On Fri, 18 Feb 2005, Barna Farag� wrote:

> Hi Anthony,
>  I was using GtkWave 1.3.19 windows port. The main problems about ps timeing
> was on my code ( my code generates directly .vcd files)... I used a wrong
> vcd fileformat spec. Thanks for help.

Glad to hear you figured out the problem.


> By the way, I start to porting the latest version to Winsux in my rest
> time... If anyone could/want help me from this list :)

-rw-rw-r--    1 bybell   bybell    2346381 Feb 18 08:30 gtkwave-1.3.54_w9x.zip

1.3 series appears clean and compiles via a ./configure ; ./make from the
MinGW command line.  As compared to the unix version it's exactly the same
except for 3 things:

1) no "open new viewer" due to lack of fork()
2) VZT loading isn't multithreaded on SMP machines (all my SMP boxes run
unix.)
3) slower.  this was covered on geda-dev in detail recently.

...because of #3 I really need to look at the gdk library source for
windows sometime to see why that's the case.  I might be better off
drawing lines/boxes into my own private bitmap through software or
whatever.  I'm really curious if they're doing something similar to an
XSync() (for windows) after every graphics op.  *shrugs*


On Fri, 18 Feb 2005, Aubanel MONNIER wrote:

|Actually, I did port a pre-version of GTKWave 2 to windows (Visual C++). I
|cannot easily access the source, but this is mainly plumbering (the hard
|parts are to get a correct config.h and the management of shared object
|code). Problem is the solutions I've used to solve this problems are quite
|ugly (use of a custom config.h + modifications at build time, all the code
|built staticlly, etc.). If you are interested, I can ask for the source.

For me, MinGW was definitely an easier way to go about porting as you
don't have to worry about fighting the Visual C++ environment.
Experienced VC++/Windows users might disagree as to what environment is
better.  =)


Regards,
Tony





Reply via email to