IDE TF Minutes 2006-12-14
A kick off call was held Dec 14, 2006 @ 9:00 pacific to discuss objectives and scope of work for the TF’s recommendation to the Interop Committee
- Jon Ferraiolo <jferrai(at)us.ibm.com>
- Kevin Hakman <khakman(at)tibco.com>
- Phil Berkland <berkland(at)us.ibm.com> representing Eclipse ATF project
- Chris Schalk <chris.schalk(at)oracle.com>
- Ric Smith <richard.allen.smith(at)oracle.com>
(Ric will be taking over for Chris as Oracle’s representative)
- Ted Thibodeau <tthibodeau(at)openlinksw.com>
- Greg Murray <greg.murray(at)sun.com>
- John Crupi <john.crupi(at)jackbe.com>
- Andrei Dragomir <adragomi(at)adobe.com>
- Wayne Beaton <wayne(at)eclipse.org>
Notes in summary, not verbatim form…
KH: I’ve been speaking off-line with Greg Murray, John Crupi and Mike Milinkovich on the topic. Mike’s stated in his experience that standard interfaces to IDEs have been a sticky issue in the past and that it’s a very fragmented space.
Therefore I’d say it’s probable that the best we can do as an organization is assert a standard way for components to be described in metadata such that various IDEs could consume and transform that information into the formats necessary? Thoughts?
JF: Yes. We can’t force an IDE to embrace what we produce, we can only influence. But by driving adoption across our members of something in this area, we can have strong influence on the IDE vendors to embrace it. This also lends itself to value for our customers who can then depend on the interoperability benefits of working with OAA conformant libraries and components.
CS: At some point getting component providers involved will be good as they can provide feedback to approaches. This exercise should not only include IDE vendors, but also the library and component vendors since they are all part of the eco-system.
JF: Borland is also a member having signed the members’ agreement. Macromedia and Microsoft would be good to have participating as well given their IDEs.
KH: Borland has spun off its tools division into a company called CodeGear. We probably need to reach out to that group now. I do not think that Microsoft is a member of the OAA. Macromedia is already on the list of participants in this task force, but it not present today.
RS: We could also survey our customer bases to get feedback from them.
KH: Good idea, but with broader participation in the IDE task force the members provide a good proxy representation of market needs. I move that we rely on the expertise of our members.
PB: The ATF team has posted a broad overview of their approach to the topic on the wiki and I see that TIBCO has now done the same.
JF: Client-side only? Both server and client-side? Or one after the other?
CS: It may also be good to define common user experiences that could be delivered such as drag and drop assembly, debugging, etc… as suggestions to IDE makers of what would be optimal.
The group collectively discussed and suggested a series of meetings and surveys as follows:
- Next meeting should include more participants to get more feedback. Especially Greg Murray from Sun as his jMaki project has done insightful work in this area.
- A wiki survey of use cases that would need to be supported, followed by the next meeting to discuss those
- A wiki survey of requirements with MUST, SHOULD, MAY rankings of what an implementation would consist of; optionally contributed with initial technical proposals and sample implementations.
- With a strawman proposal in-place invite the component and library vendors to review and advise
Orcale: Add its thoughts to the wiki page where the Eclipse ATF and TIBCO GI high-level descriptions are – especially in regards to client and server side cases and needs so that the group can better understand those.
KH as chairperson: Set up next meeting, but assure better, broader attendance and participation can be achieved.
TIBCO: Add attachments of the meta-data descriptions of its components used by the TIBCO General Interface Builder to register those components.
Next meeting scheduled for Dec-21, 9:00 am Pacific Time same phone #.