sandstorm-io / collections-app Goto Github PK

It would be really useful to have a function when accessing a grain to add it to a collection.
This would really facilitate work flow rather than having to go to the collection adding process to add a single grain.

I love collections by the way, a really great usability feature!

Hi,

I've been looking here https://github.com/sandstorm-io/collections-app/blob/master/src/identity_map.rs to find how the grains are stored

I gathered that the ID of the grain is truncated somehow and hashed with percent encode.
Then this is stored under
CollectionGrainID/sandbox/identities/
as symlinks.

I'm looking for the reverse function, aka going from the stored ID to the grain ID. Idealy, in python, perl, bash, or else.

anyone could help with this ?

PS: basically, I'm trying to find the grains associated with a collection on the server side, to Rsync those database and back them up automatically.

I would like to be able to bulk add a bunch of grains to a Collection from the grain interface, as opposed to only in the Collections interface. Is this possible?

I think this may be related to #8 also but is a more severe issue.

Let's say user A shares a collection with user B

User B may try to rename a grain by unlinking it from a collection
and then re-linking it.

This results in them no longer having access to the grain,
they need to request access again.

This seems problematic as the intention of user A was to share the grain.

I know this is probably more of a general sandstorm feature but it would be good if collections came up the top of the list in the grain list.
Also, it may be good if this occurred in the open grains list too. On a related matter, I note that there is no way to reorder grains in the open list...

In my case, I currently have a collection with three grains:

Etherpad, write access
Etherpad, read access
EtherCalc, read access

But there's no way for me to tell, for instance, that someone shared this collection can write to one file, but only read to the other two. I "just know" that, but if I forget, I could do something silly or wrong.

It would be great to be able to include grains from different sandstorm servers into a collection. I presume this may need some changed to core?

I notice that:

• I can add a collection to itself
• I can add the same capability to a collection multiple times.

Should one or both of these be prohibited?

Prohibiting the former would probably be mostly cosmetic, assuming one doesn't also prohibit more general cycles in the containment graph (which might be a good idea but might also be annoying to detect.)

Prohibiting the latter also seems mostly cosmetic but it's hard for me to imagine a non-accidental reason to do it.

entering search terms returns no matches even though they should match

ps. great to see this app, very useful!

After installing the app from the app-store and creating a collection, nothing happens when I click on "Add grain...".

Currently Sandstorm erroneously ignores this title, and actually assigns no title. In some contexts, the upstream title ends up being used instead, but in others nothing is displayed.

I'm fixing the Sandstorm end, but the collections app needs a fix as well.

When a grain is renamed (in my case it was a gitlab grain), the name change is not updated in the collection (it still shows the old grain name).
I have tried closing and reopening the grains (and collection).