Re: multiple issues with building docs (patch included)

[David Kastrup]

Graham Percival <address@hidden> writes:

> On Thu, Nov 03, 2011 at 08:40:46AM +0100, David Kastrup wrote:
>> And as long as nobody commits directly to master, dev/staging does not
>                  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> what if Han-Wen or John pops in to make a quick bugfix (say, to
> the translation handling stuff), hasn't spent 10 hours to read all
> -devel emails from the past month, and pushes directly to master?
> we haven't documented dev/staging anywhere, so we can hardly blame
> them for following that procedure.

Then I need to rebase dev/staging.

> also, what if any doc writer wants to write docs for your latest
> and greatest features, or if programmers want to simplify stuff by
> using your music function syntax stuff, etc?  Having a dev/staging
> that's more than a day away from master is not a good thing.

Never claimed any different.

> and there's no point having everbody work on dev/staging instead of
> master; we'll still get random breakage and so we'll be no better off
> than having everybody working on master.

Huh?  The point is that not that we won't get "random breakage".  The
point is we won't get unnecessary random breakage in master, and that
master is never rewound (thus providing a dependable basis for private
patches and branches), is good material for bisection, only getting
material that has passed the complete round of tests available at a
given point of time.

-- 
David Kastrup