Home Original page

PyCon DC 2003 Sprints

Martijn Faassen m.faassen at vet.uu.nl
Mon Feb 17 14:45:44 EST 2003 
Cameron Laird <claird at lairds.com> wrote:
> In article <b2joj4$lou$1 at newshost.accu.uu.nl>,
> Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>                        .
>>A few notes from a fairly experienced sprinter (as far as experience goes).
>>If the amount of participants to the sprint is expected to be large
>>(10+) it is important to do some of this preparation in advance, as 
>>otherwise you'll waste lots of time on this during the sprint itself.
>>So let everybody do a CVS checkout before the sprint, try to assign tasks
>>to tackle before the sprint.

> I'd like to see someone publish a HOWTO on sprinting?
> Is there one already?  If not, would you like to co-
> author one, or at least let me interview you for one?

Sure, drop me a mail. I think the current sprinting experience within
the Python community only goes back a few years; I think the phenomenon
started (inspired by Extreme Programming) in 2001 at Zope
corporation for kicking off Zope 3's development.

I heard a rumor that even the XP people hadn't actually done much along the
line of sprints yet, and that Kent Beck was curious to hear
more about this. Open source and sprints are a rather natural fit.

It's getting more and more popular very quickly now. Besides the quite
frequent Zope3 sprints we now have the PyPy sprint and several 
Zope related sprints (Plone sprint recently looked like a big success).
Quite a number of sprints are coming up (PyCon sprint(s), Zope 3 sprint
in Australia, one in Germany, a Silva sprint); soon we'll have a sprint
going on somewhere every week. :)

It's now spreading out of Python land; I am hearing some talk about an
'Open Source Content Management' sprint being thought about for instance.
While of course hackers have been banding together to hack in the past, I think
this sprint phenomenon is still fairly new and I suspect fairly isolated
to Python so far, at least using the 'sprint' terminology.

Sprints seem to be appealing because:

  * You meet other people doing related stuff; contacts are always fun
    and useful. This is like a conference or a workshop. Hacking with some 
    people for a few days is an even better way than conferences to get
    to know your fellow developers, even though you won't meet as many.
    (and sprints are often organized around a conference as it's a natural
     time to do it; developers are already coming to the conference)

  * It's a good opportunity to learn about the system you're hacking on, 
    both by extending it and from the sprint coach, who likely to be an 
    expert on the system. You can pick up tricks from fellow developers as
    well, as you're likely pair programming.

  * It's a great way to influence the development of the system. You can
    vent your concerns at core developers and influence the core development
    itself.

  * And since you're planning to use or are already using the system you're 
    sprinting on anyway, you're investing in your own future and business as 
    well.

Meanwhile it's also not a bad thing to be involved in organizing a sprint
for much of the same reasons, plus that the organisers get a lot of open
source street cred. :)

You almost start to wonder why nobody thought of this before. But perhaps
they had and they called it differently. Or perhaps we really are exploring
new terroritory here... After all, the modern open source phenomenon 
at its current scale is not that old -- perhaps only now the infrastructure
(involved companies and individual developers) is dense enough to make
it worth it. In Europe the Zope sprints definitely seem to attract many small 
Zope companies. Perhaps in US the dynamics are different as sprinting 
seems less popular there -- less small Zope companies around due to Zope
Corp's gravity well, perhaps, or the aftermath of the dot com crash,
or perhaps the whole US business culture is just different? (then again
Americans do attend the European sprints)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?
More information about the Python-list mailing list