[MiNT] Re[3]: usage of wind_calc()

[Martin_Elsaesser@ll.maus.de (Martin Elsässer)]

Hallo evan,

>So, are you suggesting we just not add anything new to the AES?

No! It's a good thing to add new features to the AES. But for applications
that should support this features, it must give a way to run on other OS,
where these features are missing.

>Providing emulation of every new AES feature within the program is just
>silly - its faster and less code to just not have the new features in
>the first place if you are going to provide equivalent emulation code
>anyway!

But this would be the only way for applications to run on OS without these
new features. It must give a way, to implement new features to XaAES and
build a - hopefully small - TSR which implements these features - or
emulate them - on other OS. Than applications could use new features of a
modern AES and could run on "older" OS.

>The other possibility is perhaps compiling parts of MiNT for PPC native
>code and using the NatFeat capability of aranym to run those parts
>natively.

That would be a very good thing! THings would run much faster than now, I
think!

>If my guess is right, MagicMac is so much faster because there is a PPC
>native version right?

MagiCMac 6.2 uses the JIT of MacOS Classic. And MagiCMacX has an optimized
68k-Emulator written in PPC-Assembler. So MagiCMacX is much slower than
MagiCMac 6.2, but faster than ARAnyM. If ARAnyM could run as much OS-code
as possible in PPC native, ARAnym would speed up dramatically, I suppose.

>I'm broke too.  However, thats no reason to not improve the system.

That's right!

Gruß

     Martin    [PGP-Key available]
---
Für Mails >16 kByte: martin@melsaesser.de