> (snip dream about universal PCB library)

>> Any other thoughts?

> Yes: dream on. The PIClist could not even agree on specifications for a
> PIC starter board. (And IMHO that is not even the lists's fault).

A standard designed by comitee would be about as good as most other
standards. Gnu style is the key imho. Many people talk few people
implement then the talkers get to try it and talk some more. Eventually
something that works emerges. There is no consensus between these people.
Nor is one seeked. A solution to a problem is what is being seeked.

Keep talking ...

Here are a few chips from me:

1 - The data storage should be plain text so it would be highly portable
and easily debugged, at least as a start.

2 - The parsing of the data should not mandate the use of higher level
tools (like bison/yacc). This means simple structure and strong use of
delimiters.

3 - The standard should be highly extensible (like XML - but XML breaks
#2). This includes versioning support and branching support so everyone
gets to play with his own version if they want to.

4 - The standard should cover more data than present standards, such as
combining SPICE models with geometric descriptions, pad layouts, physical
dimensions and maybe even standard application schematics and abbreviated
datasheet data where applicable.

5 - It should provide for linking in of existing (non-open) standards and
data, such as Eagle library elements (as external files), pdf files, and
whatever one can think up.

I am well aware that all this can be done in XML but XML is a moving
target and it requires a relatively heavy infrastructure for parsing and
editing.

A VHDL-like (which is Pascal-like in a way) approach would look best
(my opinion) imho.

So keep talking ...

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body