John Griessen wrote: >> RS274-X makes no distinction between traces and pads, has no way of >> naming pads, etc. Besides, even if it did, it would be up to each CAD >> vendor to actually do something useful with that capability. > > The minimalist hack I am planning is to find the silk > layer elements that touch each other first, then the intersection of > the max extents of those with pad centers and call that a group. > Then get human input to weed the groups and add the missing.
The IPC recommended footprints for even things like 0603 don't have silk surrounding the footprint. Still, if you can figure out how to get gerbv to do some of this, it would be pretty darn cool. With respect to any gerber output of pcb and any libraries used with pcb, I'd suggest that there is no reason to hack the gerber output or libraries for this since pcb can produce xy files. Now, I will admit that there is room for improvement in the xy file output but that will take some changes to pcb's internals to deal with and not just to the export HID that generates the xy files. On commercial tools which do this, how on earth do they do the refdes association? In other words, suppose I layout a simple circuit. It has two SMA connectors and is a 5th order LC ladder lowpass filter and I use 0603's for all 5 components. Even if you recognize correctly the 5 locations for L's and C's how to you make sure that it actually builds you a lowpass filter and not a bandpass one? And if it does build you a lowpass one, how do you make sure you have the correct response and not one that results from swapping the middle cap with one of the end caps? Or perhaps these tools take a netlist as well and extract connectivity from the gerbers? Just some more stuff to think about. -Dan (who has *always* had to provide some sort of x-y file but also has never seen how it is tweaked by an assembler) _______________________________________________ geda-user mailing list geda-user@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-user