iterative
ref: `dvc data status`
Report
See https://github.com/iterative/dvc/pull/7943.
@skshetry Let's start on this ASAP so that if there are concerns about the UI, they can be raised early.
That would be neat! You can make it a draft PR for now and we can have product/ QA discussions on top of it. Thanks!
@jorgeorpinel Saugat is out but the core PR should be nearly ready to merge. If you have time, please play around with it or even just check out the comments in the PR to see some example output.
Edit: Sorry, looks like you are a step ahead of me 😄
After looking into it I'm not too worried about docs. Mainly we need the new command reference and to review existing mentions of dvc status to decide which ones to replace. As for the reference, probably the more complicated part will be to decide which useful examples to include (since there are so many possible combinations of flags). On the long run the issue I see is the overlap with dvc status but that's more of a product question.
On the long run the issue I see is the overlap with
dvc statusbut that's more of a product question.
Even for the command ref for the initial PR, this is probably the biggest question that we need to address. My take is that we recommend dvc data status as a better option whenever you are focused on the status of the data and not the pipeline stages. dvc status is still needed if you have want the status of those pipeline stages (we had plans to have a dvc stage status command or similar but it's nor prioritized right now).
Wrt the command namespaces, since the basic commands like add and push relate more to data management, it would make sense for dvc status to give the new data status IMO. And to move the existing pipeline status command into dvc stage status. But that could break backward compatibility so it's probably tricky... Anyway, just a thought.
Yeah, maybe for 3.0. Ideally, we would have an improved dvc stage status and can deprecate the existing dvc status.