Feature/dataset based compilers #917
Conversation
GeorgesLorre
force-pushed
the
feature/dataset-based-compilers
branch
from
ec1c8b6
to
fef8653
Compare
GeorgesLorre
force-pushed
the
feature/dataset-based-compilers
branch
from
fef8653
to
922ab4e
Compare
| @@ -857,4 +858,4 @@ def write( | |||
| resources=resources, | |||
| cache=cache, | |||
| ) | |||
| self._apply(operation) | |||
| return self._apply(operation) | |||
There was a problem hiding this comment.
The only difference to the .apply method is that the write method does ignore the produces.
I would leave it like this but what do you think about adding a deprecation warning to the write method?
We could use this method in the future export the dataset from the working directory to a remote storage for instance.
There was a problem hiding this comment.
I'm not sure I completely understand. Are we deprecating the write ?
There was a problem hiding this comment.
Thats the question if we want to do it. The apply and write looking quite similar now.
There was a problem hiding this comment.
Thanks @GeorgesLorre! I like the workflow wording instead of pipeline in the documentation. Maybe we can use this more often.
We could explain these concepts somewhere else.
E.g.: A dataset can be materialized by running a workflow. The workflow has to be compiled first.
Co-authored-by: Matthias Richter <[email protected]>
|
@mrchtr I updated the docs based on your suggestions |
GeorgesLorre
merged commit 4ac7a0b
into
feature/refactore-pipeline-interface
Co-authored-by: Matthias Richter <[email protected]>
Co-authored-by: Matthias Richter <[email protected]>
No description provided.