[Numpy-discussion] 1.8 release

Nathaniel Smith njs at pobox.com
Mon Jan 14 17:08:50 EST 2013


On Mon, Jan 14, 2013 at 4:45 PM, Frédéric Bastien <nouiz at nouiz.org> wrote:
> I don't volontear for the next release manager, but +1 for shorter
> releases. I heard just good comments from that. Also, I'm not sure it
> would ask more from the release manager. Do someone have an idea? The
> most work I do as a release manager for theano is the
> preparation/tests/release notes and this depend on the amont of new
> stuff. And this seam exponential on the number of new changes in the
> release, not linear (no data, just an impression...). Making smaller
> release make this easier.
>
> But yes, this mean more announces. But this isn't what take the most
> times. Also, doing the release notes more frequently mean it is more
> recent in memory when you check the PR merged, so it make it easier to
> do.

Right, this is my experience too -- that it's actually easier to put
out more releases, because each one is manageable and you get a
routine going. ("Oops, it's March, better find an hour this week to
check the release notes and run the 'release beta1' script.") It
becomes almost boring, which is awesome. Putting out 5 small releases
is much, MUCH easier than putting out one giant 5x bigger release.

On Mon, Jan 14, 2013 at 9:26 PM, Ralf Gommers <ralf.gommers at gmail.com> wrote:
> +1 for faster and time-based releases.
>
> 3 months does sound a little too short to me (5 or 6 would be better), since
> a release cycle typically doesn't fit in one month.

The release cycle for 6-12+ months of changes doesn't typically fit in
one month, but we've never tried for a smaller release, so who knows.
I suppose that theoretically, as scientists, what we ought to do is to
attempt 1-2 releases at as aggressive a pace as we can imagine to see
how it goes, and then we'll have the data to interpolate the correct
speed instead of extrapolating... ;-)

On Mon, Jan 14, 2013 at 12:14 AM, Charles R Harris
<charlesr.harris at gmail.com> wrote:
> I think three months is a bit short. Much will depend on the release manager
> and I not sure what  Andrej's plans are. I'd happily nominate you for that
> role ;)

Careful, or I'll nominate you back! ;-) Seriously, though, Ondrej is
doing a great job, I doubt I'd do as well...

Ondrej: I know you're still doing heroic work getting 1.7 pulled
together, but if you have a moment-- Are you planning to stick around
as release manager after 1.7? And if so, what are your thoughts on
attempting such a short cycle?

-n



More information about the NumPy-Discussion mailing list