- From: Samuel Hurst <samuelh@rd.bbc.co.uk>
- Date: Fri, 2 Dec 2022 17:01:00 +0000
- To: ietf-http-wg@w3.org
On 02/12/2022 11:47, Lucas Pardue wrote: > While MICE might require that a tree of parts constructed over a > complete resource, I'm not sure if Sam's use case requires that > property. Instead a similar but simpler "progressive integrity" > content encoding (PrICE), where the content self-describes part > boundaries and their digests without chaining, might fit the need. The > benefit of including this type of thing inside the content encoding is > that it can pass across intermediaries and HTTP versions without > having to worry about re-encoding chunks or adding new frames. This is a really good point. And to address another comment made in reference to my use case and how you wouldn't use this for live streaming, we as a broadcaster may use 3rd parties to help deliver our content. Some of these 3rd parties may be CDNs that we have a commercial agreement with, others may be just consuming our content (manifests, media segments) and offering it via their own system, or perhaps being distributed by means outside of a traditional HTTP connection [1][2]. The authenticity can travel as metadata along all these chains, be it as a HTTP Content-Digest and Signature pair, or some other mechanism. In this age of deepfakes et al, the ability to prove the authenticity of media objects that are notionally coming from $Broadcaster$ actually are coming from them is seen as a valuable commodity. And at the same time, we're under pressure to reduce the latency incurred by adaptive media streaming, so we really need a solution that can do both. Insert joke about the PrICE being right here... Best regards, -Sam [1]: https://www.etsi.org/deliver/etsi_ts/103700_103799/103769/01.01.01_60/ts_103769v010101p.pdf [2]: https://dvb.org/?standard=native-ip-broadcasting
Received on Friday, 2 December 2022 17:01:14 UTC