Github user douglaz commented on the pull request:
https://github.com/apache/spark/pull/813#issuecomment-43579830
Yes, the user can instantiate the RDD and yes this is inconvenient. An
interface to do this would be no less inconvenient if it has the same drawbacks
(that you need to explicitly convert back the resulting sequences to the
original type).
Limiting the user to 3 cogroups is pretty much like limiting tuples to 3
elements. You may have technical reasons for that limit, but it isn't
reasonable for practical purposes. You can't just say: if you need a tuple with
more than 3 elements, use lists instead.
For tuples the current limit is 22, which is "enough for everyone". For
cogroups the limit should be lower, but certainly above 3.
---
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.
---
