E.D.G. wrote:
> The problem is that the Perl program I have developed is going to be
> sent as a freeware .exe program to researchers around the world who
> know absolutely nothing about computer programming (not that I know
> that much myself) and who are using Windows.
> Otherwise I would have just had people install a Perl compiler.
> A lot of time could have been saved.
That does indeed sound like a better option.
What's so hard about telling people they need to install Perl?
I can assure you that there are many, many perl programs in
regular use by researchers world-wide. They are even used
by non-programmers :-)
> Because of the im****tance of the effort we will have to go
> originally at least with what we have to work with.
> And that means Windows, a Perl .exe program, and pgnuplot.exe
> plus wgnuplot.exe.
That makes no sense to me. The more im****tant it is, then the
more im****tant to get it right. Basing an "im****tant" program
on a poor design and unsuitable infrastructure is just asking for
problems.
> Additionally, my
> programming
> colleague can create .exe programs using only an older version of
> Basic. I
> am having Perl call his programs when needed. And I am presently
> trying to get him organized with sending data to Gnuplot for plotting
> rather than having Basic attempt that.
>
> That Perl to Gnuplot pipe code I listed appears to work quite well.
> And I have been learning a number of procedures to use with it to get
> the desired results such as maximizing the plot window and then
> telling it to remember that configuration (right click on the top of
> the window, choose "options," and have it save the results as
> wgnuplot.ini).
>
> So, this effort is gradually getting there. But so far I can't see
> how there is any way to get Gnuplot to send information directly back
> to Perl
> through a pipe. That apparently has to be done by storing the
> information in a data file that Perl can then read.
>
> Perhaps when I talk with the Perl people in their Newsgroup they will
> have
> some type of recommendation regarding that limitation. Fortunately
> with my program that is nowhere near as im****tant as Perl sending
> commands directly
> to Gnuplot. I am relieved to have learned how to do that. Having
> Perl store Gnuplot commands in files that it would then read and
> execute worked but had problems.


|