DOS ain't dead

2008-05-13: 32-bit v00.26.11

- Added a diagnostic error when (n < n) is used. This error is displayed when a compare is used when a shift might have been wanted. To disable diagnostic errors, use the /d switch
- other bugs/modifications (0)

16-bit assembler is not upgraded - still at v00.24.95 [2002-05-12]

Home page / downloads:
http://www.frontiernet.net/~fys/newbasic.htm

> http://www.frontiernet.net/~fys/newbasic.htm

Nice toy ... what can you compile with it ? HX ? RX-DOS ? Leaked M$/DR-DOS sources ?

---
This is a LOGITECH mouse driver, but some software expect here
the following string:*** This is Copyright 1983 Microsoft ***

> > http://www.frontiernet.net/~fys/newbasic.htm
>
> Nice toy ... what can you compile with it ? HX ? RX-DOS ? Leaked M$/DR-DOS
> sources ?

Don't ask me - I know nothing. Maybe the author knows.

> > > http://www.frontiernet.net/~fys/newbasic.htm
> >
> > Nice toy ... what can you compile with it ? HX ? RX-DOS ? Leaked
> M$/DR-DOS
> > sources ?
>
> Don't ask me - I know nothing. Maybe the author knows.

It's supposed to be Masm v5.1 compatible. Maybe the "Gone Forever" source modules are good candidates for a test? IIRC they've been written with Masm v5.1.

---
MS-DOS forever!

> > > > http://www.frontiernet.net/~fys/newbasic.htm
> > >
> > > Nice toy ... what can you compile with it ? HX ? RX-DOS ? Leaked
> > M$/DR-DOS
> > > sources ?
> >
> > Don't ask me - I know nothing. Maybe the author knows.
>
> It's supposed to be Masm v5.1 compatible. Maybe the "Gone Forever" source
> modules are good candidates for a test? IIRC they've been written with
> Masm v5.1.

At least by default they don't assemble (tested with 00.26.11 and Nov. 1's srcs).

BTW, I have a dinky util I wrote that assembles (also) with NBASM32. Here's some quirks, FYI:

* not all FPU instructions are implemented
* seems very picky about "word ptr" etc. NOT being used in some cases
* very slow (at least on old machines)

But hey, it's not bad, just could use some improvement. (But it's a one-man project, closed src, so I can't help, not that I'm so useful anyways.)

---
Know your limits.h

> At least by default they don't assemble (tested with 00.26.11 and Nov. 1's
> srcs).

Yes. I also tried it with some sources. Apparently it doesn't know how to handle forward references. Luckily it knows what a label is ...

> * very slow (at least on old machines)

Well, yes, it's compiled with DGPJJ, what did you expect?

> But hey, it's not bad, just could use some improvement. (But it's a
> one-man project, closed src, so I can't help, not that I'm so useful
> anyways.)

Ok, but "selling" it as "Masm v5.1 compatible" requires some boldness.

---
MS-DOS forever!

> > At least by default they don't assemble (tested with 00.26.11 and Nov.
> 1's
> > srcs).
>
> Yes. I also tried it with some sources. Apparently it doesn't know how to
> handle forward references. Luckily it knows what a label is ...

Like I said, it doesn't like "mov ax, word ptr [blah]" (assuming "blah dw 0") for some reason (IIRC).

> > * very slow (at least on old machines)
>
> Well, yes, it's compiled with DGPJJ, what did you expect?

No, that's not it. NASM and YASM are much much faster.

> > But hey, it's not bad, just could use some improvement. (But it's a
> > one-man project, closed src, so I can't help, not that I'm so useful
> > anyways.)
>
> Ok, but "selling" it as "Masm v5.1 compatible" requires some boldness.

Yes, well, it's more MASM-y than NASM-y, wouldn't you say?