Re: c/c++ project management and debugging

[Elena]

On Dec 22, 2:12 pm, Andrea Crotti <andrea.crott...@gmail.com> wrote:
> Elena <egarr...@gmail.com> writes:
>
> > Same as: "emacs -Q" then eval sexp.  Only problem, my .emacs is more
> > than 200KB.  I would have liked to modularize it, but Emacs lacking of
> > IDE-like facilities to navigate Emacs Lisp made me give up that.
>
> emacs doesn't lack that, you don't know how to use it ;)
> C-h f and you can jump to any function you want.
> The inline documentation and the ability to trace/try out/see the
> implementation of all the elisp functions is wonderful.
>
> With semantic in your elisp file you can jump easily to any
> function/variable you want and so on.
>
> And let me understand, your .emacs is more than 200KB but you're not ad
> advanced user?
> How can that be?
>
> The only explanation is that you copied everything could sound nice
> inside and now you have no idea of what could go wrong.
> Well if that is the case first you should understand very well what's in
> your .emacs.

More or less.  When I started learning Emacs, I began copying snippets
of code without really understanding how they worked.  Nowadays, I
still have a difficult time understanding some of them.  I've become
used to the facilities they provide, therefore my refactoring is a bit
at a time.  Not all of such code gets evaluated, though.

>
> Mine is only 57Kb after a couple of years of customizations.
>
> > What I mean with "tracing":
>
> >http://www.iro.umontreal.ca/~gambit/doc/gambit-c.html#Procedures-rela...
>
> > scroll till the section "trace".
>
> > Profiling still lacking, anyway.
>
> trace-function?
> There are packages also for profiling, but I really don't think it could
> help solve your problem since it gets completely stuck (not just slow).

You are right about profiling: packages do exist which do a fine job.

Regarding tracing, I failed to explain, sorry.  I didn't mean such
kind of trace.  I was thinking about a logging facility, which logged
calls outside user control.  Such a facility, together with a profiler
and a call-tree generator, would have made easy tracking down the
function (or functions) which were causing the hang.

>
> > Done already.  Nobody has been able to help.  Currently Emacs lacks
> > needed tools.  I'm sticking to Emacs 22, which hangs much less.
>
> > I always try to be constructive in helping package maintainers to
> > track down bugs, but I'm not an advanced user.
>
> I think instead you lack needed knowledge instead, otherwise how do you
> think that the developers are able to do a wonderful work without those tools?

Nope.  I have submitted here the issue to experienced Emacs users and
they have not been able to help.  Thus it is not a matter of lack of
knowledge, it's just the lack of tools.