Mail Archives: geda-user/2014/03/04/15:29:46
karl AT aspodata DOT se wrote:
> Think of, just one symbol.
Creating a universal data model for parts is indeed key for free and
open source EDA software to have any kind of significant impact in
the world.
But it is no small task!
It is also not a task that neccessarily has anything to do with
existing EDA suites, beyond that the data model must of course not
repeat any mistakes that have already been made in existing programs.
The data model exists between EDA suites, fabs and part vendors.
(User requirements flow from EDA suit and/or fab.)
The model must encompass every real-world characteristic of parts
which EDA suites and fabs care about today, and ideally also be
future-proof enough that the same model can be extended, as EDA
suites and fabs learn to make use of even more part characteristics
in the future.
The project as I see it would have as deliverables a specification
document and a portable C library implementing the specification.
I'd suggest a permissive license such as MIT for the implementation,
so that the door is open for closed source EDA suites and fab tools
to use the library later on, thus giving the project a chance to
eliminate the ridiculous exclusivity of parts within single EDA suite
vendors, and allow for somehow sensible integration across the full
design and production cycle.
Anyone else interested in this?
//Peter
- Raw text -