Part of the new direction for me will be building tools, including
quite small tools.  Examples:

While studying Blender, I revised the recursive import script so it
merely created the @auto and @path nodes *without* doing the actual
imports.  The *next* time the .leo file opens Leo will actually do the
imports.  This turns out to be quicker and generally better.

Some planned tools:

1. Revise the @auto import logic so it is less fussy.  For example, it
could use @tabwidth to convert tabs to blanks automatically, and only
complain when doing so would create an invalid Python program.  And it
should just indent underindented comment lines without making a fuss.
More generally, it should avoid warnings and report only errors.

2. Remove gnu copy-left notices from sets of files or nodes.  Purely
for study of course, but getting rid of this cruft would be quite
useful.

3. gnu-indent followed by removing braces in .c files.  Haha.  Again,
I find extra characters just so much sand in the mental wheels.

4. Static program analyzers to produce call trees.  Some such already
exist.  With Blender, I feel kinda like the ca. 1980 days when I was
disassembling code.  Yes, the code is there, but it almost might as
well be assembly language: about the only thing that is useful are the
names of the classes and methods.

5. A debugging (or otherwise instrumented) build of Blender would be a
good study tool.  At present, I am stuck trying to get SCons to
work...

I'm going to look for more opportunities to build tools rather than
slog through morasses of code.

Edward

-- 
You received this message because you are subscribed to the Google Groups 
"leo-editor" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/leo-editor?hl=en.

Reply via email to