hey all, if we wanted to add a stream that does not have a json file in the schemas directory, do we have to update the source code for the singer tap?

very tap dependent. Some taps have options to dynamically load streams based on config files, or dynamic things. like https://github.com/ets/tap-spreadsheets-anywhere A lot of taps written by the singer-sdk maybe the one you're referring to? Most are based on fairly static api's so it's much easier and faster to just throw a schema file and stream configuration in the tap and most of the time that's most practical as most api's don't have discovery mechanisms. The "core" question here is probably getting at why can't everything be dynamic? Well it can and people have tried https://github.com/Widen/tap-rest-api-msdk is a good example. Sometimes this works for some usecases well, in complicated use cases this gets hard. For this type of tap you push a lot of "code" to configuration which some folks want

no worries! this is actually for the hellobaton one - https://github.com/dluftspring/tap-hellobaton i was thinking of adding a json file to the schema

Add it 😄

thanks!

Hope I was actually helpful haha, good luck @kevin

you were, ill read more into the dynamic loading info as well

I lean away from the dynamic side but it's really dependent, and almost always a balance

Hey @visch! one quick question to this, ive updated the source code to include the field i want. Anything to do on the meltano side to "pill" the latest source code? i reran the jobs and the destination tables in snowflake do not reflect the change

Can you open a new thread so others can help. I'm not sure how to help without more info either! Sorry 😄