OK, some small success making Vasilis Margioulas's Cells-Gtk for CLisp
and Gtk2 portable:
Vasilis's original version has been removed from the Cells project CVS
tree and installed (again, it seems) in the Cells-Gtk project here:
http://common-lisp.net/cgi-bin/viewcvs.cgi/clisp-cgtk/?cvsroot=cells-gtk
As for the new UFFI-ized version, that is here:
http://common-lisp.net/cgi-bin/viewcvs.cgi/root/?cvsroot=cells-gtk
That now works under AllegroCL and Lispworks on win32. The source as it
stands as of this writing shows only one of the nine test panels Vasilis
provided. The others can be brought into play by de-commenting them in
test-gtk.lisp. (You cannot miss it.) They all might work in AllegroCL
(they did before all the work to get Lispworks working), but under
Lispworks who knows, this is where I stopped to share since a couple of
people had offered to work on the Lispworks port.
The UFFI-ization was done weirdly, because I was hoping to keep it
working under CLisp which is not supported by UFFI. So for a while I
conditionalized the code for #+/-CLisp, and definitions were left as
CLisp-y and translated to UFFI by macrology, instead of just globally
editing everything to actually be UFFI code.
Now I think I want to fork UFFI into what will become a different
portable FFI not necessarily backward-compatible with UFFI. I gather
there has been talk of doing so for a while. I would add callbacks,
CLisp, booleans, and some handy utilities I have developed for dealing
with FF. This will include providing friendly wrappers which will, for
native FFIs which do not do it automatically, translate between nil/t
and 0/1, or nil and a null pointer.
Anyway, just do not freak out at the weird way the UFFI-ization went, I
changed horses midstream and it shows.
Contributors are welcome, as always. In case you missed the incentive,
here are four of the nine panels:
http://common-lisp.net/project/cells-gtk/
Those shots come from Vasilis's original CLisp release, running on
Linux. I would be especially interested if some brave soul wants to try
building GTK2 under X11 on Mac OS/X and see if AllegroCL or Lispworks
can run Cells-gtk over there as well. No reason other OS/X Lisps cannot
try, but then I would expect more UFFI-ization issues to fall out.
If anyone does, screenshots as well as patches are of course welcome.
kenny
--
Cells? Cello? Celtik?: http://www.common-lisp.net/project/cells/
Why Lisp? http://alu.cliki.net/RtL%20Highlight%20Film