On Fri, 2002-11-22 at 20:30, Frank Miles wrote:

> There are a few essentials for me, at least:
> (1) the netlist, which is generated from the schematic, must always be
>       able to sync up with the PCB via DRC;
> (2) when design changes occur (for this implementation or the next generation
>       hardware), re-use of the existing layout should be maximized.

I'm not sure if this is what you mean, but here is another approach to
the "problem" of back-annotation that *might* work:

Any changes in gschem will be transferred to the netlist in the pcb
program by "a click of a button".

So if you want to swop gates, or flip pin allocations on a bus driver
(two very common but non-trivial cases), then it is a
matter of doing the changes in gschem, and say GO.

To work out in practice, this has to work pretty automatically
(not too hard to do) but also rapidly. This *might* need a sort of
diff/patch mechanism in the pcb package, I'm not sure. 

> I distrust back-annotation for anything but the simplest stuff (e.g.
> reference designators). 

That is probably a very good point. A smooth system for forward
annotation is probably even better, because you would do the changes in
a tool meant for such things.

Egil

-- 
Email: [EMAIL PROTECTED]  
Voice/videophone: +47 22523641 Voice: +47 92022780 Fax: +47 22525899
Mail:  Egil Kvaleberg, Husebybakken 14A, 0379 Oslo, Norway
Home:  http://www.kvaleberg.com/

Reply via email to