X-Authentication-Warning: delorie.com: mail set sender to geda-user-bounces using -f
Date: Sat, 24 Oct 2015 01:28:05 -0400
Message-Id: <201510240528.t9O5S5qc019927@envy.delorie.com>
From: DJ Delorie <dj AT delorie DOT com>
To: geda-user AT delorie DOT com
In-reply-to: <20151024050756.GA5741@recycle.lbl.gov> (message from Larry
	Doolittle on Fri, 23 Oct 2015 22:07:56 -0700)
Subject: Re: [geda-user] Star shorts
References: <CAM2RGhQhNkShqtN8uyCSd7R+V4FgLXDZTvTC2RTdmKF7C_8ndA AT mail DOT gmail DOT com>
 <201510220136 DOT t9M1a5Uw015222 AT envy DOT delorie DOT com>
 <CAM2RGhQOpnBXbZXngNq=RcQTZN=UrU+PGzTZ9AC=xKNfu_zQkQ AT mail DOT gmail DOT com>
 <201510220149 DOT t9M1nrIe016145 AT envy DOT delorie DOT com>
 <CAM2RGhQ26+gpcAeTdgJpBEh0egOGLYcpizQ+B6kfTQLowEk5xQ AT mail DOT gmail DOT com>
 <20151022023002 DOT GA25952 AT recycle DOT lbl DOT gov>
 <CAM2RGhQ7xfMDe1knmiFkW8MAZ4Sr=ZSas8Ca9Sadv3gJT2z3gw AT mail DOT gmail DOT com>
 <201510221643 DOT t9MGhFfg003310 AT envy DOT delorie DOT com>
 <20151022170259 DOT GA28154 AT recycle DOT lbl DOT gov>
 <CAC4O8c-AGwV59t6-K+73eaZt0d43bJe0kE8E68n9pWxaa9gP+Q AT mail DOT gmail DOT com> <20151024050756 DOT GA5741 AT recycle DOT lbl DOT gov>
Reply-To: geda-user AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: geda-user AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


> There are multiple angles to this problem.
>  - the gschem end
>  - the rendition of this semantics in (one of many) netlist formats
>  - the treatment of that netlist information in board layout (or other hardware)
> 
> At the gschem end, and the netlist representation, I don't see any
> sensible approach besides having a symbol and a component.

Perhaps gschem needs a separate "star point" construct.  For example,
a large dot you can place, and any nets that connect at that dot are
considered "connected but separate".  No need for custom symbols or
pre-guessing pin counts.

As for the netlisters, they'd each have to do what their downstreams
expect.  Internally, we could have netlisters that know about star
points to call a guile function that enables them and disables
warnings about multi-named nets.

> On to PCB (or any other layout too; John Doty can close his ears for
> this part).  AFAICT, there is no established technique to implement
> such a star short in a way that will pass DRC.  There needs to be
> copper half-inside the DRC process, that definitely shows up on the
> Gerber output.  This copper can't exist during the netlist check
> (optimize rats).

<[Blue skying> We've talked about being able to "label" copper with
net names, for example labelling a polygon with "net=GND" to override
what DRC and rats guess at.  We could allow labelling copper with
"net=GND,AGND" or something, to say that it may be connected to
multiple nets.  That would allow us to use a line, arc, polygon, via,
pin, etc as the "star point".  Such labels might also help with short
detection, by limiting where DRC might guess a short would be.

Of course, anything like this would predicate a new DRC :-(

A label of "net=<ignore>" could be used for antennas etc, although if
we had true copper-in-footprint we might need to rethink how to tell
"copper that's a net" (like tying two gnd pins together) from "copper
that's not a net" (like an antenna).
</Blue skying>