Re: gEDA-user: Hierarchy viewer and database proposal

[John Doty <jpd@space.mit.edu>]

Carlos Nieves Onega wrote:

> > > Because it's not automatically generated. You can place a 7400
without 
> > > defining (or forget defining) slot and numslots attributes, and I
can't
> > > assume a 7400 has only 1 slot just because the user forgot to define
the
> > > attributes properly.
> > 
> > Why not?
> 
> Because the DRC can't trust the user. A DRC is there to find user's
> errors.

Hmm, if the user had numslots=1 on a 7400 you'd trust it? How's that
different from trusting a sensible default?

> I'm not saying the user has to add those attributes manually. That will
> be really a pain!.

You're assuming users only use pre-defined symbols. I spend a great deal
of time defining symbols. Here, all attributes must be added manually. Why
make it more difficult than it needs to be?

> Gschem has the right information in the symbol to do it itself,
> so the
> user shouldn't have to bother about them.
> What I'm saying is that gschem should define numslots and slot
> attributes when placing a part, as it does with refdes and value
> attributes.

I don't think numslots should be promoted: it's a property of the symbol,
not the instance.

Non-slotted parts should not have attached slot attributes, as it just
contributes to the clutter.

John Doty		"You can't confuse me, that's my job."
Home: jpd@w-d.org
Work: jpd@space.mit.edu