Re: gEDA-user: A little puzzled about the purpose of gschem

[Matthew Wilkins <matthew_m_wilkins@xxxxxxxx>]

   I think most people end up building a small library of gschem symbols
   that they use.  If they work with Spice a lot, the symbols will include
   the references to the necessary spice models.  Once they have that work
   of creating models complete, then it's just a matter of arranging
   everything into a circuit, cut-and-paste, etc.  The next time you need
   to simulate a circuit, you'll probably be able to reuse a large number
   of your symbols and you'll be able to get going a lot quicker.
   You should be able to do essentially all of your editing in gschem,
   then just run gnetlist and the simulator.   No editing after the gschem
   stage.
     __________________________________________________________________

   From: Madhusudan Singh <singh.madhusudan@xxxxxxxxx>
   To: gEDA user mailing list <geda-user@xxxxxxxxxxxxxx>
   Sent: Thu, April 22, 2010 1:11:11 PM
   Subject: Re: gEDA-user: A little puzzled about the purpose of gschem
     Thanks to everyone who responded.
     Just so that everyone is clear, I understand and appreciate the
   amount
     of work that has likely gone into geda. I have authored two smaller,
     unrelated LGPL projects myself and would never mock anyone for doing
     this.
     A couple of people have mentioned that I need to enter a library. My
     questions:
     Do these libraries bind the attributes to the symbols (so that I do
   not
     have to do any post-gschem drawing and pre-gnetlist work) ?
     As to the respondent who said that gschem is useful because it
   creates
     the net that gnetlist can use to generate the netlist, I am sorry to
     say that you are missing the bigger picture in the workflow. The way
   my
     initial experience was - I found that I would have to enter portions
   of
     that spice netlist in some scattered boxes in gschem, then use
   gnetlist
     to presumably do some topological analysis and write out the actual
     node numbers in the spice netlist. That is where there is a
   disconnect.
     There is no way that that is a more efficient use of anyone's time
   than
     simply drawing a circuit by hand, assigning node numbers and typing
   it
     all in in one shot.
     Now, if I could use a library that came with properly defined symbols
     (instead of just empty pretty pictures that they are right now), I
     could see the utility of doing this. Without that, its a waste of
   time.
     And to the respondent who said that GUIs are not necessarily faster
     than typing it by hand, I would have to disagree. I am hazarding a
     guess that you have not used Design Architect (and yes, I have timed
     the two approaches in the past - the DA bit was much much faster for
     even a moderately complicated circuit). The difference was - that
     beyond defining input and output nodes, I did not have to do any
     post-drawing pointy-clicky.
     All I now need are good libraries and a proper tutorial that shows
   how
     they can be used in gschem properly (see above). Someone linked a
     source - are there others ? Most vendors provide text spice
   libraries.
     How can they be converted into a form that gschem can understand ?


_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user