Re: gEDA-user: Symbol name convention?

[DJ Delorie <dj@delorie.com>]

> Parts come in multiple packages.  For example the Z8F1602 could be
> in in a 64 or 68 pin package.

The schematic symbol name has nothing to do with the footprint.  The
-1 and -2 suffixes are for cases where there are multiple ways of
symbolizing a part, such as where the bubble goes on an inverter.  So
if someone submits a new inverter symbol, it might be named inverter-5
if there are already 4 such symbols in the database.a

When you go to board layout, you'll need to give each part "device"
and "footprint" attributes, and at that point you'd worry about which
package you have.  For example, I have generic footprints for SOP14,
SOP16, DIP14, DIP16, etc, and I just assign the right footprint to
each part depending on which package I chose.  So if I have a hex
inverter, I use a generic inverter symbol in the schematics, but
specify the 74LS04 device with a TSOP14 footprint.  This methodology
is a little less obvious for microprocessors, where there's usually a
one-to-one correspondence, but you still have variants - chip speed,
installed memory, packaging, etc.

In the case of resistors, for example, the attributes might be:

	   symbol:	resistor-3
	   footprint:	AX025W  (1/4watt axial)
	   device:	47000

I think the "device" is only used for the BOM so you might as well put
the vendor's part number in there, whereas the symbol would be named
for a "generic" chip of that type.

In the case of the Z8 64 vs 68, the only differences are power pins,
so you could get away with a single symbol.  But if you're going to do
all the packages, I suggest using the same letter suffixes they use
(PM, AN, VN, etc) if needed (although Z8Fxx01 might be sufficient, vs
Z8Fxx02 or Z8Fxx03).  The -1 and -2 would be added if there were, say
different ways of representing each chip on a schematic.