Re: [PATCH 0/8] Make the BFD info manual a bit prettier

[Nick Clifton <nickc@redhat.com>]

Hi Tom,

> While I like FORTH well enough, I suspect it might be nicer to just
> rewrite chew in Python. 

My first thought is - "if it ain't broke, don't fix it" - as in, what
do you gain from rewriting chew ?  Is it just so that it is easier to
maintain in the future, should bugs be encountered, or to make it easier
to add new features, or what ?

Don't get me wrong.  If there are good reasons for rewriting chew then
I am willing to listen, but if it is working as-is then I would prefer
to leave well enough alone.


> Over in gdb, we've standardized on that for all
> our maintainer scripts; the main benefits are that a lot of people know
> it and there are plenty of libraries, etc, to use.  What do you think of
> this?

It would add a new dependency for the binutils.  Currently there are no
python scripts being used to build or test the binutils.  Not that this
is necessarily a bad thing.  But if there are alternatives to using python
that would work just as well (bash scipts ?) then maybe that would be
better.


> Finally, it seems to me that it would be nicer if chew were merely a
> documentation extractor.  Having it also generate source files seems
> unfortunate.  For example, things like 'tags' don't work, because the
> primary source code is actually in some comment somewhere.  It seems
> like this could be changed so that code is just code, comments (and
> maybe snippets of code) are extracted into the manual, and some of the
> generated headers are turned into ordinary headers.  WDYT?

I am ambivalent.  I do not use 'tags' myself, so not having it work on
the generated source files is not a big deal for me.  But I can see that
it would be annoying for people who do use it.  I do worry that rewriting
this code will introduce new bugs into an area of the binutils that is
currently working just fine.

If you are passionate about this idea, then please go ahead.  But if you
have 10 other projects clamouring for your attention, then maybe leave this
one on the back-burner.

Cheers
   Nick