Hi Eric & Richard,
> EL> Hi,
> EL> As David pointed out, the Makefiles checked into semantic
> EL> could not be regenerated via EDE. To my dismay I found a
> EL> plethora of changes waiting to be checked in, so I did.
> EL> Anyone who gets the new EDE files should be able to rebuild
> EL> the Makefiles in semantic in a way that they will come close
> EL> to what is in CVS. I'm getting clean builds on both EDE and
> EL> Semantic now.
> EL> I think the Project/Makefile may need to be double checked.
> EL> I'll look closer when I get a chance.
> It seems like mention of semanticdb-obj.el needs to be
> deleted from semantic/Project.ede, because I can't find that
> file anywhere.
> I can run "make" in semantic directory successfully after
> updating new Porject.ede from cvs db, run
> ede-proj-regenerate, then delete mention of
> semanticdb-obj.el from the resulting Makefile.
I also removed the ":partofall 'nil" option from the "doc" target of
wisent/Project.ede, so wisent.info files are build too (like
Finally in wisent/Project.ede I switched targets "languages" and
"wisent" to first compile core wisent library. So the compilation of
LALR automatons will be done by (faster) byte-compiled code.
Attached you will find a global patch for new Project.ede and
I checked the "make" process and all seems fine :-)
I can check these changes in, if no objection.
> EL> I also checked in the following semantic files:
> EL> semantic-c.el: support #elif
> EL> This may need porting back to v1p4
OK I will do that.
> EL> semantic-el.el: support defvar-mode-local
> EL> This is a nifty change that makes any language support file
> EL> (like semantic-el.el, for instance) look cool in speedbar,
> EL> treating the mode much like a class with externally defined
> EL> methods in C++.
I tried it, and it is a really cool feature. I like it!