X-Authentication-Warning:	delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient:	geda-user AT delorie DOT com
Mime-Version:	1.0 (Apple Message framework v1085)
Subject:	Re: [geda-user] pcb fundamentals
From:	John Doty <jpd AT noqsi DOT com>
In-Reply-To:	<1355593978.19071.2.camel@localhost>
Date:	Sun, 16 Dec 2012 16:06:21 -0700
Message-Id:	<E65E22D7-6585-4455-91B5-166123327198@noqsi.com>
References:	<172CCAAB-0423-43EF-8A04-5A9961F1D5B9 AT noqsi DOT com> <201212140122 DOT qBE1MoKM019255 AT envy DOT delorie DOT com> <5AA18F19-2EA9-4E7D-9378-F768D8E1E5DD AT jump-ing DOT de> <alpine DOT DEB DOT 2 DOT 00 DOT 1212140501300 DOT 26605 AT igor2priv> <50CB5D82 DOT 8060507 AT jump-ing DOT de> <E5DE4486-B23F-4945-943A-F46C6E308B90 AT noqsi DOT com> <201212141820 DOT qBEIKQDN005665 AT envy DOT delorie DOT com> <43147C45-75A5-4393-AB07-AFBFD7BD09BA AT noqsi DOT com> <CAC4O8c_YzeqQ84huGrZPL+=aK5C89phD3VgRHko40N8hWCzw7A AT mail DOT gmail DOT com> <95896B46-44E5-40BA-AB8F-2A81E6073CAD AT noqsi DOT com> <20121215010326 DOT 26307 DOT qmail AT stuge DOT se> <57596F89-BBC2-4BEE-8401-AD036143A260 AT noqsi DOT com> <1355593978 DOT 19071 DOT 2 DOT camel AT localhost>
To:	geda-user AT delorie DOT com
X-Mailer:	Apple Mail (2.1085)
X-MIME-Autoconverted:	from quoted-printable to 8bit by delorie.com id qBGN6RWM013111
Reply-To:	geda-user AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	geda-user AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Dec 15, 2012, at 10:52 AM, Peter Clifton wrote:

> On Sat, 2012-12-15 at 10:17 -0700, John Doty wrote:
> 
>> Consider blind vias, buried vias, and buried components. These are not
>> any more complex geometrically than other things pcb can draw, but
>> because they are not specifically implemented in the code, and pcb
>> cannot generally construct aggregates from simple primitive objects,
>> they cannot be drawn.
> 
> It sounds like your ideal PCB package is in fact a generic 3D modeller
> like Solidworks, Solidedge, Spaceclaim or something of that ilk.
> 
> You can model anything with that (including material properties). I
> won't expect you can make everything you model, nor that it will help
> you match up the netlist - but whatever... it is flexible.

We understand engineering in physical terms. We understand physics in mathematical terms (especially geometry). Generic 3D modeling packages get the geometric part right, but they lack the physics and engineering abstractions to relate to a schematic. Classical printed circuit design is mostly planar, so the geometry can be somewhat simplified (but not too much). Support for construction of circuits via 3D printing probably needs something completely new.

Unfortunately, while pcb has some of the high level engineering abstractions needed, they appear to lack consistent physical and mathematical foundations. It's all special cases with no clear fundamentals.

Again, it reminds me of the difference between the road net in downtown Boston versus Manhattan. The Boston road net started as pcb-like special cases (need a road from the warf to the tavern, etc.). It has been repeatedly "patched" to accommodate new "use cases". The result is difficult to navigate. Contrast that with Manhattan, which is mostly regular grid. Even though there are some irregular blockages, it is relatively easy to find a route from any point to any other point (and Manhattan is *much* bigger than downtown Boston). There are also a few "short cuts". Fundamental regularity does not need perfect implementation to be navigable.

John Doty              Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com

- Raw text -