Re: survey on multiple development versions

[David Kastrup]

"Phil Holmes" <address@hidden> writes:

> It seems to be fixing a problem we don't have.  What would be the
> benefit?

We have a problem reconciling potentially destabilizing work with the
necessity to put out releases, in particular stable releases.  That
definitely _is_ a problem we have.

I don't see Mike's particular proposal as currently tenable, either.

> I would also mention that, for this to work, there would need to be
> parallel documentation streams, too - otherwise users would read the
> docs for A, assume it's in B and rightly complain if it wasn't.  This
> would entail 1) even _more_ confusion about which is the right doc set
> to read and 2) an impossible upload.  Currently my internet access is
> screwed for about 5 hours every fortnight while I upload the new
> version.  I don't see I could lose it for 5 times that long.  It would
> also exceed my monthly quota in a single day.

Well, this is one point we should likely be addressing at some point of
time: we can probably cut down on platforms.  Of course, that alone will
not buy us enough leeway for a "5 times as much" scheme.

If we have branches with personal interests, it must become more
feasible for the respective authors with personal interests to provide
binaries if they consider that a good idea.  Any solution that will only
work via the "Phil, do more" route is not going to scale.

-- 
David Kastrup