Parrot Developer Summit, Dec 05 2010: Summary of Action Items

[James E Keenan]

The Parrot Developer Summit met for over three hours beginning at 2300 
UTC.  You can read the log starting here: 
http://irclog.perlgeek.de/parrotsketch/2010-12-05.  The summit 
particularly benefited from contributions by people who were major 
Parrot developers in past years and by people quite new to Parrot. 
Here, I'm not attempting to summarize the ebb and flow of discussion; 
I'm focusing on action points.

At the meeting I agreed to lead the effort in scheduling and organizing 
our summits for calendar year 2011.  We agreed to hold summits quarterly 
on the second weekend following our quarterly supported releases.  That 
means the following weekends in 2011, which you should mark on your 
calendars:

Sat-Sun, January 29-30, 2011
Sat-Sun, April 30-May 1, 2011
Sat-Sun, July 30-31, 2011
Sat-Sun, October 29-30, 2011

The exact day and time at which a particular quarterly summit will take 
place will be decided later.  In order to accommodate the fact that our 
developers live all around the world, we will try to have start times 
that spread the joy and the pain around.  particle and I will work on 
this process.

That schedule means that our next summit is 8 weeks away.  At this 
summit, Parrot developers committed themselves to working on certain 
areas.  Where these areas entailed design and planning issues, we asked 
them to write up material *before* the next summit so that participants 
have a sound basis for discussion and that we have ample time for new 
business and more open-ended dicussion.

In quite a few cases, we encouraged people to start working out ideas on 
their blogs and then publish more polished version on parrot-dev in the 
2 weeks before the summit.  Here is a list of such issues.  (If I have 
misunderstood or misrepresented any of the issues, please post.)

1. Google Summer of Code 2011:  Next round is likely to begin process in 
Feb 2011.  By end of Jan 2011, Parrot should have a draft plan for our 
participation.  That plan should identify some possible subject areas, 
mentors, etc.  We recognize that some GSOC projects will entail basic, 
long-term research and may not be immediately usable by Parrot. 
However, other projects can be structured such that their results are 
mergeable into Parrot at the end of GSOC 2011 (perhaps in a "merge 
party").  dukeleto leads this effort.

2. Architecture and Design:  At this summit, deficiencies in Parrot's 
meta-object protocol were identified as a major obstacle to use of 
Parrot as a platform for high-level languages.  cotto will examine these 
deficiencies in the context of getting Lorito into an iterable state.

3. Going into this summit, the API team (whiteknight and bluescreen) 
made considerable progress on a new embedding API.  The blog posts that 
whiteknight wrote, along with his recent posting on parrot-dev in 
preparation for the summit, should serve as examples for the sort of 
work we want to see in advance of the end-of-January summit.  By 
mid-January, the API team will prepare an additional report on the 
status of that work and whether it's ready for merging in 3.0. 
Documentation for these features should be thorough and usable by people 
looking at Parrot for the first time.

4. Parrot on Android:  dukeleto and lucian will begin to prepare an 
action plan, probably starting with blog posts.

5. Parrot Roadmap:  The usefulness of the roadmaps formulated at our 
summits in 2009 was severely questioned.  A simplification was 
suggested:  By mid-January (better yet earlier), cotto (as Architect) 
and whiteknight (as Product Manager) should identify the features they 
want Parrot to have in each of our 4 supported releases in 2011 
(Jan/Apr/Jul/Oct).

6. Branch Merging:  chromatic volunteered to make a document listing 
what we need to consider when we merge a feature branch.

Thank you very much.

kid51