Closed hohonuuli closed 2 years ago
For chunked responses, I'll need to create separate API classes that do not mix in ContentEncodingSupport
. ContentEncodingSupport adds GZIP to responses, so chunked response don't actually get chunked, they are accumulated and then zipped when the response completes.
The vars-annotation can attempt open videos that have thousands of annotations. Internally, it uses paging to manage the request. However, there is a performance issue with this. Need to find source of slow down and perhaps add streaming/chunked request support in Annosaurus