making review/bug filing optional for some docs changes

[Gervase Markham]

David Miller wrote:
> Personally I'm not against having trusted folks making small changes in 
> the docs.  Making wide sweeping changes or changing the content (rather 
> than just the presentation) of the docs are things that should certainly 
> continue to require review though.

But it's the wide, sweeping changes which are hard to review, because 
they tend to be done incrementally and patches tend to be very fragile.

We don't have to have the same policy in all parts of the development 
cycle - after all, the code has an open period and a freeze period. How 
about something like: when the code is not frozen, people can update the 
docs without patch review. If they want to make structural changes, they 
need pre-implementation review of the _concept_ - so core docs people 
don't turn up one day to find the entire repository rearranged. Once the 
code is frozen, all docs patches have review and approval as normal.

The idea is that this flexibility and fluidity means that the docs are 
generally in better shape on an ongoing basis, rather than being fixed 
up at the last minute - so there's much less _need_ for the big changes 
during a freeze period, because the docs at the start of the freeze are 
much closer to the final thing. So the review requirement during the 
freeze is not onerous either.

Gerv