Is BulkCheck still experimental?
# spicedb
v
It is. It was released on 1.25.0 for the first time.
h
I cannot wait to try it. It seems like a good middle way between checking single permissions on one hand and looking up all resources of a subject on the other hand.
v
Yep and in fact it does optimize that latter case, making it way faster than individual checks
e
This is great. we will be testing also.
h
The bulk check api accepts a single zed token. What is the recommended way to determine that? Currently I store a zed token with every resource.
v
Thats actually a good question 🤔 I didn't account for a scenario like that, but it makes sense you may have multiple zedtokens. Even if there was a way to compare all zedtokens and use the latest, that may not be what you want and wouldn't make good use subproblems cached on prior requests with the older zedtoken.
I'd suggest opening an issue, this may require some discussion. In the meanwhile I can only suggest to use the freshest zedtoken, if that's information you have available
h
Is the order of the Pairs of the bulk check response guaranteed to be the same as the Items of the bulk check request? Or would I have to associate them myself afterwards?
v
it's not, but each bulk check response item contains a reference to the originating request item