I wrote a custom netlister over a year ago that suppose this and hierarchical netlisting.
http://spnet.code-fusion.net/ On Tue, Mar 8, 2011 at 4:18 PM, Colin D Bennett <[email protected]> wrote: > Has there been any work or planning toward making it easier and > less error-prone to produce schematics using “tidy” single-pin nets for > power symbols and I/O port symbols? > > By “tidy”, I mean that the completely unnecessary “:1” should > not be shown on the schematic. I try to name my power nets with > descriptive names (not just VDD, etc., in most cases). Also, the I/O > port symbols (input-1.sym, etc.) are often very useful to produce clean > schematics, but if you want a truly clean schematic without the clutter > of dozens or more “:1” suffixes, the only way I know is to hide the > ‘net’ attribute and use a separate ‘value’ attribute. The problem is > that these must be kept synchronized or you will have serious errors in > your circuit! You would have to manually check every pin for > consistency. > > (NOTE: When I say “completely unnecessary”, I mean that I never use > buses/multi-pin nets so the “:1” is redundant in this context for me. > Of course it could be assumed if omitted... hint hint hidden option 3.) > > Currently I see only two choices when you use I/O port symbols or > custom power symbols: > > > Option 1: Put up with the clutter of the “:1” suffix on every net > connection. > Example: generic-power.sym > Pros: + No risk of errors: the displayed text is the real net name. > Cons: - Visual clutter in the schematic. > - I never use multi-pin nets anyway; the “:1” is meaningless. > > This even worse when you have an array of nets that end in a digit. For > instance, I have a circuit that has INNER1, INNER2, INNER3, etc., and > these use input/output port symbols so they appear as INNER1:1, > INNER2:1, INNER3:1, etc. on the schematic. It's very difficult to > parse the actual net name by eye when there are dozens of such pins. > > > Option 2: Hide the ‘net’ attribute and duplicate the name in a ‘value’ > attribute. > Example: input-2.sym (aften adding and editing ‘net’ and ‘value’ > attribute values, hide the ‘net’ attribute) > Pros: + Clean, readable schematic. > Cons: - Risk of errors if ‘net’ and ‘value’ attribute disagree. > - Tedious to edit both attributes for each I/O port symbol > on large schematics. > > > Hidden option 3: Make gschem assume “:1” for ‘net’ attribute values > without a “:N” suffix. > > This seems like it would be an easy fix for the problems resulting > from Option 1 and Option 2, but I'm no Scheme or gschem hacker. > > I understand that the ‘net’ attribute is used in a number of different > ways, but for this specific purpose it would be nice to have a good > solution if it's not too difficult. > > Another option would be to add a DRC check to the gnetlist backend to > ensure that ‘net’ and ‘value’ attributes agree, thus eliminating at > least the possibility of a hidden error in the schematic. > > Thoughts? > > Regards, > Colin > > > _______________________________________________ > geda-user mailing list > [email protected] > http://www.seul.org/cgi-bin/mailman/listinfo/geda-user > _______________________________________________ geda-user mailing list [email protected] http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
