[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gnucap-devel] output plugins
From: |
Felix Salfelder |
Subject: |
Re: [Gnucap-devel] output plugins |
Date: |
Sun, 1 Mar 2015 20:01:16 +0100 |
User-agent: |
Mutt/1.5.21 (2010-09-15) |
Hi Al.
On Sun, Mar 01, 2015 at 10:33:27AM -0500, al davis wrote:
> What I had in mind is that the existing four (alarmlist,
> printlist, plotlist, storelist) are instances of outputs. Each
> output adds a command to feed it (now alarm, print, plot,
> store).
makes sense.
> After they become plugins, we can add more. Like the existing
> ones, they can be used simultaneously. More would include Spice
> rawfiles, Qucs datasets, links to plotting programs and post-
> analysis tools like R, octave, etc.
yes. i was thinking about one "head" and one "outdata" implementation
per command (print, plot etc.)...
> s__out.cc needs big changes. Instead of explicitly calling the
> ones it knows, it should feed to all that are requested.
... then s_*.cc would just call SIM::head, SIM::outdata, which somehow
passes data to all of them. easy. but then:
void FOURIER::store_results(double X)
{
TRANSIENT::store_results(X);
[..] printlist() [..]
}
apparently, FOURIER uses the printlist. obviously, as it does two
computations in a row. i think the output plugins are not supposed to
know. and FOURIER shouldn't have to know about which plugin holds a
"printlist".
there might be several ways around this, do you see a painless or
straightforward one?
cheers
felix