2.12: new Scala SHA

[SethTisue]

https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5532/ (2.12.11-bin-466f92f)

[SethTisue]

scalariform tests failing: scala/scala#8466 (comment)

[SethTisue]

scalariform thing now has its own ticket, milestoned 2.12.11: scala/scala-dev#672

some more 2.12.x PRs were merged today, so I've bumped the SHA again and queued another run: https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5536/ fyi @lrytz

[SethTisue]

two additional repos started failing:

• scalikejdbc: https://scala-ci.typesafe.com/job/scala-2.12.x-integrate-community-build/5540/artifact/logs/scalikejdbc-build.log
• scalafix: https://scala-ci.typesafe.com/job/scala-2.12.x-integrate-community-build/5540/artifact/logs/scalafix-build.log

PRs that might be the culprits, last-merged first:

• 2.12.11-bin-ab4d97e - Merge pull request remove some refs in TraversableOnce scala#8732 from rorygraves/mike/2.12.x_quickTraversable
  • https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5546/
• 2.12.11-bin-348c541 - Merge pull request A simple TreeSetBuilder scala#8736 from mkeskells/2.12.x_sortedSet-2
  • https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5545/
• 2.12.11-bin-90bc6cf - Merge pull request Don't create a new TreeSet if there is no change to the underlying tree scala#8737 from mkeskells/2.12.x_sortedSet-1
  • https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5544/
• 2.12.11-bin-b4c0225 - Merge pull request dont create a TreeMap if the underlying tree is unchanged scala#8738 from mkeskells/2.12.x_TreeMap-1
  • https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5543/

[SethTisue]

I will dig into this a bit and once I know a bit more, I will open one or more scala-dev tickets on the 2.12.11 milestone

[SethTisue]

scalikejdbc failure is reproducible outside dbuild (on scalikejdbc/scalikejdbc@12b8f45) with core/testOnly *ResultSetTraversableSpec. bisecting shows that it regressed in scala/scala#8732 (2.12.11-bin-ab4d97e)

scalafix failure is really hard to reproduce outside of dbuild. the scalafix commit we build is on 2.12.6, not 2.12.10, and there are multiple uses of CrossVersion.full in the build. I spent about an hour on it before deciding the rabbit hole is too deep. (in case we end up down the hole again, I pushed my wip to a seth-20200228 branch in SethTisue/scalafix, branched off their v0.9.9 tag which is on 2.12.10 but, who knows, might not even be a point in time where the problem is reproducible at all); reflect/update fails for reasons I don't understand).

I suggest we fix whatever bug is affecting scalikejdbc, then hope that fixes scalafix as well. and if it doesn't, we either go back down the rabbit hole, or we use dbuild for scalafix testing. the problem with the latter is that scalafix sits atop a tall pile of dependencies that take a long time to build.

[SethTisue]

I did a series of runs on a subset of the community build that verified that scalafix regressed in the same PR that scalikejdbc did, so I'm pretty confident that scala/scala#8769 will fix the problem.

[SethTisue]

well, this is weird. the other repos are green, but now airframe is failing... I'll look into it

https://scala-ci.typesafe.com/job/scala-2.12.x-integrate-community-build/5553/artifact/logs/airframe-build.log

[SethTisue]

In local testing, the test also fails on 2.12.10 and 2.13.1. I hypothesize this is related to today/yesterday being a leap day.

and sure enough, I looked at the source code for the test, and:

  def `succinct time window unit`: Unit = {
    // 2020 have 366 days, so we need to fix the offset
    val t = TimeWindow.withUTC.withOffset("2019-01-01 00:00:00")

(Also, I looked at the source code for TimeVector itself and it has nothing to do with scala.collection.immutable.Vector or with anything else in scala.collection, afaict.)

[SethTisue]

I merged the reversion PR and excluded the airframe test (and reported the problem upstream: wvlet/airframe#953). I'll start a new run as soon as Travis-CI finishes building our new RC (2.12.11-bin-d1e9ac0)

[SethTisue]

https://scala-ci.typesafe.com/view/scala-2.12.x/job/scala-2.12.x-integrate-community-build/5554/