[Incubator] thinking about incubation sprint

[Jody Garnett]

There should be examples of existing projects and our graduation checklist.

The code review is pretty minimal, checking headers and producing a set of
known problems. If it helps the intended audience is the OSGeo board, and
to a lesser extent potential users (who can use the list of know issues
when performing audit/risk assessment).

We had a lot of frustration at the start of OSGeo with automated header
checking. Ended up asking projects to manually review each file (best done
as a team over a couple days). For GeoServer we did both grep and manual
(see GeoServer Provenance Review
<https://github.com/geoserver/geoserver/wiki/GeoServer%20Provenance%20Review>
).

Jody Garnett

On Tue, Sep 30, 2014 at 2:14 PM, Even Rouault <even.rouault at spatialys.com>
wrote:

> Le mardi 30 septembre 2014 18:08:30, Jody Garnett a écrit :
> > I was trying to think what an incubation sprint would look like .. am I
> > correct in assuming that "code review" is the sticking point?
>
> Jody,
>
> What kind of "code review" ? Checking that all source files have a license
> text
> in their header ? Or something more involved ?
> Regarding licences, I'm wondering if there are not tools that could do an
> automated summary. There are definitely proprietary ones, but I couldn't
> find
> free software ones right away with my very quick search.
>
> Even
>
> >
> > In any case two ideas:
> >
> > 1) A general cross project sprint where we gather up developers from
> > several projects, along with their mentors and try and power through
> > checking the code. It is good value as projects will encounter  similar
> > questions and have a chance to learn from each other etc... This is great
> > for community building but represents large sponsorship cost as we are
> > assured of getting developers from different corners of the globe.
> >
> > 2) The other way to play is to focus on a project at a time, and see if
> we
> > can sponsor the project mentor to "crash" an existing code sprint. This
> may
> > be better for projects with a lot of developers in the same part of the
> > world?
> >
> > Finally I a not sure if remote (via Skype) sprints are going to cut it.
> It
> > is too easy to get dragged away by your "day job" and not focus on team
> > members who are waiting. It slows down the whole thing and lacks focus.
> >
> > Any ideas/thoughts? I would be keen to hear from graduated projects on
> "how
> > they did it" - perhaps we can learn what worked and structure a "sprint"
> > around that activity.
> > --
> > Jody Garnett
>
> --
> Spatialys - Geospatial professional services
> http://www.spatialys.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/incubator/attachments/20140930/0226e25a/attachment.html>