Skip to content

Cache excess, non-requested content #82

Description

@justinquinnb

The cursor + offset approach lends itself to retrieving more content than is immediately sometimes. Encourage plugin developers to return all content retrieved (even if it's above the target amount) in the stream. As part of the fetchContent response, include a int offsetEndsAt value indicating where the first piece of content to return to the client lies in the stream. Any content prior to it will be cached, but not sent to the client.

Requires going back to a custom return type for fetchContent with the flux existing alongside that helper value.

Importantly this assumes that Providers are always capable of returning the stream of content in chronological order, which may not always be the case? Research this more. If that's not a reasonable expectation, then this concept fails because there's no way to predict where the cutoff lands.

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    Status
    Under Consideration

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions