Maurits Rijk <[EMAIL PROTECTED]> writes:
> I just had a look at the code of some of the plug-ins and I noticed that
> there is often lot's of room for improvement:
>
> 1) some constructions are plain clumsy, for example somewhere I saw:
>
> for (k = 0; k < bytes; k++) destline++;
>
> instead of simply:
>
> destline += bytes;
Easily fixable. But why did the code get into that state to begin with?
> 2) a lot of functionality has been added since the last update of some
> of the plug-ins. Result: duplicate functionality
This basically boils down to "the GIMP's user interface is a mess". I
don't know how much fixing and polishing you want to do during the
freeze. If you remove functionality from some plug-ins, you may have
to modify scripts and trigger some other chain reactions that the
freeze may not be happy with.
> 3) sometimes it's just lack of C knowledge:
>
> if (p)
> free(p);
>
> can be simply replaced by just:
>
> free(p);
If p is a null pointer then "free (p)" may (and should!) crash. You
are incorrect here.
> 4) some plugins have obvious memory leaks.
>
> So my question: is it worth the effort to carefully go through all the
> code (not only plug-ins) and clean things up?
Yes. You can use tools like memprof to do this.
> 1) source code becomes smaller. Not a big deal with Gb harddisks, but
> nice for future maintenance.
>
> 2) object code becomes smaller. My initial estimate that for example
> most plug-ins can be easily reduced with 10 a 20 % without much effort.
>
> 3) during the process we might find further bugs.
(3) is the important one. The first priority is to achieve
correctness. Performance in terms of speed or memory is secondary.
> Disadvantages:
>
> 1) this will cost time/effort (I am willing to make my contribution).
Not an issue.
> 2) we might break things that work
This can be fixed by, you know, testing things after you change them :-)
> 3) we don't spend time on other fun things liking adding functionality.
The GIMP does not need more functionality at this point; it needs to
be checked for correctness.
Federico
