What Amiga Inc. and Hyperion NEED TO DO (IMHO)

[Waccoon]

koaftder:  Different processors, different operating systems, different libs, etc and the processor is of little concern to me. So, as a developer, if the processor means almost nothing to me, why the hell should an end user care?

Because assembly is, like, all of solid awesome and stuff!  Everyone wants to write PPC assembler!

rbsfou:  Release an "Amiga Classics" title for the Wii, Playstation, and XBox 360 (all ppc i think)

The problem is that Amiga Inc. doesn't have the rights to such games.  Also, the PPC CPUs are irrelevant if the original games are 68K.  Remember that Amiga titles that utilize PPC are simply using it as a coprocessor.  A 68K CPU is still required.

Rumor has it that Microsoft (of all companies) is going to be re-releasing new Amiga games.  Sensible World of Soccer is one such title mentioned.

rsbfou:  Also do a deal with nintendo to allow sale of Amiga Games for the virtual console.

The sad truth is that 3rd party developers have been all but shut out of Virtual Console.

rbsfou:  Do a deal with Opera to supply the browser (to keep Nintendo happy if this is a concern).

First, you need decent programming APIs.  OS4 is hardly sophisticated enough to drive Opera.

As for Nintendo, their version of the browser only uses the Opera rendering engine.  The actual browser itself is completely custom (and rather poor).

rbsfou:  Make sure the geeks get their X11 server, VNC client and server, SSH client and VPN client to keep them happy.

Might be difficult given that AmigaOS is not POSIX compliant.  ;-)

The reason why these apps and tools run on so many OSes these days, is becuase most OSes follow UNIX standards.  It's not easy being an alternative OS.

Now, if Hyperion had the sense to start with a good UNIX-like OS, like QNX, and simply write a new Amiga-like interface on top, that would've been truly awesome.  That's what Apple did, and look where they are today.

Amiga could've learned a lot from Apple.  Such a shame.

Continue to supply support to the "faithful" users of Blizzards, AmigaONE and Shark PPC's (give them a Cheaper OS, especially for AmigaONE owners who have the encryption hardware built in i think).

Yeah, all five of them!

Seriously, as much as I like supporting your community, it's a simple fact that Amiga has to get new customers -- lots of them.  It would be best to make AmigaOS an x86 system, and simply offer a PPC compile.

Writing drivers for all the PPC hacks would be a real pain, though.  Remember, to support a platform, the developers have to actually have the platform.  Testing hardware, and especially documentation, isn't easy to come by concerning these old Amiga hack boards.

rbsfou:  Don't bother with X86, too much work has been done on PPC, and most of the target market would have the hardware (be it a console, a possibly-now-spare-PPC-Mac as they've gone intel, or a specialized amiga system).

I was also under the impression that the whole point of an OS was so people didn't have to be concerned about the hardware.  You write against libraries that are designed to take advantage of special hardware features, thus allowing you to still utilize "custom" hardware.  DirectX is one of the best technologies ever invented, especially since it does a hell of a lot more than just graphics -- and it runs on both x86 PCs and the PPC-based XBox 360.

Game consoles programed with PC dev tools -- and developers are happy about it?  Say it ain't so!

rbsfou:  Once your "brand" is back to where it should be and you have loads of $$$, quit messing around with those lame phone and wince games!!

Frankly, I don't think Amigans really know what they want their brand to do.  Hence, there are still so many people that insist that they should have a custom PPC machine.  I don't care what hardware it uses, so long as it's a good value and does what I want.

I'll gladly rebuild a custom x86 system to get OS4, but I'm not going to pay an ungodly amount of money for a custom PPC that, if it breaks, will take months to fix and require me to ship it to another country.  Note that the AmigaOne wasn't a very well-designed board and was full of bugs, too.  That tends to be the problem with low-volume hardware (and even that was a "generic" evaluation board!)

IMNSHO, of course.

[Waccoon]

rbsfou:  Agree, but they own the brand name, and could convince those who do have the rights to let them be used (i've covered this in my first post), and as regards CPU emulation, we have UAE which could be made into something like WHDLoad to do it transparently (see my first post again!)

If the rights to the software do not belong to Amiga Inc., how will they, or the Amiga community, benefit from ports to game consoles?

rbsfou:  My point was that they could use this as a way of sneaking ppc-OS4 onto people's consoles and it would provide incentive to give us the emulation layer.

Why would any consoles use OS4 when they have their own OSes?  Why do you need OS4 to run old Amiga games?

rbsfou:  It has a task / thread scheduler, several widget APIs all based on boopsi, portected memory (FINALLY!), truetype font support and a built in tcp/ip stack. What else could be needed?

There we have it, folks.  If you have BOOPSI, you have a full-blown GUI toolkit that makes porting Windows software a snap.

Can't say I've ever heard of portected memory.

rbsfou:  No - i LIKE AmigaOS for not copping out and being another Unix clone.

The point is that you're asking that software that employs UNIX standards should be ported to a non-UNIX OS.  That either takes a LOT of work, or you basicly have to make an entire UNIX layer run on top of AmigaOS, which is anything but efficient.

I think AmigaOS can do without an XWindow Server and all that other crap, so long as it has its own tools.

rbsfou:  Fair enough - personally i don't like x86 though, but as i said before, each to their own

All you said before is that you think x86 has too few registers.  That's hardly a good reason to dislike any CPU architecture, given that assembly is almost obsolete these days.

Even SIMD acceleration is usually accessed through APIs, rather than hard-coded assembly.  In fact, using assembly for most applications is quite bad.  Refer to problems where Apple software runs slower on a G5 than it does on a G3.

rbsfou:  Fine so let me run it on a PPC Mac (now/soon) or some x86 thing (later)!

How?  Switching architectures is not an easy thing if you design an OS the way you describe.  You have to plan for CPU abstraction ahead of time, not tack it on at the last minute.  Once you introduce Altivec/VMX, even emulation is pretty much impossible, let alone practical.

There are reasons why things exist the way that they do.

If anyone wants to make suggestions on how AmigaOS should be developed, they really should, at a bare minimum, have real experience with 10 different OSes.  Have you ever tried OS/2, VMS, RISC OS, and Inferno?  Your priorities and suggestions might be very different once you see how other "alternative" OSes have been developed in the past, and why they have failed to penetrate the mainstream market.

[Waccoon]

Dr_Righteous:  OOP lanugages have gotten SOOOOO out of control, and are the primary cause of OS/Application bloat (seconded only by pure programming laziness).

It's the over-use and abuse of OOP that causes bloat.  A good programmer knows when and when NOT to use OO.

As usual, the computers just follow instructions and don't make mistakes, but people do.

Dr_Righteous:  I'd much rather see the microcode disappear, and the transistors on the die along with it. Things would all be SOOOO much faster.

Well, so long as Core2 Duo blows away everything else in terms of performance (and it runs pretty cool, too), that comment holds no water.

Also, x86 is a winner in terms of overall performance.  Other ISAs can beat x86 in certain areas, but usually flop on general-purpose runs.  There's a reason Apple always used Photoshop benchmarks to prove the speed of PPC, although the performance could be wildly inconsistent between CPU generations.