This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: some g77 doc'toring


>Do I check this in after fixing it and doing suitable riting penance?

IMO you should just check in doc fixes ASAP.  Well, except for those
where you're unsure about facts.

That is, it's more important to get the new info in quickly and fix
it up later than to withhold it until it's "perfect".  Even though
that's counter to my own tendencies of late -- which are partly the
product of inconvenient coordination of patches to a source base
done by different people, which egcs' use of CVS mitigates somewhat.

In particular, what worries me a lot about making less-than-perfect
changes to a larger document is that, once they go in, it becomes
much harder to spot the "imperfections" as the stylistic variations
they are.

I've been bitten by this myself -- seeing two or three different
styles in a code or doc base, I happened to pick the wrong one as
what I thought was "correct" and used it in the new work I was doing,
only to find later that it was wrong, but hadn't been aggressively
dealt with.  Worse, I've globally fixed things to the "wrong" way
before discovering I'd used an imperfect example (of someone else's)
as my guide.  (And "wrong" is highly subjective -- e.g. I've munged
the g77 docs vis-a-vis using "space" versus "blank", neither really
being wrong, but only one being consistent with the Fortran standards.
Similarly, neither British nor American English is "wrong", but,
in a given document, it's best to stick with one, to avoid the
occasional misunderstanding.)

But, since we can track patches more easily than I've been used to in
the past, I assume this is going to be less of a problem.

        tq vm, (burley)


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]