Github user JoshRosen commented on the pull request:
https://github.com/apache/spark/pull/2268#issuecomment-54766446
Aha! There was a [hardcoded "1.1" string in
MimaExcludes](https://github.com/apache/spark/blob/f2b5b619a9efee91573c0e546792e68e72afce21/project/MimaExcludes.scala#L36):
```scala
def excludes(version: String) =
version match {
case v if v.startsWith("1.1") =>
Seq(
MimaBuild.excludeSparkPackage("deploy"),
MimaBuild.excludeSparkPackage("graphx")
) ++
...
```
I guess this logic serves to exclude certain incompatibilities introduced
in Spark 1.1.
What's the right fix here (beyond just rolling back the version numbers,
which I'll probably do in the interim)?
At the time of release, 1.1.0 should be binary-compatible with 1.0.0.
Therefore, when testing 1.2.0-SNAPSHOT, is it safe to just [update the
previousSparkVersion]
(https://github.com/apache/spark/blob/f2b5b619a9efee91573c0e546792e68e72afce21/project/MimaBuild.scala#L88)
to 1.1.0 and start with an empty set of excludes if `v.startsWith(1.2)`?
This might have to wait until 1.1.0 is published on Maven.
/cc @pwendell
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
