Discussion about private vs public datasets and the impact of using dots in document IDs

Any ideas why my query is returning no results when called from a google script, but returns results otherwise?

One thing that I could think of is if the dataset is set to private.

You can change between public (everyone can read) or private (need to be authed or have token to read)

It's definitely public, you can put that url in your address bar and it works

Nope, not for me

You probably have a token stored in your browser

oh you're right.. it doesn't work in a private window

didn't realise it would use your logged in session

but yeah..

Weird, I tested to set one of my datasets to private and the kept the other public and got the same result as you describe.. unfortunately that was my only idea 😕

Are your documents published? Do the

_id

attributes contain a

.

by any chance?

not sure what it means for a document to be published.. I've only been using the data lake so far

and the ids do have `.`s

I'll have a read up on document publishing

Ah. Common gotcha - https://www.sanity.io/docs/ids
Dots are treated as "paths", and the default security rules says that anything not within the "root path" is not public to unauthenticated users

So

foo.bar

is "private", while

foobar

is public. This is the way drafts work when you use the studio: they are documents with an ID prefix:

drafts.someDocumentId

ah, I think I read that page at some point but it didn't sink in. I thought using paths in my ids was just going to give me more flexibility down the line

no big deal, token auth it is

thanks for the help!

no big deal, token auth it is

Weird, I tested to set one of my datasets to private and the kept the other public and got the same result as you describe.. unfortunately that was my only idea 😕

Are your documents published? Do the

_id

attributes contain a

.

by any chance?

and the ids do have `.`s

not sure what it means for a document to be published.. I've only been using the data lake so far

I'll have a read up on document publishing

Ah. Common gotcha - https://www.sanity.io/docs/ids
Dots are treated as "paths", and the default security rules says that anything not within the "root path" is not public to unauthenticated users

So

foo.bar

is "private", while

foobar

is public. This is the way drafts work when you use the studio: they are documents with an ID prefix:

drafts.someDocumentId

ah, I think I read that page at some point but it didn't sink in. I thought using paths in my ids was just going to give me more flexibility down the line

Happy to help!