RE: [ng-spice-devel] Re: [ng-spice-frontends] how many?


To "'ng-spice-devel@ieee.ing.uniroma1.it'" <ng-spice-devel@ieee.ing.uniroma1.it>
From Jon Engelbert <jon@beigebag.com>
Date Mon, 20 Nov 2000 09:13:26 -0500
Delivered-To mailing list ng-spice-devel@ieee.ing.uniroma1.it
Encoding 101 TEXT
Mailing-List contact ng-spice-devel-help@ieee.ing.uniroma1.it; run by ezmlm
Organization Beige Bag Software
Reply-To ng-spice-devel@ieee.ing.uniroma1.it

Arno,
We haven't introduced ourselves yet.

Some history on me... I started Beige Bag software back in 1990.  I was a 
grad. student/ Teaching assistant at Univ. of Michigan the year before, and 
the EE department had just started using Mentor Graphics software for EE's. 
 I wanted to develop something easier to use for PCs and Macs... that was 
the original business strategy, and it hasn't changed a hell of a lot since 
then, except that now I'm interested in developing tools for professionals 
rather than students.  I'm pretty familiar with the Spice engine after 10 
years, but I'm not an expert at every aspect of it.  I'm also familiar with 
the xspice mixed mode code.  It had plenty of bugs, especially when the 
transient needed to backtrack and there were digital events on the queue. 
 I spent some late nights reworking the xspice engine.  Also, I've written 
my own digital simulator, which is now part of B2 Spice A/D's digital mode. 
 I've been working with Microsoft MFC classes for all of my user-interface 
code... I'm not a UNIX guy.
To find out more about me &  my company, you can visit www.beigebag.com.

I don't know if I should be sending this to the whole maling group or not. 
 I'm not experienced with this group yet.  I just found out about it last 
week.  If you think that I should send it to all, please either forward it 
to the group, or let me know that I should do it.

Amo, what's your background?

I'm hoping that I can become an active part of the development team 
eventually, but right now I'm swamped.  I like the idea of developing a 
better operating point convergence algorithm, and I like the idea of 
developing front-end capabilities for optimization.

Allowing remote spice simulation is not hard to do with sockets.  I.e., use 
sockets to send data from the simulation engine and then receive the 
results in the front end.  XSpice has this capability, but I don't know if 
it's implemented robustly.  I've played around with Microsoft MFC's sockets 
class, and remote simulation works with this technology (also with 
Microsoft OLE).

** Also, I tried downloading the source code on Friday by clicking on the 
link, and it didn't work. **  Can you help me with this problem?

-Jon Engelbert



-----Original Message-----
From:   Arno W. Peters [SMTP:a.w.peters@ieee.org]
Sent:   Sunday, November 19, 2000 4:49 PM
To:     ng-spice-frontends@ieee.ing.uniroma1.it
Cc:     ng-spice-devel
Subject:        [ng-spice-devel] Re: [ng-spice-frontends] how many?

[cross-posted to ng-spice-devel]

On Fri, Nov 17, 2000 at 10:47:29PM +0100, Riccardo Russo wrote:
>
> I'd like to know how many people is reading this mailing list, and
> if there already are projects or ideas for the ngspice frontend.

I'd say most of the action is in ng-spice-devel.

I am playing with some idea's concerning a Spice frontend.  See below
for further discussion.

> Are we going to modify some existing program (like nutmeg) or write
> a new one from scratch (much more fun!)?

I believe the best course of action is to interface an existing
scripting language to Spice.  It gives you a solid base to build upon,
instead of reinventing the wheel (aka, another scripting language)
once more.

Unfortunately, the license to the Berkeley derived Spice may not allow
such an action... licensing incompatibilities rear their ugly head
again.  Interfacing ACS with one of the many GPL scripting languages
is much easier license-wise.

What kinds of tasks could a frontend do?
* optimize a circuit for some criterium
* allow interactive exploration of circuit parameters
* interactively compare design alternatives
* allow implementation of a robust test environment
* allow remote simulation (Spice server)
* ...

Requirements for an ng-spice-rework/ACS frontend:

* programmable: variables, control loops, etc
* interactive
* command history
* simulation available as function calls
* simulation results in variables
* plotting facilities
* interfaces with libraries (networking, graphic toolkits, etc.)
* ...


Let me know what you think,
--
Arno
 << File: ATT00003.att >> 


Partial thread listing: