kicad-developers team mailing list archive

["Brian Piccioni" <brian@xxxxxxxxxxxxxxxxxxxxx>]

Well, that's a pity. I was hoping to be able to help but I rather doubt that
would be possible because by the time back-annotation is in place I won't
have the time to learn the code well enough to contribute.

Mind you I am of the school that a good plan today is better than a perfect
plan tomorrow.

If there are other non-critical sub-projects I could work on in order to
learn, I'd be keen to have the opportunity.

Regards

Brian

-----Original Message-----
From: Seth Hillbrand <seth@xxxxxxxxxxxxx> 
Sent: January 8, 2019 12:43 PM
To: Wayne Stambaugh <stambaughw@xxxxxxxxx>
Cc: Brian Piccioni <brian@xxxxxxxxxxxxxxxxxxxxx>;
kicad-developers@xxxxxxxxxxxxxxxxxxx
Subject: Re: [Kicad-developers] [KiCad-developers] Hoping to contribute but
I have some questions

Am 2019-01-08 09:46, schrieb Wayne Stambaugh:
> Seth,
> 
> This seems backwards to me.  Geographical references are generated at 
> the board level so the references changes have to be pushed from the 
> board to the schematic not the other way around.  This requires the 
> following steps:
> 
> 1) Create a netlist from the board editor.
> 2 preferred) Send netlist via KiMail to schematic editor.
> 2 optional) Save netlist to file.
> 3 preferred) Update references in schematic from mailed netlist.
> 3 optional) Load netlist from file and update references in schematic.
> 
> Either way, the code will have to be written to generate a netlist in 
> the board editor and parse the netlist in the schematic editor.  The 
> netlist generated by the board editor should have the same format as 
> the netlist currently generated by the schematic editor.
> 
> The reason I am making this netlist approach a requirement is that 
> this will be necessary for pin and gate swapping, constraint 
> management, etc.
>  It will also allow us to replace the component assignment file (.cmp) 
> for updating footprint changes in the board.  Allowing an interim 
> solution only means the someone will have to go back and merge it into 
> the netlist solution.  I would rather avoid the extra work.

Got it.  Brian, it sounds like you'll want to hang tight until the
backannotation is in place.  At which point, you can presumably call a
single function to change the reference of a component and have that pushed
back to eeschema.

-Seth