Add pallet-prune and pallet-sync commands #50
Conversation
| @@ -237,6 +238,32 @@ use `pallet--package-archives-copy' if USE-COPY is true." | |||
| (message "Pallet: unpacking %s" package-name) | |||
| (pallet--unpack-one package-name)))) | |||
|
|
|||
| ;; Command to remove unrequested packages | |||
| (defun pallet-auto-remove-packages () | |||
There was a problem hiding this comment.
On second thoughts: for the sake of argument let's say we keep this (or Cask gives us a proxy for it), and it just uninstalls surplus packages. In that case I think pallet-prune is a good name for the function (in harmony with e.g. npm prune.)
There was a problem hiding this comment.
Yeah, I like "prune" better because "sync" doesn't tell what is being synced with what (i.e. it could just as easily do what pallet-init does).
|
FYI it looks like the Cask devs might want to merge this into Cask, so we should probably hold off until they make a decision: cask/cask#320 (comment) |
|
OK, sure. I think there might still be a use case for a |
|
It turns out |
DarwinAwardWinner
changed the title
|
Ryan, given cask/cask#320 (comment), are we parking this for now? |
|
Well, if you want to, you could merge it for now with a note to update it later to use the cask functionality when it becomes available, but otherwise, yes, I'd say shelve it. |
|
Hey I'm finally getting some time to look at Pallet issues ... are you still interested in this functionality? If so would you be happy to write tests and document the feature? |
|
My current Emacs config isn't using Pallet any more, so I no longer have a personal stake in this. But I'm willing to do at least some work to properly test and document it if you think it's a useful feature. |
|
It might still be worth the effort of getting this functionality into Cask itself, as I suggested in the linked issue. But I don't know if I'll have the time to do that. And I haven't kept up with recent Cask development, so maybe something similar has already been implemented since then. |
It turns out cask already seems to have a recursive dependency walker, so I just used that instead.
This should fix #49.