mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andrew Palumbo <andrew_palu...@msn.com>
Subject [ANNOUNCE] Mahout 14.1 RC4
Date Sat, 01 Feb 2020 10:41:12 GMT
Fixing some typos and mistyping:

Hi All,

Ive finished RC4 of Apache Mahout 0.14.  Unfortunately Right off the bat, I can see an issue,
which will require a new RC (only SHA-1) checksums have been included.  Apache requires at
least sha256.  So I will -1 it myself.  I'd drop it and fix that now, but its late.

Please bear with me as this is a completely new structure of the project, which required almost
a complete rewrite of all poms, as well as my first time acting as release engineer.

@PMC,  As I've already this RC is invalid,  I don't think that a vote is necessary, but if
you all could take a look through this, and let me know if you see any issues, from trivial
to major, it would be a great help in getting this release out the door finally.  I'm going
to take a page out of Suneels book and ask for certain tasks that we need to get get RC5 down
hopefully to just running tests.

        resolver: [1] https://repository.apache.org/content/repositories/orgapachemahout-1061
        artifacts: [2] https://repository.apache.org/content/repositories/orgapachemahout-1061/org/apache/mahout/

@r<mailto:trevor.d.grant@gmail.com>awkintrevo, I know that you needed binaries, could
you please try to resolve artifacts that you need and give them a spin, and  if you have time
check to see if the binary distribution. tar.gz contains all the necessary?

@pat<mailto:pat@occamsmachete.com> could you please see if the 2.11 artifacts suit your
specific needs (and work)?

@akm could you please  give a once over (or twice, or thrice over if you have time) and see
if there are any gaping, obvious issues, especially with the binary and source release artifacts,
as well could you please let me know about the new structure (releasing 2.11 and 2.12 in the
same repo), and if it seems to work?

I'll have to concentrate my limited time, this week, on squashing these last couple of bugs.

@all please let me know if the structure seems to work.

If anyone has a quick script to verify the sigs it would  be useful but that's probably best
saved for RC5.

As well I've used the naming scheme used by SBT:

                       ${module_name}_{scala.major.minor}-{mahout version}

Please keep an eye out for any transpositions between ${scala.major.minor} and ${mahout version}.

Both SBT and maven should have no issue resolving these artifacts (I tested maven only in
an earlier RC).


It is  late, and I've done no testing other than running tests before i cut this RC, nor have
i looked closely at any of the artifacts, so anything that seems out of the ordinary, Please
take note of

There still are still issues with releasing that i have to fix, I.e  the maven-release-plugin
is not functioning correctly (I believe, haven't fully tracked down the bug) these artifacts
were .  its a new issue since RC3 and there have been very few changes since then, it may
be as simple as some spacing in the pom (first few answers i found when searching.  The issue
may be with the formatting of the XML itself.  I'll try to track this down as when I get back
to it.  And as I mentioned, the release:perform goal failed, when run as:

             $ mvn clean package install -DskipTests -Papache-release release:perform

I will try to create a Jirass for these issues, (release failing and sha1) I'm not sure how
much work I'll be able to do on this next week or over the weekend.  The release instructions
have changed slightly and I will document these in the Jiras or on the PR with the new required
~/.m2/settings.xml.

I will be on and offline for the next few days, so @akm, if you want to release before i get
back to this, I'll finish documenting the release steps in either the Jira, or the PR that
I sent to you yesterday [3].

If we run these few smoke tests, and sanity checks (and as many more as anyone has time for),
it will go a long way in raising the probability of RC5 being the final.

Of course, as always, any testing and feedback is welcomed and greatly  appreciated from anyone
on the list.

Thanks very much,

Andy


BTW:

It seems that the RC was not assigned a test number for the the artifacts [4]. e.g:

<dependency>
     <groupId>org.apache.mahout</groupId>
     <artifactId>mahout-core_2.11</artifactId>
     <version>14.1</version>
</dependency>

This should not be an issue, as there is no other version 14.1 in maven central or other repos,
nor are there a scala 2.11 or 2.12 versions of anything [4].  However if you've tried testing
a 14.1 RC before, please remember to use the -U option when building, or clear out your ~/.m2
./repository.

[1] https://repository.apache.org/content/repositories/orgapachemahout-1061
[2] https://repository.apache.org/content/repositories/orgapachemahout-1061/org/apache/mahout/
[3] https://github.com/apache/mahout/pull/384/commits/8794c42f910ed6fc3ac93d4081a88385520cc84b
[4] https://repository.apache.org/#stagingRepositories
[5] https://mvnrepository.com/artifact/org.apache.mahout/mahout-spark_2.10/0.13.0





Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message