I would like to second Jody's comments here regarding the timing of the code
sprint. Having only attended two of these conferences, I can't speak with
any sort of real authority, but from what I've seen, delegates tend to be
pretty flogged by the end of the week, be that from thinking or from

Having the code sprints scheduled at the *beginning* of the conference seems
to me would offer several benefits:

1) people are at their freshest -- ok maybe the 19 hours in 47C weren't
exactly a sponge bath... but you are essentially feeling "on vacation", you
haven't yet had to wake up and have your act together by 9am *sharp*, you
haven't yet been out drinking till 5am four nights in a row (generally),
jetlag might be working to your advantage (or you could at least find a few
friends to code with you at 4am), and but most of all, you've been excited
about this trip for months and now it's finally happening. game on.

2) time can be used as a last minute sprint to get a release out the door --
or at least an rc. This is always a nice thing to be able to announce to
people *during* the conference, ie "And if you want to try this out...."
"just download the new 5.7 release!" versus "download the latest 5.6 release
and then apply this patch at www.example.org/hacks/great_new_feature.patch"

3) The energy of having the developers all together huddling could make for
super-valuable last-minute additions to presentations or workshops

4) Any and all unfinished business started during the code sprint can be
followed up during coffee breaks and post-prandial brandy sessions in the

5) Last day of conference is really last day of conference, and you can
remorselessly go out and celebrate that last night

Furthermore, I'd say bumping it up to at least two or three days is
*definitely* worth looking into. Especially when you consider that you do
get some oddball/outlier types showing up to these things, it's hard to get
a plan set up, get newbies directed, *and* manage to actually write some
code... that's all pretty hard to do in just one short day -- especially if
there are time limitations like we experienced this year.

I would say that especially in the case of paying steep airfares to attend,
companies/individuals would probably be more than interested in spending a
couple more days huddled together banging on work.

anyways, just some thoughts.

On Wed, Oct 8, 2008 at 1:07 AM, Jody Garnett <[EMAIL PROTECTED]>wrote:

> I would go for 3 days - it worked out well for the GeoTools community.
> Three days is enough time to effect change; but not so much time you get
> bogged down.
> I did find timing of a sprint right after FOSS4G to be a bit of a trouble
> for some (in addition to being tired ) reports this year indicate that
> sprints were hampered by the occasional hangover (no doubt due to wish
> others a good trip home the night before).
> Jody
> Paul Ramsey wrote:
>> Everyone loves a good code sprint... or do they?
>> 2007 brought you the one-day sprint (with the GeoToolsers and uDiggers
>> going for an extended weekend sprint).
>> 2008 brings you another day.
>> 2009 is still thinking about it.
>> How much sprinting would you do? 1, 2, 3, 5, 7 days?
>> I am wondering if the right way to handle the sprints is to turn them
>> from something the conference quietly subsidizes to something that
>> OSGeo pays for directly.  That way the conference organizers don't
>> feel like they are having it taken out of their hide, and it can be as
>> long as people like. Also, it fits directly into the OSGeo mission of
>> promoting the development of the software.
>> Book-keeping-wise it's a left-pocket-to-right-pocket transaction for
>> OSGeo, but from a authority and decision making PoV it removes the
>> issue from the plate of the conference team and puts it into the hands
>> of the software promoting team (whomever they may be).
>> Paul
>> _______________________________________________
>> Discuss mailing list
>> Discuss@lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/discuss
> _______________________________________________
> Discuss mailing list
> Discuss@lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/discuss
Discuss mailing list

Reply via email to