Sakai Technical Meeting
March 30 2005
Meeting Notes
Sandra will take ownership of the Sakai@ucdavis.edu listproc and will add the new people. At present, the listproc includes the following:
Kirk Alexander - absent at the Mar 30th meeting.
Alex Alfierri - absent at the Mar 30th meeting
Scott Amerson - present at the Mar 30th meeting.
Thom Amsler - present at the Mar 30th meeting.
Tom Arons - present at the Mar 30th meeting.
Chris Brandt - present at the Mar 30th meeting.
Daniel Cotton - present at the Mar 30th meeting.
Brian Donnelly - present at the Mar 30th meeting.
Scott Fischbein - present at the Mar 30th meeting.
Lawrence Fyfe - absent at the Mar 30th meeting
Liz Gibson - absent at the Mar 30th meeting
Kamal Gill - present at the Mar 30th meeting.
Jon Gorrono - present at the Mar 30th meeting.
Peter Korthage – absent at the Mar 30th meeting
Beau Patrette - absent at the Mar 30th meeting
Alisa Sukha-Akhom - present at the Mar 30th meeting.
Ray Tai - present at the Mar 30th meeting.
Sandra Stewart - present at the Mar 30th meeting.
Harvey Delano - not on the listproc but present at the Mar 30th meeting. Sandra Stewart will add Harvey to the Sakai listproc.
The team discussed project organizational issues. We will probably have separate instances of Sakai running for now. We'll look forward to a convergence in the future.
1. School of Medicine has a Sakai 1.5 Windows installation, with a MySQL database. They're running a Tomcat java server, and Apache web server software. Everything is on one computer for now. They are currently looking at three separate Windows-based servers for their production implementation - web server, application server, and database server. They are currently looking at capacity issues for production hardware. Dan and group are already financed for hardware and are not looking for funding.
For Sakai project, the School of Medicine represents Medical Informatics, Family Nurse Practitioner and Physicians Assistants, and Clinical Research Center for their Fall 2005 deployment.
2. Veterinary Medicine has one test instance running also. They're using Apache and Tomcat also. They're using Windows 2003 Server and MySQL. For application development, programmers work on separate instances of Sakai on their laptops before porting to the Sakai server installation for testing.
Vet Med will purchase production hardware after a review of possible options. Vet Med's hardware is already financed.
Vet Med also has uPortal running and has plans to deploy it with Sakai. They are in discovery for how they will integrate the two.
3. Mediaworks, with Jon Gorrono as lead programmer and Kirk Alexander as the lead manager, also has an instance of Sakai up and running for development and testing. They are running Linux Red Hat on a Dell server with Hsql as the database. Mediaworks is using Confluence for documentation and Subversion for version control. They will use Jira for bug tracking.
Programmers for Mediaworks use separate installations of Sakai on their desktop computers for application programming before porting to the Sakai server installation for testing. However, their current hardware is inadequate for testing.
Mediaworks will need funding for the purchase of development/testing and production hardware.
Middleware, headed by Tom Arons, will need access to all these environments to write authentication and authorization management. Tom Arons will seek access to each of the above mentioned development/test computers for Middleware development of authentication and authorization modules in support of Sakai.
Tom also offered that others could use a Middleware server to house documentation and version control software. The group decided to use Confluence for documentation, Jira for bug tracking, and Subversion for version control. Tom Arons and Sandra Stewart will investigate the acquisition of these two software products and report back to the group. We will extend invitations for everyone to access Confluence and Jira on the Middleware server.
There was some discussion regarding the most appropriate OS for the Sakai installations. Kamal reported reading that Windows was not appropriate for production Sakai. There was no action item on this.
We agreed that if we expect to converge our separate Sakai installation in the future, it would be better if we all used the same database software. There was no action item on this. However, we agreed that we should plan on sharing requirements documents where it makes sense. We'll store these documents in Confluence.
We also agreed that we would all use the base Sakai product for our development. That is, we'll always download and store a pristine version of Sakai (in Jira). Any modifications we need to make to the product we'll make in our separate installations. This means that each time the Core Sakai Team releases a new version of Sakai, we will re-install all our local modifications.
Thomas Amsler reported that Berkeley has a production installation of Sakai. Sandra Stewart will contact the Berkeley programmer, Fred Beshears.
We agreed to have these meeting for project management issues every two weeks. Sandra Stewart will make these arrangements. Programmers will also meet among themselves, particularly once everyone has space in the PSL building.
The remaining of the meeting was spent going over the project plan. Sandra Stewart will update the project plan and send to the members. Programmers are encouraged to send additional tasks that they want added to the project plan.