pretty much

unless you can just keep the static nature in the name itself

thanks for confirming 🙂

really awesome project as always; i'm a big fan of all your hard work!

Hey there! I need some advice about my schema, I left comments in a test relationships tab. I would be grateful for any help https://play.authzed.com/s/IWXrzdU2W-v1/schema

Suggestion: Would be helpful to run LookupResources and LookupSubjects in the playground. Feels like it would be a great way to validate the schema you're working on, by checking permissions in both ways. What users can access this thing? What things does this user have access to?

yeah, we plan to do it

Hello! I've been playing a bit with the serverless offering. When I went to write ~10k relationships, I ran into what looked like rate-limit errors. Might have been an error-based circuit break, but it was hard to tell. While there are plenty of ways for me to lower my rate of API use for one-off tasks like this, is there any documentation for what the API rate limits are for AuthZed serverless?

We do have rate limits on the serverless system and we're in the process of documenting them (today, in fact). Right now, there is a Write QPS limit of around 5 per second, but we'll be increasing that over time as well

Haha, what timing. Thanks!

Is that per token, per client, per system, or something else?

per system

keep in mind, these are the limits for prod systems

dev systems will be lowered over time

since they are intended only for development and testing

Is that something that would be adjustable if the need ever arose for a given prod system? Or would that be venturing into Dedicated / Enterprise territory?

yes, but it is currently planned

awesome

so no firm timeline on when that will be there; if you need adjustable or higher QPS today, Dedicated / Enterprise is the solution

@Joey , is there a public backlog/github issue link for this?

apparently not; I'll create one

Hey there, Does caveat today support checking timestamp against a range ? For example, xyz resource can be access only from time 11 am to 2 pm . Now when a request to access resource is initiated at 12pm , how can the same be validated ?

yeah, we support a timestamp built-in type. You can store in your relationship the upper boundary and lower boundary, and then write the caveat expression like:

caveat working_hours(now timestamp, start timestamp, end timestamp) {
  now >= start && now <= end
}

I'm not sure 100% we have implemented the comparison operator though, but if we didn't it should be an easy addition

@vroldanbet https://github.com/authzed/spicedb-operator/blob/main/CONTRIBUTING.md#common-tasks does not explain the dev workflow, are you manually running docker build, tagging the image, pushing to your own registry, and hacking the yaml to take them image from there? I am asking since I'd like to patch in your PR tinally be able to try spicedb end to end

It's true there is no dev workflow defined, feel free to open an issue. The way I do it is to run the operator myself from my IDE, and pass the arguments to make it talk to my

kind

kubernetes cluster in my workstation.

then I'll mess with the custom resources in my kube cluster to validate behaviour, debug...

not sure if that's the right way, but it certainly works for me

will do, thanks