Thiago Bellini Ribeiro

Marking this as `documentation` since it would be nice to add some documentation about how this works

Hi @Mapiarz , I get what you mean. Indeed the current implementation, and specially the optimizer, can either be used with the default limit/offset pagination, which is pretty basic and...

@Mapiarz > I have to disagree, that's not the reason why the cursor has to be opaque. In short, it's so that clients can be decoupled from implementation details of...

@salcedo , > 1. Documentation isn't in line with reality. I'm spending more time reading code and at this point, adding logging at various places to better understand what it's...

> I just checked the lib out and it doesn't suggest anywhere that using non opaque cursor is the way to achieve 'jump to page'. So nothing to complain there....

Hey @MaehMaeh , This is actually the same issue as the one described here: https://github.com/strawberry-graphql/strawberry/issues/2893 We still need to find a better way of handling unions for connections when using...

Hi @humphrey , That is indeed a problem without a solution atm. What I do in my projects is to `cast` the return value (i.e. `return cast(UserPreferencesType, UserPreferences())` in your...

> I assume that explains why "self" is actually the Django model instance in the following code. I've been working around this by specifying the model as the type. PyCharm...

> instantiating TypeUser, I need to fill all of that, including chats, if not, it will throw an error. Now, I am not sure if this is the right way...

> @bellini666 the specific use case for me is to describe a session object, it's not really derived from a django model but it's more of a computed object where...