Re: [ng-spice-devel] Answer to MANU and Arno and design doc.


To ng-spice <ng-spice-devel@ieee.ing.uniroma1.it>
From Paolo Nenzi <pnenzi@ieee.ing.uniroma1.it>
Date Mon, 6 Dec 1999 21:07:40 +0100 (CET)
Delivered-To mailing list ng-spice-devel@ieee.ing.uniroma1.it
In-Reply-To <384C40F7.7AFAE1E3@wanadoo.fr >
Mailing-List contact ng-spice-devel-help@ieee.ing.uniroma1.it; run by ezmlm

On Mon, 6 Dec 1999, Manu Rouat wrote:

> Not very important - if we are not satisfied with it's speed, then we can
> always consider rewriting the functions that we need and optimize them.
> The important thing is to get a functionnal simulator quickly.
I agree!


> verilog grammars, which tends to show that lex/bison are more useful with 
> nicely formatted inputs.
I am wandering what solutions have develped other implementors.

> I am a personnal friend of Daniel Veillard, member of the W3C and author of 
> libxml. He has often said that an xml netlist would be very useful to us.
XML is a standard for everithing (well, seems to be). I think that he can
give us some good advices on how to write a DTD for netlists. 

> I will try to see if he can write something about the advantages of xml
> for us. However, I do know the xml is essentially an interchange format, not
> really meant to be hacked by hand. The advantage is that xml is fast and
> easy to parse (and libxml can be used for that).

Yes, it's a markup language that encapsulates the parsing rules of its
content.

> OK, LyX and SGML output then.
Then we will make a doc directory with sgml files and LyX files.

> This can all be found in GSL. The online HTML docs for GSL seem to suggest
> that code written using it will be quite easy to read.
Yes, I have downloaded the pdf docs and read them. There a re a plenty of
functions. I am wandering how we can provide partial derivatives of all
quantities, any idea?

> > Modified Nodal Analysis: Yes. Are there other interesting methods ?
> 
> yes : have a look at:
> 
> http://www.aplac.hut.fi/publications/ecctd-91-2/main.html
Wow, gyrators, at the end I see why I have stuided them in my Circuit
Theory I!!!!! It has been a misterious component for me ;-)))

> http://www.iue.tuwien.ac.at:8000/diss/grasser/diss/diss.html
> that one is VERY good!
Yep. Fantastic. Manu .if you find similar docs, send me the URLs!!!

> This is what most simulators do. They have a default algorithm, but it can
> be changed by the user (under ELDO, you can use : '.options be' for 
> Backward-Euler for example)

Aren't we talking of reinventing the wheel! ;-)
The new links will be on the web tomorrow.
Paolo


Partial thread listing: