Re: gEDA: Proposal: Make symbol files nestable

[Stephen <stephen-at-NOSPAMelantec_dot_com>]

This was exactly what I was describing back in December...

> From owner-geda-dev-outgoing@seul.org Mon Dec  3 10:52:11 2001
> From: Stephen <stephen@elantec.com>
> To: geda-dev@seul.org
> Delivered-To: geda-dev-outgoing@seul.org
> Delivered-To: geda-dev@seul.org
> Date: Mon, 3 Dec 2001 10:51:58 -0800 (PST)
> Subject: Re: gEDA: Symbol Conventions
> X-To-Get-Off-This-List: mail majordomo@seul.org, body unsubscribe geda-dev
> 
> You may want to consider using the PSpice way of doing things.
> If you look at the old Microsim PSpice "Schematics" program, (which incidently
> was infinitely better than the Orcad way of doing things), the attribute
> 'AKO' or "a kind of" was used where the symbol graphics which represented
> a simulation symbol was used for many different symbols
> 
> Thus a BC108 was "a kind of NPN" and a 2NXXXX was "a kind of NPN"
> and it would pull in the symbol graphics from the AKO symbol.
> 
> In the same way in your example , HS1234_DIP is "a kind of" HS1234
> and a                             HS1234_SO is "a kind of" HS1234
> 
> that way if a symbol has an AKO attribute then it will pull in the
> symbol graphics from the one specified in the AKO attribute.
> 
> This still means you end up with a lot of symbols in the library,
> but the underlying symbol only has to be created once.
> 
> Just a thought...
> 
> Stephen G. Brickles


> From owner-geda-dev-outgoing@seul.org Mon Feb 25 15:15:55 2002
> From: Magnus Danielson <cfmd@swipnet.se>
> To: geda-dev@seul.org, bakker@thorgal.et.tudelft.nl
> Delivered-To: geda-dev-outgoing@seul.org
> Delivered-To: geda-dev@seul.org
> Date: Tue, 26 Feb 2002 00:16:51 +0100 (CET)
> Subject: Re: gEDA: Proposal: Make symbol files nestable
> Mime-Version: 1.0
> Content-Transfer-Encoding: 7bit
> X-To-Get-Off-This-List: mail majordomo@seul.org, body unsubscribe geda-dev
> 
> From: "J.D. Bakker" <bakker@thorgal.et.tudelft.nl>
> Subject: gEDA: Proposal: Make symbol files nestable (was: Type numbers for 
transistors)
> Date: Mon, 25 Feb 2002 14:37:07 +0100
> 
> > Hi,
> 
> Hi,
> 
> > I've followed the discussions on gschem symbols for a while, and it 
> > seems to me that there is a conceptually simple solution to all the 
> > 'one symbol, multiple packages' issue: nestable symbols.
> > 
> > Instead of the current situation, where we have a 7400.sym, a 74LS00  
> > .sym, a 74HC00.sym etc, all of which effectively have 95% overlap, 
> > wouldn't we be better off with:
> > 
> > NAND2.sym:
> > <drawing and pin definitions for a NAND2>
> > 
> > 7400_generic.sym:
> > C 0 0 1 0 0 NAND2.sym <for lack of a better #include directive>
> > numslots=4
> > slot1=1,2,3
> > slot2=4,5,6
> > slot3=9,10,8
> > slot4=12,13,11
> > <power and ground nets, if desired>
> > 
> > SN74AC00PWR:
> > C 0 0 1 0 0 NAND2.sym
> > device=SN74AC00PWR
> > footprint=TSSOP-14
> > manufacturer=Texas_Instruments
> > url="http://www-s.ti.com/sc/psheets/...
> > 
> > The same goes for transistors, diodes, LEDs and the like. This would  
> > IMO solve a lot of problems by abstracting symbols from devices.
> 
> This is effectively doing what I have been proposing for a long time,
> namely separate the symbols and components. In such a system the
> symbol is only the graphical representation where as the component
> declarations describe the details for a certain component.
> 
> > I know that in the Open Source/Free Software tradition I should have  
> > attached a patch at this point. Sorry, no time yet. It's a poor 
> > excuse, but it's the best one I have.
> > 
> > Thoughts ? Opinions ? Flames ?
> 
> Indeed a quite concrete proposal.
> 
> Cheers,
> Magnus