Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Antoine Pitrou
It would be good to test all Python versions in a cron build, but I agree we may not need to test all Python 3 versions in per-commit builds. Regards Antoine. Le 07/08/2018 à 03:14, Robert Nishihara a écrit : > Thanks Wes. > > As for Python 3.5, 3.6, and 3.7, I think testing any one of them

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
OK, website update PR is up https://github.com/apache/arrow/pull/2387 Everyone OK to announce at 12pm Eastern tomorrow? On Mon, Aug 6, 2018 at 9:03 PM, Wes McKinney wrote: > I just pushed the documentation update. I had to comment out the JS > doc build but otherwise didn't have any issues > > O

[jira] [Created] (ARROW-3005) [Website] Update website and write blog post for 0.10.0 release announcement

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-3005: --- Summary: [Website] Update website and write blog post for 0.10.0 release announcement Key: ARROW-3005 URL: https://issues.apache.org/jira/browse/ARROW-3005 Project: Apa

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Robert Nishihara
Thanks Wes. As for Python 3.5, 3.6, and 3.7, I think testing any one of them should be sufficient (I can't recall any errors that happened with one version and not the other). On Mon, Aug 6, 2018 at 12:01 PM Wes McKinney wrote: > @Robert, it looks like NumPy is making LTS releases until Jan 1,

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
I just pushed the documentation update. I had to comment out the JS doc build but otherwise didn't have any issues On Mon, Aug 6, 2018 at 8:59 PM, Krisztián Szűcs wrote: > Conda packages (arrow-cpp, parquet-cpp, pyarrow) are also available now on > conda-forge. > > Tested pyarrow on OSX + PY27/3

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Krisztián Szűcs
Conda packages (arrow-cpp, parquet-cpp, pyarrow) are also available now on conda-forge. Tested pyarrow on OSX + PY27/35/36, all of them are working. On Aug 6 2018, at 11:16 pm, Wes McKinney wrote: > > Wheels are up https://pypi.org/project/pyarrow/#files > On Mon, Aug 6, 2018 at 4:38 PM, Phillip

[jira] [Created] (ARROW-3004) [Documentation] dev/gen_apidocs.sh builds docs for master rather than a pinned commit

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-3004: --- Summary: [Documentation] dev/gen_apidocs.sh builds docs for master rather than a pinned commit Key: ARROW-3004 URL: https://issues.apache.org/jira/browse/ARROW-3004 Pro

Re: Arrow & plasma - java sample to store complex objects

2018-08-06 Thread Wes McKinney
hi Gerard, This is the right place to ask questions. The Slack channel was closed (see prior discussions on the mailing list); few Java developers were on Slack anyway so it wouldn't have been a good place to get help. Using Java with Plasma is very bleeding edge territory. I don't know if anyone

Re: [DISCUSS] Contribution of Gandiva to Apache Arrow

2018-08-06 Thread Wes McKinney
hi Ravindra, Thanks for the update. I will start a vote to approve the donation of the codebase soon if there are no objections. The merge will be subject to completing the standard ASF IP Clearance and having a Software Grant filed by Dremio. I noticed some things in the Gandiva codebase vs. the

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Li Jin
I don’t think so. I manually checked it out under arrow/site. On Mon, Aug 6, 2018 at 6:20 PM Phillip Cloud wrote: > Is it being checked out as a submodule? > > On Mon, Aug 6, 2018 at 6:07 PM Li Jin wrote: > > > I've generated the docs but unfortunately is stuck on a weird git issue. > > For som

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Phillip Cloud
Is it being checked out as a submodule? On Mon, Aug 6, 2018 at 6:07 PM Li Jin wrote: > I've generated the docs but unfortunately is stuck on a weird git issue. > For some reason git doesn't want to add a few files after "git add ." (10 > python files): > > modified: docs/python/parquet.html >

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Li Jin
I've generated the docs but unfortunately is stuck on a weird git issue. For some reason git doesn't want to add a few files after "git add ." (10 python files): modified: docs/python/parquet.html modified: docs/python/plasma.html modified: docs/python/search.html modified: docs/python/

[jira] [Created] (ARROW-3003) Unable Java doc in dev/gen_apidocs/create_documents.sh

