Mail Archives: geda-user/2023/03/18/04:53:54
Vladimir:
> "karl AT aspodata DOT se [via geda-user AT delorie DOT com]" <geda-user AT delorie DOT com> writes:
>
> > Why then don't we cooperate on making sch2pcb great ?
> > Do you want to convert it to scheme or still doing things in C ?
> > Tell me your plans with it.
>
> I'm working towards converting the program to Scheme as it is a higher
> level language than C and it allows for better testing and faster
> progress. Thereafter I'm planning to transform two related
> lepton-netlist backends to so named (newly introduced) 'module
> backends', which will give us a chance to replace system calls in the
> program with just using available Scheme functions. The base for it has
> been added recently, I just need some time and motivation to go ahead.
> One discouraging fact for C lovers is I work towards converting most of
> C to Scheme. I, for one, like C. OTOH, I like Scheme much better. In
> this particular case, IMO, the C code is overcomplicated, and any work
> on it takes much more time compared to what I could spend if I did it in
> Scheme.
>
> In two words, since about two or three years ago I had a new Scheme
> prototype version of the program. I realized then, that it is not good
> enough tested and I need much more work on testing it in order to offer
> it to the users. If I do all testing myself, it'll spend much more
> time to come to some result :-)
Why not laeve sch2pcb as is and develop the scheme version instead,
perhaps we could call it lepton-2pcb (lepton implies sch so that is
redundant) or have it as a menu item in the schematic prg. ?
I can do testing and possible some scheme but I'm not fluid in scheme,
just tell me what to do.
From what I have guessed, my goal, the program should do:
. read lepton and pcb configs to find the libs.
Unfortunately pcb has three config files: pcbrc preferences settings,
which one has the final say about libs ?
. read the sch file, find refdes and fp or refdes and source attributes
. do a netlist as usual
. find the footprint file[s]
. find the source file[s]
Unfortunately neither sch and pcb file formats resolve the
issue of files with the same name in different directories.
. put all fp's in the *.new.pcb file as usual
. get the source files pcb sibling, adj. refdes (add the source files
refdes) and save it in the same dir as the orig. sch file
and then possible:
. remove not used fp in the target pcb file
. write out a nice message to the user
or tell pcb that
. there is fp's remove
. there is fp's to add
. there is subpcbs to add
///
An extension to the last part, is that pcb could tell lepton to
. change an attribute of a symbol
. change some connections
///
All the above goes for pcb-rnd also if possible.
Regards,
/Karl Hammar
- Raw text -