incubation checkup was Re: [Board] Meeting Reminder - this Friday

[Jo Walsh]

dear Frank, all,
On Tue, Mar 06, 2007 at 09:00:43PM -0500, Frank Warmerdam wrote:
> >>to produce a report to the board and membership.
> >- a kind of survey of all the projects of their experience of
> >  incubation
> folks are pretty tapped out, so I'm afraid of generating a whole bunch
> more work with no one to do it, revising things.
> However, I have initiated a mentor report on incubation status for the

Nod, I wasn't envisaging something that would take up a lot of
peoples' time, more something the PSC chair could write down in 5 or
10 minutes and get feedback from their project on or not, as they
wished. 

I am interested in getting more information out of the projects and
the incubator is just the place to go where they are gathered
together. 

> >I am concerned that incubation can be seen as too much of a chore. 
> something then I think we have to set a reasonable target level for
> projects to graduate incubation.

Argh, I didn't mean to lower the bar on what we have, which is
lightweight as is reasonable, probably and well thought out. 
It is just difficult when a project's entrance to the foundation is
instantly all these legal issues and a burden of inspection, that I
want to see that balanced with more positive support. I know expecting
a lot more people to give up their free time isn't sustainable. 

> I wasn't aware of a code audit issue with GeoTools (perhaps you mean

All the code audit issues glue togethrer in my mind into a lump.
 
> We have mentors for each project, but there is a limited pool of mentors
> with a good background in the incubation process.

Perhaps at least a 'backup' or 'second' for a mentor? That doesn't
have to be someone with the same amount of seniroty or experience but
knows the osgeo process and is connected to the codebase
professionally? In some places this sort of thing happens organically
because of the closeness of pieces on the stack. Where it doesnt i
think it should be a role of the foundation to encourage it.
 
> >GeoNetwork is looking for some help with bug tracking and release
> >management, i think. 
> 
> Looking for advice?  Actual do-the-work help?  Setting up infrastructural
> services?  Really, I don't think we have a bag full of developers just
> looking for something to do.  But I'm sure we could at least squeeze some
> advice, and infrastructure help out if desired.

Advice primarily, i think, and perhaps another pair of eyes. Is this
paying for this the sort of thing that its appropriate to work with 
a project to seek project directed sponsorship? You've taken a lot 
of initiative with GDAL to do this; how easy is it for another project
to get logistical help to use OSGeo as a support seeking vehicle?

But where the interesting parts are, are often in the edges between
projects - e.g. GeoNetwork + Jeroen are keen to join edges more with
GeoServer and with gvSIG to form something like an 'integrated suite'.
There's the PostGIS/GRASS/QGis axis as well... 

Cross project but not foundation wide and sometimes the kind of thing
that could be a good target for a sprint  - Jody is proposing some
metadata madness at FOSS4G - but benefit from more structured support
- call it 'strategic development'? ;) 

Is this something the foundation wants to directly help projects work
on, or wants to encourage the development of consultancies to produce?
If so how can we be most helpful with least resources? Or is enough of
the right movement happening organically in the world that we do
enough by sitting here being a repository of trust and social hub?
 
> There are times I have thought we ought to have two levels of code
> audit and rigor.  One high level for libraries or applications that
> are likely to get embedded in other things, and a more modest level for
> top level applications.  But I'm not sure that is reasonable either.

Hm I do see that there is a difference in the quality of the audit
process. For libraries so much more is automatable for testing.
Especially as we are starting to get more graphical apps applying I
would really like to see some interface stability / usability QA being
done as part of incubation... moon on a stick, right? But this is
where the userbase will be put off or convinced, on the surface.
  
I don't mean to stir, am not that au fait with what's happening in
incubation and it all sounds positive. 

cheers,


jo