2018-08-06 Thread Li Jin (JIRA)
Li Jin created ARROW-3003: - Summary: Unable Java doc in dev/gen_apidocs/create_documents.sh Key: ARROW-3003 URL: https://issues.apache.org/jira/browse/ARROW-3003 Project: Apache Arrow Issue Type: Tas

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Phillip Cloud
Excellent. Thanks for testing that out Krisztián. On Mon, Aug 6, 2018 at 5:32 PM Krisztián Szűcs wrote: > Great! Just tested to be sure, the osx wheels (all three python versions) > are working. > > On Aug 6 2018, at 11:16 pm, Wes McKinney wrote: > > > > Wheels are up https://pypi.org/project/p

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Krisztián Szűcs
Great! Just tested to be sure, the osx wheels (all three python versions) are working. On Aug 6 2018, at 11:16 pm, Wes McKinney wrote: > > Wheels are up https://pypi.org/project/pyarrow/#files > On Mon, Aug 6, 2018 at 4:38 PM, Phillip Cloud wrote: > > The release artifacts are uploaded to SVN.

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
Wheels are up https://pypi.org/project/pyarrow/#files On Mon, Aug 6, 2018 at 4:38 PM, Phillip Cloud wrote: > The release artifacts are uploaded to SVN. > > On Mon, Aug 6, 2018 at 4:01 PM Wes McKinney wrote: > >> We can comment out the JS doc build for now, so it's not a blocker, >> but would be

[jira] [Created] (ARROW-3002) Inconsistent DataType Hashing

2018-08-06 Thread Sam Oluwalana (JIRA)
Sam Oluwalana created ARROW-3002: Summary: Inconsistent DataType Hashing Key: ARROW-3002 URL: https://issues.apache.org/jira/browse/ARROW-3002 Project: Apache Arrow Issue Type: Bug

[jira] [Created] (ARROW-3001) [Packaging] Don't modify PATH during rust release verification

2018-08-06 Thread Krisztian Szucs (JIRA)
Krisztian Szucs created ARROW-3001: -- Summary: [Packaging] Don't modify PATH during rust release verification Key: ARROW-3001 URL: https://issues.apache.org/jira/browse/ARROW-3001 Project: Apache Arro

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Phillip Cloud
The release artifacts are uploaded to SVN. On Mon, Aug 6, 2018 at 4:01 PM Wes McKinney wrote: > We can comment out the JS doc build for now, so it's not a blocker, > but would be good to fix > > On Mon, Aug 6, 2018 at 3:59 PM, Paul Taylor wrote: > > Looks like typedoc is using the wrong version

[jira] [Created] (ARROW-3000) [Python] Do not build unit tests other than python-test in travis_script_python.sh

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-3000: --- Summary: [Python] Do not build unit tests other than python-test in travis_script_python.sh Key: ARROW-3000 URL: https://issues.apache.org/jira/browse/ARROW-3000 Projec

[jira] [Created] (ARROW-2999) [Python] Do not run ASV benchmarks in every Travis CI build to improve runtimes

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2999: --- Summary: [Python] Do not run ASV benchmarks in every Travis CI build to improve runtimes Key: ARROW-2999 URL: https://issues.apache.org/jira/browse/ARROW-2999 Project:

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
We can comment out the JS doc build for now, so it's not a blocker, but would be good to fix On Mon, Aug 6, 2018 at 3:59 PM, Paul Taylor wrote: > Looks like typedoc is using the wrong version of typescript. I can take a > look at updating it later this afternoon if it’s blocking the release. > >

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Paul Taylor
Looks like typedoc is using the wrong version of typescript. I can take a look at updating it later this afternoon if it’s blocking the release. On Mon, Aug 6, 2018 at 12:55 PM Li Jin wrote: > I was trying to build the documentation but hit a JavaScript error. I am > not familiar with JavaScript

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
You hit https://issues.apache.org/jira/browse/ARROW-2925. We can skip updating the JS documentation for now @Paul, can you take a look? On Mon, Aug 6, 2018 at 3:54 PM, Li Jin wrote: > I was trying to build the documentation but hit a JavaScript error. I am > not familiar with JavaScript and Type

[jira] [Created] (ARROW-2998) [C++] Add variants of AllocateBuffer, AllocateResizeableBuffer that return unique_ptr

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2998: --- Summary: [C++] Add variants of AllocateBuffer, AllocateResizeableBuffer that return unique_ptr Key: ARROW-2998 URL: https://issues.apache.org/jira/browse/ARROW-2998 Pro

