Wouter van Ooijen wrote :
> > Maybe one should look into truly separate source files for
> > the different OS'es (just the parts that have to be different,
>
> I would go that way.

OK. I'll see how that could be handled. There are also other
issues like :

The "handle" you get when opening a file (or com port) in the
OS/2 code is an "int", in Windows it's a "pointer to void" (data
type "HANDLE"). So every function declaration that used this
handle must be different.

Rob has used "ANSI control codes" (or "escape sequences") to
highlight things in the screen output. I could not find out how
to enable this in a Win2K command window, so I had to make
an OS-specific part of this also. And besides, I'd like the code
to run on any Win2K box without special configuration.

So it's not just the low level functions that are different, some of
the core variables are of a different type, and must sometimes
have other values.

> > And, there is still the issue with keeping up with the "original"
> > XWisp...
>
> Well, I tried to add one more target chip to the Wisp628
> firmware but it is full, realy full. I could of course move on to
> a 16F648A, but I think I will redesign the whole thing, except
> for the hardware. So JE: how fast do you need this native port?

In about two weeks. In time for the Dwarf and Wisp delivery...

But, I don't see any major problems so far, part of the work is to
get the conditional compiling to work cleanly to keep the code to
build on OS/2.

Jan-Erik.

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu