octave-maintainers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: graphics issues (was: Re: moving toward a 3.0 release)


From: Sebastien Loisel
Subject: Re: graphics issues (was: Re: moving toward a 3.0 release)
Date: Sun, 1 Oct 2006 07:26:03 +0200

| and label placement logic in .m files. You're the one who kept asking me for
| the least amount of stuff to be done in C++ as possible, and the most
| possible in .m files.

Sure, it's a reasonable goal, but if some code already exists to do
the job and there's a way for us to use it, there's no point in
reinventing it as a .m file.  So does OpenGL have good routines for
autoplacement of tic marks and axis labels given data ranges and font
metrics?

Again, my professional opinion is that this worry about having some library deciding the position of tick marks for you is bogosity. You will spend more time adapting your code to a foreign library and waiting for the plot to come up than you would making the decision where the tick marks go and feeding it to OpenGL. I speak from experience.

Now if you had to code the tick mark logic in C++ it might be different.

Paul also said:

> OpenGL does not have good routines for autoplacement of tics.
>
> I wrote Tcl code which would be easy to translate.
>
> Note that OpenGL does not have support for text either.
> There are a couple of libraries which provide it but I
> haven't used them myself.

So as Paul points out, this is not hard. And as for the text, Paul hasn't needed it but clearly from my screenshots I've implemented it already, so it can't be that hard.

Code reuse is great, which is why we want to use Octave and not NLISP http://nlisp.info/nlisp-faq.html, but reusing someone else's wrapper around fread() is just a waste of time.

Sebastien Loisel


reply via email to

[Prev in Thread] Current Thread [Next in Thread]