[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Can not compile MASH source code !

The problem here is that Makefile and Makefile.vc are maintained
separately.  Dali stuff has been moved to the optional part for
compilation under unix.  Dali has to be downloaded separately just as Wei
pointed out.  In Makefile.vc, just comment out ENABLE_PSVP.

As for the other bugs, I suspect it is the difference between VC6 and VC5.
Mash has not been very friendly with VC6.  I remember Peter modifed a
bunch of the code so that it would compile nicer under VC6, but I'm not
sure if he checked in those changes.

This brings up the question of maintaining the windows port in Mash.  I
would hate to see Lloyd going through another set of 500+ compile
warnings.  I have tried to download cygwin and compile Tcl/TK.  No success
yet, but it's probably just a small bug in my distribution.  The question
I'd like to bring up is to keep support Makefile.vc or to move to cygwin.
Currently there are pro and con with both parts, and it's not very clear
to me which way is the best to go.  Any thoughts?

Here are some pro and cons of each I can think of:

VC:
+has worked
-need to maintain another Makefile.vc, which cannot be generated
	automatically
-different versions: VC5, 6, 6+?
-It's M$, but then again, everyone seem to have access to one, except me
*Debugging can use Develop studios, but I have not gotten it to recognize
	the Makefile and create a workspace.

Cygwin:
+Uses the virtually same tools as unix build
+free (that's why I'm trying it as I have no access to a VC)
+Uses gdb to debug, with some nice GUI
-Still fairly beta
-Developer has to download cygwin (although the process is probably easier
	than installing VC)
*Two different Make modes: Unix and Win32, which is still not clear to me
	when building Tk.  I'm assuming Win32 is better, as it doesn't need X
	libs, but who knows?

--Paul

On Sat, 11 Nov 2000, Lawrence A. Rowe wrote:

> Which Windows OS are you using?  The code will compile after some effort
> on WNT, but we are still trying to resolve some problems with W9x.  The
> problems you mention don't sound like the ones I ran into, but it has
> been a while since I did it.
> 
> Can someone else help him?
> 
> As for Dali, that is a library of routines for doing image processing.
> It is used in one of the applications that you won't be using, the
> parallel special effects system (psvp), so you don't need to worry about
> it. The encoder-h261 is more important.  That code is in the core of the
> system and should compile just fine.
> 
> Paul, Peter, or Lloyd, can you take a look it this problem?
> 
> We really need to get the clean compiled version for windows done.  I'll
> try to do a compile on my WNT desktop of the new version that Lloyd just
> checked in with the compile errors fixed.  Peter, any progress with W9x?
> 	Larry
> -- 
> Professor Lawrence A. Rowe          Internet:  Rowe@BMRC.Berkeley.EDU
> Computer Science Division - EECS       Phone: 510-642-5117
> University of California, Berkeley       Fax: 510-642-5615
> Berkeley, CA 94720-1776            URL: http://bmrc.berkeley.edu/~larry
>