Im currious... these days feel im slowing down on topics im interested in programaticaly. There are alot of tallented people out there that are doing great things that I personally dont wish to compete with. So im often wondering what to take on next.

To that end, i thought i would ask just what things about MASM32 and Win32 do you wish was somehow 'simpler' or better in some way. Or certian aspect of HLL's that have not been somehow provided for in Masm32 (Personally i dont think there is anything HLL can do that Masm32 cant, but im open minded ;) )

Anywho.. it thought i would open a forum for open "gee that would be great" conversations. At the very least it should be entertaining to see what kind of diverse opinions i can dig up ;)

Regards,
:NaN:
Posted on 2004-06-14 00:18:32 by NaN

Personally i dont think there is anything HLL can do that Masm32 cant, but im open minded

Automatic register allocation, link-time code generation... ;)
Posted on 2004-06-14 08:54:41 by f0dder
Masm wish list... Hmmm.... Maybe Fasm's virtual directive and some compile-time funcions of HLA added to Masm. :)
Posted on 2004-06-14 14:37:27 by Vortex
I get the feeling this idea isnt going to go very far.....
Posted on 2004-06-14 15:36:14 by NaN
Trolling aside, masm fits my purposes fine... that is, provide people with example code, quickly hack something together, or sometimes for linking with C code.

For more advanced stuff, I tend to use nasm or fasm, as they're better suited for some of the weird and tricksy stuff. I also tend to use nasm or fasm when linking with C code.

If I did full-application assembly coding, or nasm/fasm didn't exist, there'd be some amount of things I wanted in masm, though.
Posted on 2004-06-14 17:16:26 by f0dder
Well im somewhat of a purist. I find MASM32 complete for the most part. I've never needed to try other assemblers because MASM hasnt failed me yet.

Regards,
:NaN:
Posted on 2004-06-14 18:33:28 by NaN
direct binary output, ability to change the "expected execution address" without changing where the code is assembled... useful when writing bootsectors and other tricky code.
Posted on 2004-06-14 18:46:04 by f0dder
Is there a MASM released yet with SSE3 or x86-86 support?
Posted on 2004-06-14 18:54:14 by Scali
There is masm for x86-64, was discussed here more than half an year ago :)
Posted on 2004-06-15 01:11:22 by Ultrano
Yea well I must have missed that thread then. Where can I download it?
And what about SSE3?
Posted on 2004-06-15 16:06:32 by Scali
Hi Scali,

You can download masm64 from:

ftp://ftp.sandpile.org/ml64.zip
Posted on 2004-06-16 12:36:17 by Vortex
Well, as far as I can see, it can generate only 64 bit code (it doesn't seem to understand .686, .Model, .XMM etc directives... Only if I remove them all, it works). And it doesn't understand SSE3 either.
So I wonder, how can I write SSE3 code?
The only way I have found so far is to use inline asm with Intel's C/C++ compiler.
Posted on 2004-06-16 12:53:33 by Scali
Hi Scali,

And it only uses FASTCALL from what I could tell when I tried it, MS pretty much gutted MASM to make it a side tool for C only from the looks of it. Though I think it was just a quick fixup to be able to say they supported the instruction set until a full version is finished. I had alot of issues when I tried it and ended up just deleting it as a useless tool, better to wait until the full version comes out in a processor pack.
Posted on 2004-06-16 14:52:59 by donkey
Handle correctly if/whiles and other structures (register allocation?), and not intel sintaxis , also for support old sintaxis, a internal translator should output a separate file with the sintaxis that is the 'new' one.

There are a lot of other things, but that is apart ;).

Have a nice day or night.
Posted on 2004-06-17 12:40:36 by rea