[EMAIL PROTECTED] wrote:
On Mon, 13 Aug 2007, Abdelrazak Younes wrote:
How about copying stuff from those and pasting it as introductory
comments in a suitable source file?
An even lazier approach would be to let those interested to just look
at the code ;-).
Lazier for you, but more work for many others;-)
No, as I said, the incremental svn diff are already available. Reviewing
such a big patch is masochistic.
In addition, there's an
ulterior motive to my approach: Some documentation of the intended
structure/design ends up in the source code rather than hidden in the
commit messages or the developers' list.
It is already more or less. More doxygen comments are always welcome of
course.
Documentation in the code is
_very_ helpful to developers trying to understand what it does (and what
it is intended to do -- these doen't necessarily coincide..).
I am always careful to document my code. There's a site somewhere that
do statistic on the svn commit. You will see there that my code/comment
ratio is one of the highest.
I repeat again: the changes are not that complicated. Most of the
patch is about:
- bv->buffer() which becomes bv.buffer()
- all test on nullity of bv->buffer() are obviously deleted.
These changes even I could see, but the extent hides non-trivial changes
such as those you describe below. I could have reviewed bv->buffer() to
bv.buffer(), but for other changes I don't know enough about the code.
Me reviewing only parts of a patch is a waste of time - hence separate
patches can help. A small patch requires less effort and thus increases
the likelyhood someone reiews it. A small scope of a patch (or a simple
patch) increase the number of developers able to review it.
Again, the incremental patches are already available in my branch.
Abdel.