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Li Jin
I was trying to build the documentation but hit a JavaScript error. I am not familiar with JavaScript and TypeScript ... Any suggestion? Error log: /apache-arrow/arrow/js /apache-arrow + npm install npm WARN optional Skipping failed optional dependency /chokidar/fsevents: npm WARN notsup Not c

[jira] [Created] (ARROW-2997) [Python] Scripts for uploading conda binary release artifacts to anaconda.org under @apache account

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2997: --- Summary: [Python] Scripts for uploading conda binary release artifacts to anaconda.org under @apache account Key: ARROW-2997 URL: https://issues.apache.org/jira/browse/ARROW-2997

[jira] [Created] (ARROW-2996) [C++] Fix typo in cpp/.clang-tidy

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2996: --- Summary: [C++] Fix typo in cpp/.clang-tidy Key: ARROW-2996 URL: https://issues.apache.org/jira/browse/ARROW-2996 Project: Apache Arrow Issue Type: Improvement

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
I have just rebased master on tag apache-arrow-0.10.0. I will rebase the PRs that might be affected On Mon, Aug 6, 2018 at 2:55 PM, Krisztián Szűcs wrote: > Wes, I can help You with the forge packages. I'm creating the PRs. > > On Aug 6 2018, at 8:52 pm, Wes McKinney wrote: >> >> I'm going to st

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Wes McKinney
@Robert, it looks like NumPy is making LTS releases until Jan 1, 2020 https://docs.scipy.org/doc/numpy-1.14.0/neps/dropping-python2.7-proposal.html Based on this, I think it's fine for us to continue to support Python 2.7 until then. It's only 16 months away; are you all ready for the next decade

[jira] [Created] (ARROW-2995) [CI] Build Python libraries in same run when running C++ unit tests so project does not need to be rebuilt again right away

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2995: --- Summary: [CI] Build Python libraries in same run when running C++ unit tests so project does not need to be rebuilt again right away Key: ARROW-2995 URL: https://issues.apache.org/j

[jira] [Created] (ARROW-2994) [C++] Only include Python C header directories for Python-related compilation units

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2994: --- Summary: [C++] Only include Python C header directories for Python-related compilation units Key: ARROW-2994 URL: https://issues.apache.org/jira/browse/ARROW-2994 Proje

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Krisztián Szűcs
Wes, I can help You with the forge packages. I'm creating the PRs. On Aug 6 2018, at 8:52 pm, Wes McKinney wrote: > > I'm going to start kicking along the conda-forge packages. If anyone > can assist with these, it would be much appreciated. > > I will update the website and write a blog post sum

Re: [RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
I'm going to start kicking along the conda-forge packages. If anyone can assist with these, it would be much appreciated. I will update the website and write a blog post summarizing the 0.10.0 release (it's the biggest release we've ever done). I'm going to rebase master as soon as I merge the bu

Re: [VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Wes McKinney
Thanks Kou. I also noted the NodeJS 8.11.x LTS failure. Opened https://issues.apache.org/jira/browse/ARROW-2993 about better documenting this On Mon, Aug 6, 2018 at 4:26 AM, Kouhei Sutou wrote: > +1 (binding) > > Ran release verification script on Debian GNU/Linux sid > > * gcc version 8.2.0 (D

[jira] [Created] (ARROW-2993) [JS] Document minimum supported NodeJS version

2018-08-06 Thread Wes McKinney (JIRA)
Wes McKinney created ARROW-2993: --- Summary: [JS] Document minimum supported NodeJS version Key: ARROW-2993 URL: https://issues.apache.org/jira/browse/ARROW-2993 Project: Apache Arrow Issue Type:

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Robert Nishihara
> > Also, at this point we're sometimes hitting the 50 minutes time limit on > our slowest Travis-CI matrix job, which means we have to restart it... > making the build even slower. > Only a short-term fix, but Travis can lengthen the max build time if you email them and ask them to.

