2023-06-14 23:40:42 +00:00
|
|
|
It would be useful if there was a way to make git-annex stop transferring
|
|
|
|
|
after the current files finish being transferred. Sometimes I am
|
|
|
|
|
moving files to a remote, and I might not pick the same remote to move them
|
|
|
|
|
to next time, so interrupting a transfer unfinished can result in wasted
|
|
|
|
|
work and space used temporarily storing partially transferred files.
|
|
|
|
|
|
|
|
|
|
This should not be ctrl-c, because it could take a very long time for the
|
|
|
|
|
current transfer to finish. Maybe SIGUSR1 or something like that?
|
|
|
|
|
--[[Joey]]
|
2023-06-14 23:53:55 +00:00
|
|
|
|
|
|
|
|
> An alternative approach, which avoids the complexity of needing to send a
|
|
|
|
|
> signal, would be for git-annex to remember what transfers got
|
|
|
|
|
> interrupted, and provide a simple command to resume them. --[[Joey]]
|
2023-06-23 20:31:54 +00:00
|
|
|
>
|
|
|
|
|
> > That could not be something like the (existing)
|
|
|
|
|
> > `git-annex get --incomplete`, because for that you have to remember
|
|
|
|
|
> > you were running a get and run effectively the same command again.
|
|
|
|
|
> >
|
|
|
|
|
> > For an interrupted `git-annex move --to foo`, it would not do to need
|
|
|
|
|
> > to remember what remote a file was being sent to. The goal would be
|
|
|
|
|
> > something like `git-annex resume` that remembers for you.
|
|
|
|
|
> >
|
|
|
|
|
> > For this, the transfer logs would need to be extended with the operation
|
|
|
|
|
> > and remote.
|
|
|
|
|
|
|
|
|
|
[[!meta title="resume interrupted move/copy"]]
|
|
|
|
|
[[!tag confirmed]]
|
