Code Recommenders projects want to do a 1.0 release with Juno. We
already spoke about graduating Code Recommenders a while ago  and I'd
like to follow up the discussion here on the list.
started in January 2011 and has since then quickly made progress. For
Juno we have the opportunity to join the Java, RCP, and Scout package to
present Code Recommenders to a broader audience - which would in my
opinion be a big thing to get the idea or Recommenders in Software
Development out to Eclipse users and developers.
as such is still rather small. We have 5 active committers, 3 employed
by Darmstadt University of Technology, one from Queens University, and
one independent. In the past year we had roughly a dozen contributions
of varying sizes (one-line fixes up to complete features such as local
code search engines, code snippet completion, or code summarization
features) - of which seven have been applied so far.
also has a good recognition in the Eclipse user community as the
Eclipse Community Awards have shown.
The project also has some
effect in research circles. Other universities consider joining Code
Recommenders (as Queens), actively start reusing components for their
own research work (Bejing), or consider doing joint research on areas
related to code recommenders (McGill).
The code base is
getting stable and the CI build is doing its work - although the latest
integration into the Juno package popped up a few issues that have to be
Features such as Subwords or Chain completion
have improved significantly in the past months and their concepts are
now evaluated by JDT and may find their way into many more areas than
just code completion, see for example .
In terms of
documentation and recommendation model coverage there is still some work
to do. However, we are confident that we'll improve on this until Juno.
whats your opinion on this? Is there anything you'd like to see
addressed before graduation can happen?