Attendees: Julia McCarthy, Chris Middlebrandt, Julie Holloway, Charlie Halloran, Mark McCraw, Robert DeMason, Merri Jensen, Josh Hester, Jason Losh, Mark Weitzel, Michael Scharf
Agenda
Review minutes from last week
Update on tooling use case entry
Enhancement requests entered
P2 Deep Dive Meeting Summary
Jason, Mark, Charlie to summarize P2 runtime meeting
Identify SDD deliverables for COSMOS 1.0
Tools
Runtime
Repository
Tentative: continue use case review
Focus on deployment end user use cases
Follow-up on action items
Charlie/Chris, continue working w/ IBM legal on committing BTG code
Josh, Mark M., Chris to start entering formal tooling use cases
Charlie, Mark M., Jason, to discuss P2 and setup deep dive meeting on P2
Charlie, Mark W., Jason, to discuss code commit process
Minutes
Tooling use cases and enhancement requests from IBM have been entered, SAS needs to reformat their tooling use cases and enter their enhancement requests still.
Target due date: today or tomorrow (2/19, 2/20)
Charlie still working through IBM legal to get approvals for end user use cases.
There are several terminology conflicts (for example, IU and artifact are separate in P2, but not in SDD).
Their IU is an xml file but it is very simplistic, there is no schema defined.
It will most likely require a significant amount of work to make P2 process an SDD.
Mark W. asks if we have created a decision matrix for which technology to use for the Runtime and for the CL1 vs CL2 non-functional requirements.
TPM folks at IBM didn't like Java and Mark W. said that part of him felt that we go w/ P2.
Suggested that we start with a wiki entry and Charlie can circulate back to IBM.
If we base on P2, we should clearly state that so that the P2 team will know. Then we would do the coding work and attach it as a patch/ER, and they would then accept it as a contribution -or- we could put the code in COSMOS as an extension.
What happens if our patches/ER's outpace the actual P2 framework?
We want to ensure the success of P2, they won't want to take extensions if it risks their success, so we would keep patches to the framework to a minimum. We would just have to keep communication with the P2 team open and negiotiate with them (so everything we need is still on the table).
Mark M., Jeff, & Jason to publish their P2 deep-dive findings onto wiki.
Michael asks if we have looked into Maynstall or Eclipse Packaging Project.
Julia is looking into the interchange format (which is like the old media descriptor from IUDD days)
SDD Programmatic Interface (SPI) & validator should be tooling contributions
Change Analyzer
Julia to send design docs to Charlie who will determine if the legal agreements with SAS cover this.
Basically, any code over 200 lines has to go through special IBM legal process. Charlie is lumping the BTG, SPI, and validator as one code contribution, and the Change Analyzer as a separate contribution.
We skipped the deployment use cases, Charlie still working with IBM legal, so we'll keep this as a tentative agenda item for next week.
Committer vote has already started for Jason, Julie is still pending.
As of 2/22/08, Jason has been approved for committer status.
Action Items
Charlie/Chris/Julie, continue working w/ IBM legal on committing BTG code, Charlie to send status.
Josh & Mark M. to continue work entering formal tooling use cases and Enhancement Requests.
Mark M., Jeff, & Jason to create wiki page for P2 findings.
Create Decision Matrix wiki page for implementation technology and for CL1 vs CL2 non-functional requirements. (Owner?)
Look into additional projects that Michael mentioned, links are provided in the Minutes section above.
Julia to send Charlie Change Analyzer design doc, Charlie to engage IBM legal about contributing code.
Mark W. to answer outstanding questions on generated code for BTG and proposed repo structure and package naming conventions.
Mark W. to move forward with vote for Julie and Jason as code committers with COSMOS.
