Is it possible to omit the need of specifying path?

[cnbuff410]

What I meant is, currently if I want to sync a folder, I would do

rclone sync source:path dest:path

But the most convenient way would be I simply go to that folder and do

rclone sync

Because remote path is always hard to remember for storage service like GCS. Is it possible? I imagine having some kind of mapping metadata saved in config file would do it.

[ncw]

We could add a path into the config relatively easily so then you'd have to type

rclone sync source: dest:

What do you think of that?

[cnbuff410]

That's fine. Just curious though, why not even ignore "source: dest:" part?

[cnbuff410]

I guess what I would like to see is to have "push" or "pull" command just like git. If I go to a rcloned folder and do "rclone push", it like "rclone sync local remote", same case for "rclone pull"

Then you may want user to do "rclone init remote" first, and record that info into rclone.config. So everytime when user use "push" or "pull", rclone could automatically figure out what's the remote.

Does it make sense?

The way I currently using is to do it using bash script. so like if I want to have a folder "A" keep sync with a remote, I need to create a folder AA, then put "A", "push.bash", "pull.bash" into AA. It's fine, it's just nicer to be able to remove the meaningless "AA" folder.

[zeshanb]

Allowing users to give unique name to each cloud storage point is very well done. Thanks for making this unique GO application allowing anyone to diversify their data storage. GO is a perfect choice to make something like this.

I can name inbound cloud storage as downloadfrom: and where I want to transport the data to uploadto: . This way there isn't any confusion during large data transactions.

rclone [goodie options] copy downloadfrom:container uploadto:bucket/folder-name

Thank you many folds.

[Animosity022]

This looks to be covered these days with the use of the alias backend.