I did that before I ran the above command to do a test

I’ll try again

Has there been anybody working on kotlin support?

I’m still getting the issue above even after I did a clean-all and added that short statement within my pants.ini

Someone mentioned Kotlin support a while back, IIRC.

Oh no, that was Clojure.

when displaying dependencies with

depmap,

is it possible to show the dependencies classifiers ? I do

./pants bundle foo

and I get a warning

Duplicate classes and/or resources detected in artifacts: (a.jar, b.jar, b-classifier.jar, c.jar)

but when I do

./pants depmap foo --depmap-graph > graph.dot

and then

grep b graph.dot

only

b

appears (and not

b.jar

neither

b-classifier

). How can I track where the duplicates come from ? and is it possible to keep the lib jar names instead of the dependency target name with a dash and a number ? (what does this number means ? it looks like big library are chunked but then I have

lib-1.jar

,

lib-3.jar

,

lib-5.jar

and no trace of 2 or 4 … ) also I don't get what exactly clean

clean-all

. If I run

./pants clean-all

and then

./pants bundle foo

, foo is bundle but I don't see any recompilation.

ok I kind of get the information I'm searching with

resolve

. Is there a way to get a dot graph of

targets

and resolved 3rd party binaries ?

after rebasing my dev, on

./pants bundle foo

I have a

creating dist/job-front.jar
15:25:36 00:02       [create-monolithic-jar]
15:25:36 00:02         [add-internal-classes]
15:25:36 00:02         [add-dependency-jars]
15:25:36 00:02         [cache].
15:25:36 00:02         [bootstrap-jar-tool]
                       ==== stderr ====
                       Unexpected problem writing target jar /home/lorilan/dev/platform/dist/foo.jar: java.util.zip.ZipException: invalid entry size (expected 175 but got 116 bytes)

-ldebug

does not provide any more informations … any ideas ?

Is anyone going to be at Scale by the Bay conf?Maybe give somewhat related talks? Would love to visit and maybe chat afterwards. I know @aloof-angle-91616 mentioned that he might give a talk. Wonder if that happening.

i have been assured that it is indeed happening! just a 5-10 minute one on how we effectively create maintainable webs of versioned tools (toolchains) that can build a large variety of software instead of trying desperately to keep 3rdparty dependencies in sync (a problem much worsened with the current C/C++ ecosystem). at the end there'll be a very miniature demo of building tensorflow custom operators with the new native code/`python_dist()` integration. i should find my time though

@red-balloon-89377 is also going to be there i think?

I’ll be popping in and out of talks, yes. Definitely around to chat.

Also @aloof-angle-91616 When’s your talk?

This OSX

test_dumps_logs_on_terminate

failure seems unrelated to my change: https://travis-ci.org/pantsbuild/pants/jobs/454683251 Does it ring any bells to anyone?

Retrying the shard did not help.

Note that error does not reproduce locally

hmm, never mind, maybe it is related to my change

No, I’m back to thinking it isn’t related

I’m doing something weird, but does anyone have any ideas around this error?:

pants.option.errors.ParseError: Unrecognized command line flags on scope 'binaries': --baseurls

(i.e. how do I specify a global-scoped flag if there’s also a scope which is the prefix thereof)

This is a blind guess, but maybe you can put it in a pants.ini?

Known apparently: https://github.com/pantsbuild/pants/issues/5282 But this works:

$ PANTS_BINARIES_BASEURLS="['bob']" ./pants options | grep baseurls
binaries_baseurls = ['bob'] (from ENVIRONMENT from env var PANTS_BINARIES_BASEURLS)

Thanks!

i have been informed that the lightning talks in scale by the bay are to occur during the "unconference" portion today (https://docs.google.com/spreadsheets/d/1TCnUpG8jQPqzDijF_d7Cwl1AZLDY3V0-OSWyKQdQirI/edit#gid=0), and was informed that the schedule was left intentionally open "in the spirit of lightning talks". will let you know if that becomes more clear (as of now, it is still not clear). regardless i'm wearing a pink twitter shirt and colored hair so please say hi anyone who's at the conference!

Getting

./pants test --cache-test-ignore tests/python/pants_test/backend/jvm/tasks:jar_publish_integration -- -vs
...
                     E   	               Waiting for background workers to finish.
                     E   	00:32:00 00:14   [complete]
                     E   	               SUCCESS
                     E   stderr:
                     E   	.....timestamp: 2018-11-16T00:31:57.685637
                     E   	Signal 15 was raised. Exiting with failure. (backtrace omitted)
                     E   	
                     E   	
                     E   	INFO] killing nailgun server pid=12458
                     E   	INFO] killing nailgun server pid=12518
                     E   	
                     E   	.INFO] killing nailgun server pid=12579
                     E   	INFO] killing nailgun server pid=12596
                     E   	
                     E   	timestamp: 2018-11-16T00:32:00.199617
                     E   	Exception caught: (sqlite3.IntegrityError)
                     E   	  File "/home/root/pants/src/python/pants/bin/pants_loader.py", line 75, in <module>
                     E   	    main()
                     E   	  File "/home/root/pants/src/python/pants/bin/pants_loader.py", line 71, in main
                     E   	    PantsLoader.run()
                     E   	  File "/home/root/pants/src/python/pants/bin/pants_loader.py", line 67, in run
                     E   	    cls.load_and_execute(entrypoint)
                     E   	  File "/home/root/pants/src/python/pants/bin/pants_loader.py", line 60, in load_and_execute
                     E   	    entrypoint_main()
                     E   	  File "/home/root/pants/src/python/pants/bin/pants_exe.py", line 39, in main
                     E   	    PantsRunner(exiter, start_time=start_time).run()
                     E   	  File "/home/root/pants/src/python/pants/bin/pants_runner.py", line 62, in run
                     E   	    return runner.run()
                     E   	  File "/home/root/pants/src/python/pants/bin/local_pants_runner.py", line 159, in run
                     E   	    self._run()
                     E   	  File "/home/root/pants/src/python/pants/bin/local_pants_runner.py", line 233, in _run
                     E   	    run_tracker_result = run_tracker.end()
                     E   	  File "/home/root/pants/src/python/pants/goal/run_tracker.py", line 457, in end
                     E   	    self.store_stats()
                     E   	  File "/home/root/pants/src/python/pants/goal/run_tracker.py", line 400, in store_stats
                     E   	    StatsDBFactory.global_instance().get_db().insert_stats(stats)
                     E   	  File "/home/root/pants/src/python/pants/stats/statsdb.py", line 82, in insert_stats
                     E   	    ri['version'], ri['buildroot'], ri['outcome'], ri['cmd_line']])
                     E

on linux only. any thoughts?

how is build

id

determined?

source: https://api.travis-ci.org/v3/job/455670120/log.txt

running a single test in

tests/python/pants_test/backend/jvm/tasks:jar_publish_integration

is fine, so somehow all the pants invocations from the tests get assigned the same id, in linux