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Antoine Pitrou
Also, at this point we're sometimes hitting the 50 minutes time limit on our slowest Travis-CI matrix job, which means we have to restart it... making the build even slower. There's something perhaps suboptimal in the way we build Arrow C++ on Travis: - first we build it for no particular Python

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Antoine Pitrou
Not wanting to answer for Wes, but those are two sides of the same coin: reducing CI overhead and complexity helps increase developer productivity. Reducing CI overhead is not a goal *in itself* (unless there are money issues I don't know about) ;-) The productivity cost of being Python 2-compa

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Robert Nishihara
Wes, do you primarily want to drop Python 2 to speed up Travis or to reduce the development overhead? In my experience the development overhead is minimal and well worth it. For Travis, we could consider looking into other options like paying for more concurrency. January 2019 is very soon and Pyt

[RESULT][VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Phillip Cloud
With 4 binding +1 votes (myself, Uwe, Wes, Kou), 2 non-binding +1 (Krisztián, Li), and no other votes, the vote passes. Thanks everyone! I will upload the Java packages as per the release management wiki. Would some folks please volunteer to get the Python packaging, documentation and website upd

[jira] [Created] (ARROW-2992) [Python] Parquet benchmark failure

2018-08-06 Thread Antoine Pitrou (JIRA)
Antoine Pitrou created ARROW-2992: - Summary: [Python] Parquet benchmark failure Key: ARROW-2992 URL: https://issues.apache.org/jira/browse/ARROW-2992 Project: Apache Arrow Issue Type: Bug

Arrow & plasma - java sample to store complex objects

2018-08-06 Thread Gérard Dupont
Hi, Not sure this is the right channel for a "user" oriented question but the slack channel on heroku seams to be down... TL;DR: is there some hidden tutorial/java samples to store complex data objects in arrow and access (put/get) with plasma? I'm currently exploring the unit test from the java p

[jira] [Created] (ARROW-2991) [CI] Cut down number of AppVeyor jobs

2018-08-06 Thread Antoine Pitrou (JIRA)
Antoine Pitrou created ARROW-2991: - Summary: [CI] Cut down number of AppVeyor jobs Key: ARROW-2991 URL: https://issues.apache.org/jira/browse/ARROW-2991 Project: Apache Arrow Issue Type: Task

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Wes McKinney
> The 40+ minutes Travis-CI job already uses the toolchain packages AFAIK. > Don't they include thrift? I was referring to your comment about "parquet-cpp AppVeyor builds are abysmally slow". I think the slowness is in significant part due to the ExternalProject builds, where Thrift is the worst

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Wes McKinney
hi, On Mon, Aug 6, 2018 at 7:52 AM, Antoine Pitrou wrote: > > Le 06/08/2018 à 13:42, Wes McKinney a écrit : >> hi Antoine, >> >> I completely agree. Part of why I've been so consistently pressing for >> nightly build tooling is to be able to shift more exhaustive testing >> out of per-commit runs

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Antoine Pitrou
Le 06/08/2018 à 13:42, Wes McKinney a écrit : > hi Antoine, > > I completely agree. Part of why I've been so consistently pressing for > nightly build tooling is to be able to shift more exhaustive testing > out of per-commit runs into a daily build or an on-demand build to be > invoked by the u

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Wes McKinney
hi Antoine, I completely agree. Part of why I've been so consistently pressing for nightly build tooling is to be able to shift more exhaustive testing out of per-commit runs into a daily build or an on-demand build to be invoked by the user either manually or by means of a bot. If you search in J

Re: [DISCUSS] Re-think CI strategy?

2018-08-06 Thread Krisztián Szűcs
Hi, A straightforward way would be to run non-critical CI jobs as nightlies. Nightly package builds work pretty well, see the following link https://github.com/kszucs/crossbow/branches/all?query=nightly the notification logic requires improvement though. We should also run integrations tests regul

[DISCUSS] Re-think CI strategy?

2018-08-06 Thread Antoine Pitrou
Hello, Our CI jobs are taking longer and longer. The main reason seem not to be that our test suites become more thorough (running tests actually seems to account for a very minor fraction of CI times) but the combined fact that 1) fetching dependencies and building is slow 2) we have many conf

Re: [VOTE] Release Apache Arrow 0.10.0 (RC1)

2018-08-06 Thread Kouhei Sutou
+1 (binding) Ran release verification script on Debian GNU/Linux sid * gcc version 8.2.0 (Debian 8.2.0-1) * Ruby 2.6.0dev * NodeJS 10.8.0 * Note: NodeJS 8.11.2 (LTS) failed tests FlatBuffers 1.9.0 can't build with GCC 8.2.0. So I need to use the master like the following command line: