Re: [Gnucap-devel] Gnucap Integrated waveform displayer Details

[Felix Salfelder]

On Sun, Jun 15, 2014 at 12:20:02PM +0530, Abhinav singh wrote:
> If we want the same then why dont we save the plot in temporary file
> and it will change if user do the same type of simulation again
> otherwise remains as such.This seems a better option to me instead of
> creating whole dataset related to each simulation.

temporary files are messy. dumping/loading waves into/from might be
useful, but is not required for plotting. this could be implemented
as seperate commands (some day).

the creation of the dataset is already implemented, what's wrong with
it?

> > -------------------
> > similarly, the plot window should allow some interaction on graphs, like
> > change colors, style or delete them. (read: it should be possible to
> > add these features some day).
> 
> 
> How you wants to do it by using an interface or through command window?

i was thinking of a backend that is accessible from a graphical
interface. particularly, the gtk process should be aware of the data it
has been passed, and the waves should be identifiable in a useful way.
i'm not sure what you mean by command window.

> > one set of waves for each simulation type does not really improve the
> > situation. why not just copy a wave to the gtk process as-needed and
> > attach a meaningful identifier? this is still a hundred times better
> > than brute-force store everything, like other simulators do.
> 
>  I tried it but failed may be i was doing something wrong.What
> actually happen gtk widgets redraw with every expose event so it
> retrace whole drawing procedure and if data get changed anyhow then we
> dont get desired output.

what did you try and how? what did you expect? when does data change and
why?

cheers
felix