Re: Operator for syncing s3 buckets (and recursively copying)?

2019-06-05 Thread Austin Bennett
Thanks, airflowuser, Would love to get feedback on AIRFLOW-4736 and AIRFLOW-473 5. Specifically, anything that anyone would suggest as conventions/issues ahead of submission. I haven't found

Re: [2.0 spring cleaning] Changes in import paths

2019-06-05 Thread Philippe Gagnon
Hello community, Here is my take on this: Case #1: Solution B, but I would like for us to consider putting incubating modules in a separate repository instead of vendoring them with Airflow itself. There are pros and cons to this approach. On one hand, I think that we need to take responsibilit

Re: [2.0 spring cleaning] Changes in import paths

2019-06-05 Thread Kamil BreguĊ‚a
I have done a deep investigation of this issue. Please, look at the beginning of chapter "Case #7". I prepared a repository with simple demo and also shared a screenshot from my IDE. Photo: https://imgur.com/a/mRaWpQm Repo: https://github.com/mik-laj/airflow-deprecation-sample Today I looked again

Re: [2.0 spring cleaning] Changes in import paths

2019-06-05 Thread Ash Berlin-Taylor
> Case 7: Zope deprecation or not > *B*: I prefer to use Kamil's in-house hack than zope, but I feel we could > improve it slightly to remove a bit of boilerplate - it looks like the > warning.warn() message should be very similar in all deprecation cases and > possibly we can define it in one plac

Re: [2.0 spring cleaning] Changes in import paths

2019-06-05 Thread Jarek Potiuk
I think it's very important to discuss this and choose the right approach and follow up quickly. This is very important to introduce some sanity in the structure of 2.0.0. I propose that anyone with a strong opinion voices his or her concerns here and then maybe a week from now we can get to a cons