[Grok-dev] Grok forest sprint report

[Martijn Faassen]

Hi there,

Last week we had a very successful sprint in a lovely converted barn in 
the forest in Germany.

The atmosphere was great, we had great hosts and great company, and 
thanks also to Uli for arranging everything! I personally came to the 
sprint a bit dispirited, but so much cool stuff was going on and 
discussed that I'm now feel quite energized.

Some of the things we did (please add to this when I forget);

* many many bugs in the launchpad bug tracker were closed, reevaluated, 
and updated.

* the long-standing Martian improvements finally landed in Grok itself.

* various strange group authorization issues were explored and tests 
were written. We now are in a better position to eventually rewrite the 
way this works.

* started work on a dedicated static resource publisher as part of 
hurry.resource. This would combine the lessons learned from a while of 
hurry.resource usage, z3c.hashedresource and some experimenting I did 
with search, with some new ideas. hurry.resource would totally take over 
the publishing of static resources, taking that job away from
Zope. Jan-Jaap, is a branch checked in?

* various directives were moved to lower-level packages for greater reuse.

* splitted off grokcore.json from Grok itself. grokcore.rest and 
grokcor.xmlrpc are to follow.

* discussions on the future role of the "grok" package now that it will 
contain less code. grok's __init__.py would contain conditional imports 
from the various Grok packages, with a warning if an import wasn't 
possible. By default, grokproject would install all grokcore.* deps, but 
a developer may remove some.

* discussed a new publisher and decided on a few approaches to take

* got a sketch for the new publisher on bitbucket: 
http://bitbucket.org/faassen/dawnlight

* demonstrations and discussion about many things Dolmen. We also 
visited actual real life dolmen in the woods nearby, about 5000 years old!

* demonstrations and discussions of javascript-driven views and forms.

* advanced testing (including javascript with selenium) discussions

* explored the py.test testrunner for a few projects

* work on zeam.setup, a potential buildout/setuptools replacement

* started to maintain community documentation in Sphinx on bitbucket

* work on running Grok without the ZODB; there's a grokcore.setup branch 
for this now.

* implementation of a json schema implementation that can also do 
content conversion for JSON input, along with some other features: 
http://bitbucket.org/faassen/jsonor. JSON plays a more and more 
important part in web development and this will help us get better 
support for it.

Regards,

Martijn