Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-12-11 Thread Jarek Potiuk
+1 to Daniels's suggestion. On Tue, Nov 28, 2023 at 5:25 AM Amogh Desai wrote: > Agree with @Daniel Standish on this. > We should rename get_uri to reduce confusion. > > I am fine with renaming to to_uri_repr and then deprecating get_uri > > > Thanks & Regards, > Amogh Desai > > On Mon, Nov 27,

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-27 Thread Amogh Desai
Agree with @Daniel Standish on this. We should rename get_uri to reduce confusion. I am fine with renaming to to_uri_repr and then deprecating get_uri Thanks & Regards, Amogh Desai On Mon, Nov 27, 2023 at 8:56 PM Andrey Anshin wrote: > > Recently, a method was added `to_json_dict` which give

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-27 Thread Andrey Anshin
> Recently, a method was added `to_json_dict` which gives the json repr. This method is internal method which use in 1. Connection Serialisation, and it is a replacement *to_dict* method which was used before https://github.com/apache/airflow/pull/35723 2. Use in public method *as_json* Personall

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-27 Thread Daniel Standish
Just some history ... and a renaming suggestion... When get_uri was added, generating the connection URI was a real pain, and Airflow provided no way to generate one, and JSON repr was not yet implemented. By adding get_uri, a user could define the connection object we needed in python, and airf

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-26 Thread Tzu-ping Chung
Some context: I’ve been looking into putting the Connection URI to more use, most importantly to consolidate it with OpenLineage’s URI format. Skipping the details, one of the conclusions I had is get_uri on Connection is not actually that useful, due to how different libraries (SQLAlchemy, Airf

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-26 Thread Amogh Desai
A little late to the party here but convinced by what Daniel had to say. I am ok with the idea of adding this into the deprecation bucket in the coming time. Implementing the get_uri in the hook seems quite reasonable to me on the grounds that the Connection class' get_uri is meant to be general,

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-20 Thread Daniel Standish
> > I still think different people have different expectations of the Airflow > URI format. My expectation is that this is only a SerDe mechanism around > Airflow Connection. If we have a look at the codebase of Airflow we found > that this is not always true and we try to use it as SA URI. Yes,

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-19 Thread Amogh Desai
Hey All, Very detailed discussion, Andrey and others. I am hearing everyone out here and I kinda liked the idea from Jarek about prefixing with "airflowconnections+". I share everyone's problem here because I have been trying to get Hive HA connection type work which is of the format: jdbc:hive2:

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread Andrey Anshin
Hey guys! Thanks for all your feedback and suggestions. I see here couple of examples with postgresql, however I would like to remind that postgresql it is not a valid connection type which doesn't register into the providers manager, and we have an exclusive hack to turn this connection type into

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread utkarsh sharma
I do think the URI format for airflow connections also serves the purpose of conveying the intent, similar to how we URI in a browser to connect to a website. We use Ariflow Connection URI to connect to one of the services supported by Airflow. Due to this reason, I lean toward Jarek's proposal of

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread cary hawkins
I would also be a -1 for deprecating URI connection representation, this is one of the defacto standards across Python. I do agree with the suggestion from Jarek that it could be less confusing if it follows SQLAlchemy style formatting. That is a known format that users are likely going to be fa

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread Bolke de Bruin
So -1 on deprecating. I think we should do the opposite and go all in. The suggestion by Jarek makes sense in this case, as it remains and is fully standards compliant. The JSON representation isn't: it is our own crafted format. My 2 cents, Bolke On Sat, 18 Nov 2023 at 22:15, Jarek Potiuk wrot

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread Jarek Potiuk
I agree with Ash that it's much easier to define a lot of simpler connections this way. But it is also very confusing the way now how you can mix the "real" url with "airflow connection" URL. And Daniel is very right about the magnitude of breaking change. But possibly there is a way to eat cake

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread Daniel Standish
The thing that makes *me* hesitant to deprecate is the sheer magnitude of breaking it would bring (even though we're only talking about a hypothetical 3.0 release), balanced against the actual pain it causes. I.e. it's confusing to use, and takes up space in docs (when, if removed, we could just sh

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-18 Thread Ash Berlin-Taylor
-1 — for the simple case the URI format is perfect: AIRFLOW_CONN_MY_DB=postgresql://user:pass@host We can push JSON format for more complex cases, and/or limit URI to only the simple cases as long as we don’t remove it entirely. -ash > On 17 Nov 2023, at 22:44, Andrey Anshin wrote: > > Greet

Re: [PROPOSAL] Deprecate URI Connection representation in favor of JSON

2023-11-17 Thread Daniel Standish
I would also like to see it deprecated. That said, I am not convinced there is anything that we cannot encode using URI though. I think the problem is just when one tries to use the same URI to mean two different things, e.g. both airflow connection URI and sqlalchemy URI. They are different. H