Hi,
I think it would be good to get a Wiki up and running soon. I like
the look of Confluence - having fine grained
control over access to the Wiki could be beneficial with our interaction
with the JSR expert group.
If nobody objects, can I ask either Robert or Ant (our PMC members) to
open a ticket on Jira INFRA to
create a Confluence Wiki? I look the look of the OpenJPA site -
exporting static HTML from Confluence
looks ok.
Thanks,
Stuart
Steve Poole wrote:
Hi,
Having got the holiday season out of the way I'd like to push now on getting
the Kato project moving. The infrastructure items that need
completing/discussing are
1. Creating a Wiki. I vote for a Confluence flavor of wiki rather than
MoinMoin. Apparently from this page (
http://cwiki.apache.org/CWIKI/#Index-HowdowerequestaCWIKISpace%253F) the
CWIKI guys would like a PMC member to take the admin role. The name of the
space would be "kato" presumably :-)
2. Create components in the Kato Jira instance. We have the following
obvious areas - is this a reasonable split?
- Test Compliance Kit. Items under this component would cover TCK
development and issues arrising about execution validity.
- API specification - items here would cover user story creation,
bugs in the spec and spec development
- Reference Implementation (nuf said)
- Test Harness. The TCK is only one part of the test suite. What we
are developing needs a complex test setup and execution environment - its
big enough to need its own component
3. Create at least two versions in the Kato Jira Instance. We have an
obvious 1.0 delivery version and a need for somewhere to store items that
will happen after that. Not sure if there is any standard form?
4. SVN repository needs basic trunk/branches/tags structure setup
5. For spec development we would like to provide the JSR Expert Group
members with the oppotunity to be able to contribute directly to the
specification (ie the API interface classes ) and to the TCK testsuite.
Some of the EG members cannot contribute to the RI development directly (due
to IP issues with exposure to Sun JVM internals) and so I'd like to explore
how we can resolve this?
6. We need a website for Kato setup - I don't have much exposure to the
alternatives proposed - which is best ? or do we roll our own?. Note that
we need to publish the API spec (in the form of javadoc pages) and the
binaries that we create.
7. The kato website, once created will need to be linked-to in various
places on other ASF pages. We need a list of these places made and
eventually get our links inserted.
8. ASF have a Hudson server and I'd like to use it to do the build and
basic unit testing. We'll need a job on Hudson set up.
9. TCK testing is likely to be expensive in cpu terms. I don't know if
we can lean on ASF hardware for this activity. IBM has machines that are
already building and testing DTFJ and there is no problem with them being
used to build and test Kato for the duration of the JSR development. However
I'd like to be clear on what we need to do to ensure that the Kato project
has a good hardware story for the future.
Thoughts?
Thanks
Steve
