Re: gEDA-user: why separate xgsch2pcb?

[John Doty <jpd@xxxxxxxxx>]

On Feb 9, 2010, at 9:45 AM, Peter Clifton wrote:

> I'm not totally opposed to teaching gschem how to call gnetlist and
> export various netlist formats,

It would be nice to teach gschem to call "make": that's the way to  
put the pieces together.

This should be a simple scripting problem, except that the gschem  
scripting API is undocumented, so one needs to be a world's expert on  
gschem/libgeda internals to write it. If I was one of those world's  
experts, I'd want to give the API documentation high priority, if  
only to get myself out of design loops like this.

> but where it starts to fall apart is
> this..
>
> A single schematic is rarely a sufficient description to generate your
> design netlist. Only very small designs will 1 schematic -> 1  
> board, and
> that means the "File->Export->PCB netlist" option will not produce
> useful output for (probably) the majority of board designs using PCB.
>
> We have no concept which allows us to find (given one open schematic),
> which other schematics must be netlisted to together in a given  
> project.

It seems especially tricky from the point of view of schematic reuse.

> This is where the prerequisite for "better integration" comes -
> introducing the concept of a project.

"Make" is the tool. The flow contains more than just gschem->netlist.  
There are all sorts of data products in a project: BOM, simulations,  
documentation, etc. Often more than one board, and maybe even an  
ASIC. "Make" can help keep track of all this. And it's enormously  
handy to have a Makefile for a complex project that has been on ice  
for a couple of years. No need to remember the build procedures...

John Doty              Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd@xxxxxxxxx




_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user