gEDA-dev: Re: Bug reports for gattrib

[Stuart Brorson]

On Sun, 1 Apr 2007, Stephen Williams wrote:
> Stuart Brorson wrote:

>> However, my question is:  Why are you attaching attributes to the same
>> component which have the same name?   Is it something like
>> vendor/vendor part no. for different vendors?  And what should the
>> correct behavior be?
>
> I've used it to attach power pins in chips with lots of 'em.
> The net= attribute has special, magical powers and allows me to
> use a lot less space yet still make power connections visible.
> (Sometimes text is more compact and expressive then pictures:-)

OK, both respondents said they had problems with the net= attribute,
so I guess that is the only one causing problems.  It is also the only
one which makes sense to attach multiple times to a single component,
right?

The idea is that you want to have component attributes like this:

net=GND:5,7,9,12
net=VCC:2,4,6,10

>From the perspective of gschem, gnetlist, gattrib, and the rest of gaf
the net attribute is a little different from other attributes.  It is
acutally an attribute which designates *connectivity* to a pin,
instead of some component property.  That is, its value has to do with
the component's pins, not to the component as a whole.

In the ideal world, gschem would insist that the net= attribute would
be assigned to a *pin*, and would tie a particular pin to a particular
netname.  However, I realize this is a PITA, and the current usage of
net= is more convenient.   Therefore, I'm not suggesting we change the
usage of the net= attribute.

My proposed solution is this:  gattrib should just ignore the net=
attribute for now.  That is, it should't touch it at all.

In the future, when I implement pin attributes, net=
should somehow be handled by the pin attrib tab.  Probably each pin
displayed in gattrib's pin spreadsheet would have an attrib
designating which net it's attached to.  But when you wrote out the
.sch file from gattrib, it would reattach all net= attribs back
to the component, instead of the individual pin.

However, for now, net= should just be ignored when found.  Agreed?

Stuart