On Nov 7, 2007, at 5:17 PM, Chris Albertson wrote: > On 11/6/07, Dave N6NZ <[EMAIL PROTECTED]> wrote: >> >> Steve Meier wrote: >> >>> What do we expect the schematic DRC to catch? >> >> The problems that apply to *my* situation and *my* technology. Also, >> violations of the in-house design style guides that only I use -- for >> instance, require a partnum attribute whose value exists in the >> purchasing database. >> >> In other words, generic DRC is largely pointless. DRC should be >> implemented as a DRC engine and a rules database. Of course, one >> hopes >> that a (or two or three) good, reasonably generic rule sets can ship >> with the release. End users can start with that and tweak to >> their own >> needs. >> >> The trick, it seems to me, is to come up with a clean and readable >> syntax for the rule set. The language needs to be declarative at its >> core, and allow user plug-ins written in C (so that I can run SQL >> queries against my parts database, for instance.) > > Have you looked at Prolog? It is an interesting language. A Prolog > program is a set of patterns and actions. For example to write a > sortting > program you could give as a pattern "x[i] >x[i+1]" and the action > would > be the exchange x[i] and x[i+1]. In Prolog data and program have the > same syntax
Yes. Like Lisp. And our DRC back ends are already written in Guile, a Lisp dialect. > > For DRC in Prolog, one way would be to first "facts" about the > devices and > the pins. These "facts could be written in Prolog and stored with > the symbol. > Rules would be patterns. These are declaritive and would say things > like > Power inputs must by connected to power sources. Source must be able > to drive loads,.... > > One of the nice things about adopting an existing language is that > there > are existing interpreters and impotently books and tutorials. Yes. We already have such a language, Guile. It makes little sense to me to create an additional dependency on another similar language. > > People have written very complex programs in Prolog, several so called > "expert systems" and it would be agood way to implement features like > autoplacement and autorouting. > -- > ===== > Chris Albertson > Redondo Beach, California > > > _______________________________________________ > geda-user mailing list > [email protected] > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user John Doty Noqsi Aerospace, Ltd. http://www.noqsi.com/ [EMAIL PROTECTED] _______________________________________________ geda-user mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
