I'm kind of curious why there are 3 similar commands:

run

,

el

and

elt

. It seems el & elt allow

--select

and

--exclude

to be specified where

run

does not. I saw that

elt

was being deprecated in favour of

el

and the cli docs for el mention the following:

The command meltano run is the recommended way to run cross-plugin workflows in a composable manner.

Are we moving towards having just 1 command, or does it make sense to keep both

run

&

el

. I'd love to understand when to prefer

run

over

el

& vice versa.

As I understand it the

el

is happening as part of the

run

.

meltano run

creates a run with an ID and some status about how it went etc. in Meltano's own database. I don't think

el

does that..

I think it still makes sense to keep both

run

&

el

. The latter like you mentioned supports

--select

and

--exclude

and we don't have plans for adding them to

run

.

Thanks for the response Edgar, I guess I'm still struggling to understand when one should use`run` if all of its functionality (+ some more) can be achieved via

el

When you want to track the results. Suppose you want to run the same

el

command at a certain interval, then using the

run

API that becomes a "thing" with metadata like "when did the run start/end", "were there any problems" etc.

@haleemur_ali Well, al least three reasons: 1.

run

has a different implementation and works on Windows, if that's important to you 2.

run

can run mappers, e.g.

meltano run extractor mapping loader

3.

run

can also run more complex commands: https://docs.meltano.com/reference/command-line-interface/#run