Kai-Martin Knaak wrote: > I like to use separate power symbols for ICs. However, I constantly hit > the usability issues associated with this approach. Support for multi- > part symbols is quite fragile. > For example, gnetlist just checks the footprint attribute of the first > symbol it encounters. That is, the footprint actually used depends on the > relative position of the various parts in the *.sch file. There is no > check for completeness, no warning if a power symbol is missing. This is > clearly an invitation for nasty surprises. > > My time schedule is a bit relaxed in the weeks to come. So I thought, I > might engage in contributing to the geda project by tweaking this aspect > a little. ...
Great idea! > ... I figure, that it may not be too difficult for a low time > hacker like me. This is a list of sub goals for this little project: > > 1) gnetlist should look for a footprint in every instance of a refdes > > 2) add a known attribute "parts" that lists all symbols of a component > (should the entries in the list be separated by space, or something > more sophisticated?) > > 3) gnetlist should complain if the lists of attributes are inconsistent. > > 4) gnetlist should complain if there are different footprints present > > 5) gnetlist should complain if a component is not complete (part missing). > Yes. This avoids forgetting to place U4D and then its input is floating or the feedback trace missing. > 6) gschem should read the parts list and insert all parts at once. > > 7) auto number of gschem should keep multi-part symbols with the same > refdes. > > Next step would be to treat slotted components like multi-part symbols. > 7) and your sentence after that is the main reason why I do not use gEDA and currently do not recomend it to clients. (Ales: Don't explode now ...) While you are at it, another feature would be really useful: Be able to select that auto-renumbering leaves U1A, U1B, U1C and U1D in exactly this order. It can re-assign it to U12A, U12B and so on but not mix with other refdeses and not switch B and C and so on. Such a lock-in isn't needed for all designs but anyone who has done very drift-critical analog designs knows what I am talking about. Another nice feature would be if gschem would allow parts with inherent supplies on the A-slot, meaning it'll always be there when you place it but not on the B, C and so on parts. This is how Kicad does it. Many engineers prefer that versus a "floater" power symbol that you scoot over the part. > Any comments? > No, just that you had a good idea there. This could really give gEDA a boost and more acceptance. -- Regards, Joerg http://www.analogconsultants.com/ _______________________________________________ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user