> =46WIW, I absolutely hate the m4 definition scheme. Everytime I do a large= > =20 > layout (Which I'm doing now...), m4 turns into a ditch pig that I have > to edit/customize for hours. > > m4 is the only true gripe I have with gschem. Sure, I'd like to see some=20 > features added in the future, but as far as using the program as it is > right now, m4 frustrates me.
I agree that M4 is nasty pig. However, gschem doesn't use M4 at all. I assume you are ranting about PCB, right? As far as PCB goes, some folks -- including myself -- don't use M4 footprints at all. We use only newlib footprints & hand edit them using emacs (or whatever) because that gives us total control over every aspect of the footprint. I can then inspect each footprint to verify that the pin holes are the right diameter, the solder mask is properly cleared from the pads, the silkscreen lines meet the proper design rules, etc. I tend to keep all my newlib footprints in a dedicated directory which I place in the footprint search path. Yes, the drawback is that you need to actually create all the footprints yourself, but that's what most layout designers do anyway. Never blindly trust somebody else's footprints -- verify everything, else you will end up with a footprint whose hole is too small for the component (for example). One of the regular members of this list has talked about selling a CD full of pre-built and verified newlib footprints for PCB. That's a good idea, IMO. (Others may disagree, but they don't have to buy the CD.) Once this happens, you'll have an alternative to the built-in libs. (Of course, you'll still want to check them. . . . . ) Stuart > > Regards > > Marvin
