gEDA-dev: Draft: Creating and maintaining a spice model library

[Werner Hoch]

Hi Peter,

On Dienstag, 6. November 2007, Peter TB Brett wrote:
> On Tuesday 06 November 2007 18:08:22 Werner Hoch wrote:
> > Mmhh, well. I've read the source. Should I look at a different
> > place?
>
> Yeah -- it generates detailed doxygen comments, including 2 dedicated
> pages on how to use the functionality.  Should it be put in a
> different place?

No of course not. But I've only read the function description and not 
the file head of the s_clib.c file to write my code.
I usually do not create the doxygen docs.

 *  The <b>component library</b> is made up of a number of
 *  <b>component sources</b>, each of which in turn makes available a
 *  number of component <b>symbols</b>.  Each source may be either a
 *  directory on disk containing symbol files, or a command in the
 *  system PATH which can generate gEDA symbol data (e.g. from a
 *  database).  A component source is represented by a #CLibSource
 *  instance.

The first sentence does not mention the scm interface, but everything 
else is really great.

> > 2. Copying of my generated symbols is slow in gschem. If I place
> > 100 symbols and create a copy, it currently takes quite long (about
> > 20s).
>
> Hmm... is it any slower for the file-based symbols compared to the
> command-based ones?

Yes. I don't notice any delay when copying file-based symbols. 

> If so, that's bad and shouldn't be happening -- the symbol data's
> supposed to be cached.

gschem calls my script over and over again.

Regards
Werner