Jon, Alex and I had a planning meeting this morning to come up with a
proposal on what repositories we will need to get started. We also have
some questions that we will need some clarification on.

The overarching strategy is that we want to separate code that is subject
to different build/release schedules and/or different distribution
mechanisms.  This is to simplify our maintenance issues as well as simplify
user access and ability to get started quickly.

1) we propose the following repos:

   - *apache/incubator-datasketches*:  We propose that this would be the
   website source where the README would include a list of links to the
   specific repos below.
   What changes to our current
   https://github.com/DataSketches/DataSketches.github.io website source
   will be required to comply with ASF's website generation process?
   This will determine whether we can move this site at the beginning or
   wait until later.  This is not versioned and no artifacts are generated.
   - *apache/incubator-datasketches-site: *This would be the website binary
   generated by some ASF tool and subject to the comment above.
   - *apache/incubator-datasketches-java: *What we currently have in
   https://github.com/DataSketches/sketches-core.  It is versioned with an
   artifact to Maven Central
   - *apache/incubator-datasketches-cpp: *What we currently have in
   https://github.com/DataSketches/sketches-core-cpp It has an GitHub
   version tag with release notes.
   - *apache/incubator-datasketches-postgresql: What we currently have
   in https://github.com/DataSketches/sketches-postgres
   <https://github.com/DataSketches/sketches-postgres>.* Adaptor for
   PostgreSQL. It has an GitHub version tag with release notes, and a script
   for generating an artifact
   - *apache/incubator-datasketches-hive: What we currently have
   in https://github.com/DataSketches/sketches-hive
   <https://github.com/DataSketches/sketches-hive>.* Adaptors for Hive. It
   is versioned with an artifact to Maven Central
   - *apache/incubator-datasketches-pig: What we currently have
   in https://github.com/DataSketches/sketches-pig
   <https://github.com/DataSketches/sketches-pig>. *Adaptors for Pig.  It
   is versioned with an artifact to Maven Central.
   - *apache/incubator-datasketches-memory: What we currently have
   in https://github.com/DataSketches/memory
   <https://github.com/DataSketches/memory>. *DataSketches-java, pig and
   hive depend on this and other systems outside of DataSketches also depend
   on this. This has virtually no external run-time dependencies and would be
   ported first. It is versioned with an artifact to Maven Central.
   - *apache/incubator-datasketches-characterization:* This would be a
   merge of our https://github.com/DataSketches/characterization and
   https://github.com/DataSketches/characterization-cpp sites.  This is not
   versioned and no artifacts are generated.

2) The remaining repos currently in https://github.com/DataSketches
organization
are not critical and we can defer to later. They might either be obsoleted
or deleted.

3) Any comments on the above proposal?

4) The DataSketches link
<https://incubator.apache.org/projects/datasketches.html> is broken on the
Apache Incubator page.  What do we need to get that set up?

5) Both *Jon Malkin* and *Edo Liberty* have signed and submitted their
ICLAs. Why are they not listed on
https://whimsy.apache.org/roster/ppmc/datasketches?  Both of them should be
part of the PPMC.

6)* [email protected] <[email protected]>* should also be on the PPMC
list.

Reply via email to