These are chat archives for CommBank/maestro

15th
Apr 2015
Stephan Hoermann
@stephanh
Apr 15 2015 00:05
@SamRoberts yeah, collections, etc. are something we want to support we just had or have to wait for parquet to be further along.
Stephan Hoermann
@stephanh
Apr 15 2015 00:23
@samroberts can you look at the ebenezer issue as priority.
Sam Roberts
@SamRoberts
Apr 15 2015 00:25
yep
Todd Owen
@toddmowen
Apr 15 2015 01:18
Oops, merging PR #363 into master via the “Merge” button on github gave a pretty ugly result in the git log. Merge bubble, right?
Sam Roberts
@SamRoberts
Apr 15 2015 01:18
yeah, don't do that :smile:
Todd Owen
@toddmowen
Apr 15 2015 01:19
@SamRoberts is it worth doing a forced push to fix it?
Sam Roberts
@SamRoberts
Apr 15 2015 01:19
I would generally:
git rebase -i origin/master
# squash all commits, or do whatever makes sense
git push -f
git co master
# update master to your branch, e.g. merge --ff-only
git push
pushing your branch and then pushing the same commit to master means git will automatically close your PR
um, I think it would be nice if you did it now
not a huge problem if you don't: we've had people making the same mistake in the past
Todd Owen
@toddmowen
Apr 15 2015 01:34
Cheers, I tidied it up. Had to make sure nobody would yell at me for suggesting a forced push to master.
Stephan Hoermann
@stephanh
Apr 15 2015 01:36
Force pushs are bad because they screw up the commit history for others. If you do it really quickly you won't affect anyone. :smile:
Sam Roberts
@SamRoberts
Apr 15 2015 01:37
sure. It's not ideal, obviously. Not sure off the top of my head, but we may produce maestro artifacts on artifactory which are tied to your initial commit, which after a while should no longer exist in our repo. But you'll have created another artifact which is "later" than that one to replace it in a few minutes, and people take the latest version they see, so it should be ok
Sam Roberts
@SamRoberts
Apr 15 2015 02:58
can someone please review CommBank/ebenezer#98 ?
I expect this to fix the bug that @snehalvidyan reported yesterday
Rowan Davies
@rowandavies
Apr 15 2015 03:00
@shmookey (as discussed) You should be able to add tupled to Function1 by creating a RIchFunction1 similar to: https://github.com/CommBank/maestro/blob/master/maestro-core/src/main/scala/au/com/cba/omnia/maestro/core/scalding/RichExecution.scala
But, this is likely overall at least as messy as generating different code for Function1.
Luke Williams
@shmookey
Apr 15 2015 03:01
:)
thanks
Jithin John
@jithinjohn
Apr 15 2015 05:58
@stephanh @SamRoberts Can i update util.etl. with latest maestro version to start using #360
Sam Roberts
@SamRoberts
Apr 15 2015 05:59
you should always be able to update util.etl to use the latest maestro version
Jithin John
@jithinjohn
Apr 15 2015 05:59
:) thanks
Sam Roberts
@SamRoberts
Apr 15 2015 07:01
ok, I have dealt with the compat project in CommBank/ebenezer#97, could someone please review it again?
it is now a rather large PR