I think your concerns about the human ambiguity of a 'net' command with
no signal names are valid ones-- this crossed my mind as well.
(and yes, you caught my error in the change comment)
Personally, in terms of how to expose an explicit anonymous (??) net to
hal, I tend to prefer
net ??? pin => pin
where ??? is something that we think is exceedingly unlikely to be used
in any actual configuration file. But signal names have so few
restrictions on their forms that there's nothing (except a pin name!)
that seems to be forbidden now for that first argument in a 'net'
command.
"net * pin => pin" is the next closest I can think of. "*" is not quite
forbidden as a signal name, but it behaves weirdly enough with e.g.
'show sig *' that probably nobody would have used it.
I'm going to put this on the back burner for now. I think the branch
should make a fine starting place for implementing "anet" as well, with
just the last commit needing to be reworked. Perhaps someone else would
care to do this.
Jeff
------------------------------------------------------------------------------
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
