Re: Status of on-going PRs

[Frank McQuillan <fm...@pivotal.io>]

Looks like all of the PRs for 1.10 have been merged, except
https://github.com/apache/incubator-madlib/pull/75
which will spill over to 2.0.  Thank you all who contributed on doing this.

We are still doing some final checking on E2E functional test suite across
postres, greenplum and apache hawq, so not officially at code freeze yet.
But getting dangerously close.

Frank

On Tue, Jan 31, 2017 at 12:22 PM, Rahul Iyer <ra...@gmail.com> wrote:

> Hi Satoshi,
>
> Thanks for compiling this list. Please find my comments inline.
>
> On Tue, Jan 31, 2017 at 3:04 AM, Satoshi Nagayasu <sn...@uptime.jp> wrote:
>
> > Hi all,
> >
> > As release manager for 1.10, I just did a quick review and created a
> status
> > list of the on-going PRs.
> >
> > https://github.com/apache/incubator-madlib/pulls
> >
> > If you have comments, please let me know. I will update the status.
> >
> > Status of the PRs
> > -----------------
> > Use relative path for installation in GPDB/HAWQ #94
> >   -> Need to be tested with GPDB/HAWQ.
> >
> > Build: Use only major version for GPDB 5, HAWQ 2 #91
> >   -> Need review?
> >
> ​Testing is complete for both PRs. Requires a review.
> ​
>
> > Allow encode_categorical_variables() to use the svec type. #93
> >   -> Need more work by the developer (me).
> >
> ​This would be better merged within the 1.10 release.
> Adding it to the next version would require special handling by upgrade
> since there is a change in argument type (hence requiring drop/replace
> during upgrade).
>
>
> >  K-means: support for array input #89
> >   -> Need more review, or ready for committer?
> >
> ​This looks ready to merge. ​
>
> >
> > JIRA: MADLIB-927 Changes made in KNN-help message-test cases-etc #81
> >   -> Need more work by the developer.
> >
> > HAWQ2.1: Changes the cmake to assume any HAWQ 2.X system is 2.0 and #79
> >   -> Need review, or ready for committer?
> >
> ​This is superseded by #91 and will be closed with it. ​
>
>
> > Include boost::format in MathToolkit_impl.hpp. #76
> >   -> Already merged. The PR can be closed.
> >
> ​I forgot to close this with the commit message and can only be manually
> closed by the contributor. If not closed soon, I'll close it with a future
> commit.
> ​
>
> > SVM: Implement c++ functions for training multi-class svm in mini-batch
> #75
> >   -> The doc needs to be updated?
> >
> ​This requires substantial more work and discussion as the scope of the
> work is not defined. We will have to ​release without it.
>
>
>
> >
> > Regards,
> > --
> > Satoshi Nagayasu <sn...@uptime.jp>
> >
>