Forum OpenACS Q&A: OCT Meeting Agenda 2005-12-27 11:00pm UTC

1: OCT Meeting Agenda 2005-12-27 11:00pm UTC

Posted by Dave Bauer on 12/27/05 07:34 PM

Items I'd like to address today.

1) Bug#2758 (dueling commits, Don made a change, and it has been reversed. The reversal breaks some dependent code including bug-tracker)

2) Bug#2759 Addition of new table with typos in the definiton into acs-kernel packge 1 day before release. Discuss feature freeze and how to better enforce it and better test release process. 5.2.0 install for oracle is broken.

3) 5.2.0 upgrade fixes (in progress) 5.2.0 upgrade fails on sites installed from 4.6.0 (i.e. openacs.org)

2: Re: OCT Meeting Agenda 2005-12-27 11:00pm UTC (response to 1)

Posted by Caroline Meeks on 12/27/05 10:20 PM

http://openacs.org/forums/message-view?message_id=353402

Should the default for PrivacyControlEnabledP be 0 rather then 1?

My understanding is its a very specific dotLRN microfeature designed to allow Sloan to meet a specific US privacy law. I think it should be turned off by default.

4: Re: OCT Meeting Agenda 2005-12-27 11:00pm UTC (response to 1)

Posted by Ben Koot on 12/27/05 11:25 PM

If this is a Sloan/US law issue, I don't see how this should be allowed to mess up an international toolkit. Sorry, but my sense of humor is a bit low right now. This stupid situation kept me busy a whole day, pissed of my client and made me look a fool.

Cheers
Ben

3: Re: OCT Meeting Agenda 2005-12-27 11:00pm UTC (response to 1)

Posted by Malte Sussdorff on 12/27/05 11:27 PM

2) I think we need to have a "postgresql-core" package where people can dump in code which ought to be in core packages but which we dont commit there as it might break core due to Oracle. Makes it considerably safer to commit early and often. Then we could block commits to core packages and only allow these to be made before alpha and only by core developers who have access to both databases. This will surely keep the number of core commits down :).

Enjoy your meeting, I realize that I'm considerably too tired to have OCT meetings from midnight to 1am, so unless we have an occasional meeting in the afternoon, my OCT membership will be restricted to sleepy attendance (if at all) and comments in the forums.

5: Re: OCT Meeting Agenda 2005-12-27 11:00pm UTC (response to 1)

Posted by Dave Bauer on 12/27/05 11:57 PM

Malte,

We might have to work out something else for time.

Regaring 2) developers need to be aware of feature freeze on the release branch and commit code accordingly. Custom is to ask for help on Oracle, Don always offers to test code on Oracle since he has a local install.