[Coco] HDB-DOS build fixes. Untested as of yet.

[Robert Gault]

Juan Castro wrote:
> Fixes are in http://users.vialink.com.br/jcastro/hdbdos14/
>
> The 38400 code wasn't being compiled in for the CoCo 1. There was code
> repetition. Now the server does recognize the commands from the CoCo, but
> all reads get a CRC error according to the server log. I see lots of
> retries coming from the CoCo.
>
> One more thing about the conditional assembly: the code repetition in
> dwwrite was like this:
>
> If blah
> Do something
> Else If symbol H6309-1 is nonzero
> Do ABC
> Else If symbol BAUD38400 is nonzero
> Do THECOCO1THINGIWANT
> Else If symbol H6309-1 is nonzero
> Do ABC
> Else ...
>
> I cut out the first IF for the H6309-1 block, which is identical to the
> third. But looking at it, it shouldn't make a difference if H6309-1 is not
> defined. But it does. Different code is generated. I think there's still
> something screwy with conditional compilation. Maybe the assembler
> interprets H6309-1 as a calculation yielding a nonzero value. Maybe I
> should find-and-replace this symbol and use an underline character instead.
> Gotta test more.
>

I agree that the Drivewire HDBDOS1.4 for a Coco1 does not work. I've not had the 
chance to carefully look at your patches but the .patch file seems to deal only 
with dwwrite.asm.

Both the dwwrite.asm and dwread.asm may need to be patched as the same 
convoluted nested conditionals are present in both.