[Cosmo-dev] [Sum] Dec 10-16

[Ted Leung]

Here's the summary of the activity on cosmo-dev for the week of   
December 10, 2006.

New threads:
-------

Mikeal Rogers checked in the first version of a script to test Cosmo  
database migration.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002314.html>

Randy Letness reported that he finished a migration framework for  
Derby, and in the process rewrote all the migration code in Java, so  
that there is one codebase.  He also mentioned that the new framework  
could be configured to auto-run on Cosmo startup.
+ People were generally fine with the new Java framework, but nobody  
like the idea of auto migrating on Cosmo startup
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002315.html>

Randy Letness upgraded Cosmo to use Spring 2.0.1 and acegi-security  
1.0.3.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002325.html>

Vinu Gopal posted his status on WebDAV ACL's for Cosmo.
+ Brian Moseley asked why the group UI was deferred, but didn't get  
an answer.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002327.html>

Ted Leung posted the first version of a more flexible Javascript Unit  
Test runner.
+ Matthew Eernisse replied with a code review.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002328.html>

Vinu Gopal had some questions about how to implement groups in CMP.
+ Brian Moseley replied with some suggestions for a more RESTful way  
to implement groups.  Vinu had already implemented a different way.   
It was unclear as to whether the RESTful way would be adopted.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002329.html>

Bobby Rullo and Matthew Eernisse swapped a bunch of messages  
regarding changes to ModelDialog code that resulted in some good  
explanation of Javascript subtleties.  You need to read the whole  
thread, though.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002330.html>

Jeffrey Harris asked about how plans for handling unrecognized  
iCalendar data in the new EIM based world.
+ Several possible ways of doing this were discussed, but this is  
probably a bit into the future.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002339.html>

Ted Leung asked for proposals for 0.6 bugs that should be deferred to  
0.7.   He didn't get any responses.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002355.html>

Brian Moseley reported on details of account and url migration when  
Cosmo 0.6 comes out.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002357.html>

Jared Rhine has tested the new Java based migration tools and they  
seem to work, with the exception of a minor build related bug.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002362.html>

Travis Vachon reported on some backed refactoring that was needed for  
the bookmarkable URL part of the Casual Collaborator workflow.  This   
work is all done and he is moving on to compressing Cosmo's  
Javascript downloads.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002366.html>

Continuing Conversations:
-------

The discussion on getting handles to UI elements in the browser (for  
testing) continued:
The choices seemed to be between randomly generated ids for each  
element vs names generated from the element hierarchy of the DHTML   
Matthew Eernisse summarized the arguments in <http:// 
lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002365.html>, and asked for more input from the community.   No one  
else spoke up.
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002321.html>

Ted Leung replied to Adam Christian's query about Cosmo UI  
Performance Testing.  He pointed Adam at the measurements that the  
Chandler project is making  as an example.
+ There was additional conversation trying to flesh out specifics as  
they related to Cosmo
<http://lists.osafoundation.org/pipermail/cosmo-dev/2006-December/ 
002338.html>