On Dec 5, 2007, at 9:49 PM, Michael Busch wrote:
Hi all,
we were talking about releasing Lucene 2.3 already some weeks ago, but
it seems that the list with issues targeted for 2.3 doesn't get
smaller.
Currently there are these 7 issues in the list:
- LUCENE-1061: Adding a factory to QueryParser to instantiate query
instances (Unassigned)
- LUCENE-1044: Behavior on hard power shutdown (Mike)
- LUCENE-1052: Add an "termInfosIndexDivisor" to IndexReader (Mike)
- LUCENE-1001: Add Payload retrieval to Spans (Grant)
Moved to later.
- LUCENE-1045: SortField.AUTO doesn't work with long (Grant)
We need a decision here on how best to change this. We have two
viable patches, one that keeps the status quo, more or less and one
that changes FieldCache to be a class instead of an Interface and also
removes FieldCacheImpl. It is NOT back-compat. but the feeling is
(and I asked on user) that this change is acceptable. It would,
however, require a recompile, as Yonik points out. Probably not a big
deal either as long as we advertise this up front, but I don't feel
comfortable making the decision (I am happy w/ either patch.) The
change to class, I think makes more sense in the long run.
I suppose the other alternative is to make the changes to the
existing, as in the first patch, then deprecate ExtendedFieldCache and
mark FieldCache such that it will be changed in 3.x to be a class,
since that will require a recompile, presumably.
Either way, once the decision is made, it is pretty trivial to make
the appropriate changes.
- LUCENE-1077: New Analysis Contributions (Grant)
I will probably commit this in a day or two (I just uploaded a
patch). It is all new functionality to provide various SinkTokenizer
implementations and a few convenience classes for working with
Payloads. Since these are additive, it shouldn't hold up the release.
- LUCENE-1079: DocValues cleanup: constructor & getInnerArray()
(Doron)
I think we all voted earlier this year to have more frequent releases
instead of very big ones. But I believe that this release has
already a
whole bunch of new features (e. g. DocumentsWriter, background merges,
IndexReader.reopen()), so I think we should stop adding more
features to
the 2.3 list (except bug fixes of course) and plan a bit ahead how
long
the above mentioned issues are going to take.
So everyone on the list above: Could you give an estimate on when the
patches will be ready to commit? And are there issues on list that
shouldn't block 2.3? I think 1061 shouldn't block the release, because
nobody is assigned.
Another question to everyone: Are there other issue that you really
want
to get into 2.3 that are not on the list?
I'd like to get LUCENE-584 in, but I have to read through the recent
comments to figure out how much work is left. It should not block
2.3 if
all other issues are resolved.
I think a good target would be to complete all 2.3 issues by end of
this
year. Then we can start a code freeze beginning of next year, so that
we'll have 2.3 out hopefully by mid/end of January '08. I would
volunteer to act as the release manager again.
Best,
-Michael
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
--------------------------
Grant Ingersoll
http://lucene.grantingersoll.com
Lucene Helpful Hints:
http://wiki.apache.org/lucene-java/BasicsOfPerformance
http://wiki.apache.org/lucene-java/LuceneFAQ
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]