdated apps developpers choosen to keep original API instead of updating it to support this must have feature.
In the day of commodore, ppl didn't hesitate to update the OS even at some compatibility price.. like, AOS1.3, and AOS3.x were very diffrent running AOS3.x requiered you to update the Amiga ROM chip and all.
nowdays devs (OS4,AROS,Morphos)just stick to the 3.x API as if it were pure truth, true gold .. this is too bad. AmigaOS clearly need something to prevent apps to crash down the whole OS.
i hate to say but we all know about the company that own 90% of computer's system market share that company which delivers the most crap OS ever to 90% of the PCs around the world.. that OS filled with all the bloat and fat you can come up with.. that OS who get plagged with security threats and loads of virus, you know what OS i'm talking about, made the same error..
for compatibility with the past they just stick with crap API instead of coming up with some better stable , clean, secure fundations.
well i don't expect anything from That company anyway, i use Linux of course.. but it's just sad to see AmigaOS devs are making the same mistakes as That company..
Macintosh on the other hand showed imho the right thing to do.. at some point you need to cut the cordon, and start with something clean.. "up to date i should say.
for compatibility there's emulators.. UAE is up to the task, for ex. but there's other options too.
'just my 5cents.
nowdays devs (OS4,AROS,Morphos)just stick to the 3.x API as if it were pure truth, true gold .. this is too bad.
The problem is that the existing applications written for the AmigaOS API expect there to be no memory protection. If you add that, you break compatibility with the old software, and an OS with no software is pretty useless.
Also, I believe AROS is attempting to get AmigaOS 3.X implemented fully before they start changing things. This is the easiest way to get everyone working toward the same goal. They could argue for years about the best way to implement memory protection (or whatever other feature you want added to AROS) but saying "We'll get AmigaOS 3.1 re-implemented first." Is the easiest way to get something done.
no..unfortunatly there's still no memory protection for all flavor of amigaOS systems (this including AROS, Morphos, AmigaOS4, Amithlon..).
I believe (but am not sure) that AmigaOS 4.0 supports limited memory protection. The application has to explicitly ask for it, though. If it was on by default, old code for AmigaOS 3.X would not run without extensive modification.
AROS lacks Memory Protection, Ressource Tracking,...
Why sticking with this old 20-years API ?
Application compatibility ? AROS isn't binary compatible with any AmigaOS. The only apps that can be recompiled (with some changes) are also old...
So: why sticking with the old API ?
The fact that it does not have these features and that they cannot be added will serverly limit the interest anyone can have in this OS. Not to mention the developper point of view where the lack of memory protection is a real problem...
I really don't get where they think they can go without these features...
And that's too bad 




