[news.eclipse.technology.ldt] Re: What is the contribution?

["johnc" <jcamelon@xxxxxxxxxx>]

e.g. An aspect that several projects will probably need to deal with
involves the management of remote systems and tools. Wind River's new device
project plans to deal with this problem (I believe) but there will be (most
likely) a project wide dependency upon the CDT.  Other projects could
benefit from these constructs and interfaces (remote embedded Java
debugging, TPTP, just off the top of my head) and unless the device project
componentizes their solution with this in mind, a dependency upon CDT
inhibits the adoption of these constructs for other domains.  Often pure
self interest does not produce the best componentization.

While I understand that BEA wants to get its feet wet in areas that it cares
more about, what I would like this project to be the collaboration point for
all language-related projects to come up with a roadmap regarding
componentization.  The short term answer is to put more stuff in the LDT or
the platform, but we'll need to come up with a way of managing this process
as projects evolve (and even mutate!).

Cheers,
JohnC
www.eclipse.org/cdt

"Tim Wagner" <twagner@xxxxxxx> wrote in message
news:d0qdlp$i9a$1@xxxxxxxxxxxxxxxxxx
> BEA originally proposed a somewhat wider scope for the LDT project, but
> initial guidance from people inside Eclipse was that too grand a scope
might
> set the wrong expectation or cause a certain amount of confusion.
>
> After EclipseCon and a good deal of verbal feedback, it seems clear that
> several areas of parallel development are indeed interesting to the
> community, including enhanced Java support, a reified language model and
> generic editor services, support around metalinguistic generation (like
> Chris Daly's work), dynamic/interpreted language support, and so forth.
>
> With regard to contributions: BEA is already actively engaged in working
> with the JDT team on an initial project, bringing APT/Mirror support (JSR
> 269) to the Eclipse Java editing experience, and this includes migrating
> concepts and in some cases code from our existing code to the JDT. We'll
> apply the same logic to other LDT projects: If there are helpful
algorithms,
> data structures, or language-based APIs we've devised, then we're happy to
> share them with the larger community. What I don't want to do is drop a
huge
> ZIP file and suggest it as the answer to a complex constellation of
language
> and compilation challenges...it's really the developer support and overall
> community interest that should drive specific BEA contributions.
>
>
>
> "Bob Foster" <bob@xxxxxxxxxx> wrote in message
> news:d0orud$3u1$1@xxxxxxxxxxxxxxxxxx
> >I asked this question, "What is the contribution?" at the late-scheduled
> >EclipseCon LDT session. The speaker turned it into, "Is there a
> >contribution?" and answered to the effect that BEA has a lot of code on
the
> >subject and might consider making some of it open source.
> >
> > To which I say, hmmm.
> >
> > The thing that struck me most about the talk is how grand was the
vision,
> > yet most comments here are to the effect that the scope is too limited.
I
> > suspect that the difference is that the BEA team has more experience in
> > the trenches with actually trying to generalize multiple language
support
> > in the context of practical examples, like JSP, annotations, etc., while
> > the commenters have the luxury of still being able to discuss the
subject
> > in the abstract.
> >
> > This would be a good conversation, if there were one. What is needed is
> > some comment from BEA.
> >
> > Bob
> >
> > Bob Foster wrote:
> >> With all those BEA heads, it's obvious this is a BEA-inspired
initiative
> >> (which is a good thing) and probably means BEA intends to contribute
some
> >> technology to kick the project off. It's entirely appropriate to
discuss
> >> what that technology is and how its approach differs from or works with
> >> other initiatives in this area, like the previously mentioned SSE, and
> >> Chris Laffra's generative plugin work.
> >>
> >> I and I'm sure many others are very excited by work in this area. I'm
> >> looking forward to your Thursday 2 pm session at EclipseCon (for other
> >> attendees, the session didn't make the program - check the fliers near
> >> the messages board) where some of these questions will no doubt be
> >> answered. But please remember the millions of Eclipse developers who
> >> aren't at EclipseCon, and put that info on the web.
> >>
> >> Bob Foster
>
>