Mail Archives: geda-user/2015/03/17/15:11:10
On Mar 17, 2015, at 12:26 PM, Peter Stuge <peter AT stuge DOT se> wrote:
> John Doty wrote:
>> What puts *me* off is the lack of any clear conceptual foundation
>> to stand on. There are no elementary objects to group together to
>> make complex objects. There’s no grouping hierarchy. Instead, there
>> are predefined complex objects
>
> I might have asked before but I'm not sure whether it was answered:
>
> Which elementary objects would you use instead of the current
> predefined ones,
Define actual elementary geometric objects, polygons, circles, etc. with *no* properties other than their geometry.
Then define properties that objects might have: material, network affinity, negativity (it’s a hole), …
Then allow construction of more complex objects from simpler ones.
> and what would be a couple of concrete advantages?
The chaos of certain constructs being allowed only in certain contexts would be relieved. The confusion of the unnecessary differences between a footprint and other layout fragments would disappear. Arbitrary grouping would support hierarchical design. Users could construct composite objects not possible today. Buried vias have been the subject of repeated feature requests, but they should be trivial to construct from elementary objects.
Look at how IC layout works, where the bottom is geometric shapes drawn on masks. A particular coordinated arrangement of such shapes is a transistor. Construct a NAND gate from transistors, that’s a low level cell. Construct an R/S flip-flop from NAND cells for a higher level cell. Keep working your way up until you’ve captured the whole design.
John Doty Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com
- Raw text -