Github user pwendell commented on the pull request:
https://github.com/apache/spark/pull/5363#issuecomment-89756227
Hey Sean,
Thanks for posting this. So my straw man alternative is to just add
hadoop2.5 and hadoop2.6 profiles and associated documentation. My concern is
that this current approach is not future proof in the likely even that for some
future Hadoop (say 2.7 or 2.8) requires a different build profile. Someone will
google and find this documentation that suggests they can use the hadoop2.4
profile, then they could get confusing build errors. From what I understand
Hadoop makes no firm commitment around dependencies...
A simpler thing to expose to the user is just to say we do profiles for
each minor version of Hadoop. There is a very modest maintenance cost, but we
as the maintainers of the build can bear that cost much more efficiently than
users who don't understand the nuances of our build interactions with Hadoop. I
worry we've already spent more time explaining to people why we won't add this
feature than the time cost of just adding it.
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
