Re: gEDA-user: Tools for timing diagrams for digital signals

[Andrew Bardsley <bardslea@cs.man.ac.uk>]

On Sat, 15 Feb 2003 steve@spiketech.com wrote:

> Andrew,
> 
> I don't expect you to go off and write it - I'm just
> trying to define what the other tool would be used for
> as a point of discussion of the differences.

Yeah sure, I'm just throwing around my own GTKWave ideas
too.  I can think of good uses for some of this stuff for
following signalling protocols.
 
> Again - the free-hand creation of the timing diagram itself
> is a more interesting tool because it has a more general
> application. I already understand that you wouldn't
> personally use such a tool.

Someone said (with a sigh) `Visio' as the tool to use. Perhaps
instantiating up-transition/down-transition/etc. objects on a
diagramming tool (on an automatically generated grid template)
and then post-processing that diagram to draw the correct traces
over the placed objects (and remove the objects/move them to
a different layer).  'cause the problem with drawing timing
diagrams always seems to be the way that most diagramming tools
can't handle the concept of a kink in a wire (a transition)
being a real thing you might want to move around.  Or maybe
I'm using the wrong diagramming tools.
BTW. I probably would use such a tool as timing diagrams can
be handy for highlighting examples of beahviour (even if I
don't use them for spec.s)  I'd love to have
an ASCII art timing diagram generator, actually.
It'd be very handy for email!
 
> Annotation of diagrams in GTKWave would have some
> utility for documentation after the fact - while I
> would imagine you could use the other tool before the
> fact.

Yes, that's the kind of role I was thinking of.
Actually, my thoughts had progressed from thinking about
how handy annotation would be for generating timing diagrams
to how useful the same annotation would be for me for understanding
the relationships between signals in my circuits.

> The "TimingDesigner" tool can actually be
> used to cut test benches that generate the waveforms
> that are drawn as one example of tieing to useful
> testing needs.

That is an interesting idea!

- Andrew
    __________________
___/Dr Andrew Bardsley\_________________________________________
University of Manchester Dept. of Computer Science  Amulet Group
Research Associate bardsleyATcs.man.ac.uk  Tel: +44 161 275 6844
Snail: Room IT302, Man. Uni., Oxford Rd, Manchester, M13 9PL, UK