Re: gEDA-user: Schematics and layouts' nets thoughts

[Steve Meier <smeier@xxxxxxxxxxxxxxxxxxx>]

Didier Villevalois wrote:
> Hello gEDA users,
>
> While making scripts to process gschem's schematic files and pcb's
> layout files, I bumped into two things that makes me scratching my head.
>
> Let me ask you whether i should make feature requests...
>
> Schematics files have nets defined as lines which have only start and
> end coordinates. So to know if a net's end is on a symbol instance pin
> you have to process each symbol instances and determine whether one
> their pin is as that net's end coordinate...

No, a pin can connect to the middle of a net segment.

>
> I always thought that schematics should be logical files and not only
> graphic files. So would that be possible to have two additional
> arguments to nets (one for each net's end) to reference a symbol pin or
> another net's end.

I think the confussion is that libgeda uses the term net when it realy
should be a line segment of type net or of type bus or just a graphical
line.

net type pins connect to net type segments

bus type pins connect to bus type segments

graphical line segments are just their for us humans

A net exists after the netlister has processed the schematics and
symbols. A net will have a name and then a list of pins associated with
the component. But no segments at least until layout at that time the
segments are added.

net_1 U1-2 U3-4      or net_1 has two pins one being the 2nd pin of U1
the other being the 4th pin of U3


>
> Also, PCB's layout file have lines in layers to define electrical
> connectivity. However, if one wants to know which net this line is for,
>  one has to compute closure of the line and see if it crosses an element
> pin and then finding the corresponding net in the embedded net-list.
> This is what already does PCB, if I understand correctly how it works.
>
> Maybe lines could have an additional argument that can be used by PCB to
> reference a net from the embedded net-list. That way scripts could
> easily extract subsets of a layout from a set of element refdes by only
> computing a net-list subset and then corresponding lines (without having
> to recompute what net a line is defined for).
>
> Should i make two feature requests or those ideas completely stupid ??

making feature requests may be premature. Trying to understand the
complexity of the data structures, their use and how to improve them is
far from stupid.


>
> Thanks for your help.
> Best regards,
> Didier.

_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user





_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user