I think UAE inside of AROS is overkill.
The main use of UAE is running hardware-hitting applications and games.
I would like to see a simple high level amiga emulator, that just simulates 68000 opcodes, and for all I/O that uses the proper API calls, divert these to the equivalent AROS routines. This means lots of stuff won't work, but it is a much cleaner and simple approach.
Some stuff will work.

Could then just run individual programs via the shell (eg "run68k myprog")
ie More like the way Wine runs.

Yeah Wine would be nice too (yeah I know I'm dreaming)

well for that we have UAE...the issue is make UAE merged with UAE as much as good and possible...UAE works fine, merge it with AROS is more a extend feature!!!

Quote:No Lawyers allowed! :massacre:

I could be wrong but when people talk about running "old" amiga software, don't they often mean games?

well, AROS can eventually have new games. :zidane:

personally i don't care about games, but some useful programs - paint programs, mp3 players, etc. ARE important to develope. i guess it will be a combination of who makes the most noise and who is having the most fun programming.

paolone wrote:
Quote:

Instead, AROS fanaticism! (j/k)

Quote:

We are all users with their needs, their dreams and their own ideas about AROS. I sold my Amiga 1200 ten years ago or even more, then I've been a Windows, Linux and MacOS user, and nowadays I have great faith in AROS, as it may become what Linux has never been in these years: a NICE, USEABLE and USER-FRIENDLY operating system, which doesn't need years of computing knowldge to be used. As an advanced user, I simply don't need any legacy Amiga App on AROS: I think we don't need old applications which could run slowly (but faster than on a classic Amigs), first of all because they will be less stable thain in their original ambient, secondly because they're simply outdated. What we need on AROS are brand new applications which can compete with Windows and Linux software.

But when you propose AROS to a old Amigan, his first question will probably be "wow, may I run my old amiga software on AROS?". They're not so happy when you say "NO, but you can still run UAE". So, here's the need for a UAE integration. Maybe it could be a nice way to attract amiga users, but I'm afraid it will attract also lawyers and companies which don't like anything that could compete in their market.

Amigan fanatism is a old plague. But I don't think AROS developers should care about it.

hi,
i dont care lot of uae ...cause uae as it is, is very stable, but i didnt understand "in bring AROS emulation (sandbox) to AROS as well", but the thing it could be nice is a virtual machine that could run amiga files (that could have an special flag)....but that is not something i am concern due that lot of amiga apps are outdated and could be nice replacements for that...the uae is more a nostalgic feeling in games and demos, not in really apps (as imagine, sculpt and this on, maybe a nice suite on POV could do the job) and a psx emulator could do the games (as mame had been very nice in that area)...if more people stop concern less on nostalgic implementations, maybe we could see some nice things going on...taking care to not make AROS too POSIXlike....but we have lots of source for good thing that could work very nice on AROS...and the AMIGA programs are not that easy supported...imho!!!!(open sources)...the very big quest for AROS is to make a very strong multimedia support, since that was the stronger in amiga...but at actual state of aros...all is just a matter of time...i think the most complicated thing was done, now we need a community growth...and lots of amigans are really fannatic, and didnt like the ideia of aros...for me the look´n´feel is nice,but much better than that is the enviroment! =o)

by DonnieA2 on 2004/4/12 3:44:08

Quote:

Although in my personal POV mirrors yours (if I need Amiga apps, I'll run UAE), I'm wondering how much of the work (code)bringing UAE intergration into AROS could be used in bring AROS emulation (sandbox) to AROS as well? If both can share alot of the code, it maybe well worth supporting intergration of UAE into AROS as a step towards the sandbox.

Dammy

I would just say that from my perspective I could care less about Amiga 68K binaries on AROS. I want it to be it's own beast. We should have a platform that's got it's own software and doesn't rely on running past binaries more than it already does. AROS is not exactly a "NEW" project in the open source world. It's coming along. I don't want to see this project die or some other company come after it because it looks TOO much like someone else's product.

It's a cool OS on it's own. I hope it remains so..

Well, the 68K library structures would need to be identical. It doesn't influence the AROS libraries at all. At most, your thunk routine would maybe have to check some 68K library fields on entry, thunk to the native routine, update some 68K library fields, then switch back to the emulated 68K code returning from the library call.

It's extra overhead, but it's all native so it's rather fast compared to the emulated 68K code.

Well, as for thunking, we still run into endian issues. And for that to work, the structures must be identical, in offset and contents, to the original Amiga 68K headers.