The point of view of ECB (which is currently the heaviest tool using CEDET, at least IMHO ;-) to this topic:
I second the suggestion of Ming-Wei. It is very big potential root of many problems if a tool like ECB has to deal with two slightly different versions of CEDET - on one side the emacs-integrated one on the other hand the author version... we need interface stability.
I vote for a marriage! :-)
Thanks to all people working on CEDET regardless which one.... ;-)

Von: Chang Ming-Wei []
Gesendet: Dienstag, 23. Februar 2010 04:58
An: Eric M. Ludlam
Cc: Oleg O Andreev;
Betreff: Re: [CEDET-devel] Different interfaces in bzr Emacs and CVS Cedet.

Hi Eric,

    Thanks for the putting a lot of effort in CEDET. I am thinking that synchronizing CEDET CVS and CEDET in emacs 23.2 is very important. It is probably the most important thing right now. Emacs 23.1.93 is going to pretest pretty soon. Once emacs 23.2 is released. People is going to develop a lot of new stuff on top of CEDET so similar situations (like what happened in this thread) are going to happen again and again. Therefore, I would like to make a bold suggestion: maybe we should combine CEDET CVS and CEDET in emacs 23.2 first before releasing CEDET 1.0. What do you think?



On Mon, Feb 22, 2010 at 5:04 PM, Eric M. Ludlam <> wrote:
On 02/22/2010 12:58 PM, Oleg O Andreev wrote:
> Hello!
> I've recently updated Cedet from 1.0pre6 to CVS HEAD and I've noticed
> that function 'semantic-ia-get-completions', which was present both in
> 1.0pre6 and in Cedet, bundled with Emacs from bzr trunk, disappeared in
> CVS version. This function is used by different completion mechanisms,
> such as company-mode or auto-complete. As far as I can see, there is a
> function 'semantic-analyze-possible-completions' which is roughly
> equivalent to 'semantic-ia-get-completions'. The question is, will
> 'semantic-ia-get-completions' be present in future versions of Cedet, or
> should completion modules be rewritten, so that
> 'semantic-analyze-possible-completions' is used?

Hi Oleg,

The function semantic-ia-get-completions was an internal function, and I
did not realize that multiple other tools used it.  I could certainly
put it back if needed.  It had some extra features that caused results
to sometimes be inaccurate.

The function semantic-analyze-possible-completions is the preferred
function to use.  Tools can use that in conjunction with
semantic-analyze-current-context to hold information and make subsequent
calls faster.

I can do whichever is best.


Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
Cedet-devel mailing list