Re: [Moq] Charter adjustment

[Maxim Sharabayko <[e-mail]>]

HESP looks extremely interesting! Thank you Bernard for sharing!

It seems to be a re-thinking of LL-HLS with low-latency in mind.
One of the main outstanding points is low stream initialisation and channel switching time.

Ian said: “This begs the question of "Why is MoQ better?"

HESP is more a protocol to deliver (distribute) a content to an end-user, and does not cover ingest well enough.
While MOQ “will develop simple low-latency media delivery solution for ingest and distribution.”


-- Maxim

From: Moq <[e-mail]> on behalf of Christian Huitema <[e-mail]>
Date: Friday, 22. July 2022 at 08:05
To: Bernard Aboba <[e-mail]>
Cc: Ian Swett <ianswett=[e-mail]>, Suhas Nandakumar <[e-mail]>, Lucas Pardue <[e-mail]>, Mo Zanaty (mzanaty) <mzanaty=[e-mail]>, Ali C. Begen <ali.begen=[e-mail]>, Ted Hardie <[e-mail]>, MOQ Mailing List <[e-mail]>
Subject: [EXTERNAL] Re: [Moq] Charter adjustment

Thanks for the pointer to the HESP spec, Bernard. MoQ and HESP are very different approaches. I think that HESP strict adherence to HTTP implies risks of head-of-line blocking. Probably not an issue if low latency is defined as "under 400ms", but certainly important if we aim for "under 200ms" or "under 150ms".

-- Christian Huitema
On 7/21/2022 8:33 PM, Bernard Aboba wrote:
https://datatracker.ietf.org/doc/draft-theo-hesp

On Thu, Jul 21, 2022, 7:23 PM Christian Huitema <[e-mail]<mailto:[e-mail]>> wrote:

On 7/21/2022 7:13 PM, Bernard Aboba wrote:
Ian said:

"HTTPS proxies have been around for a long time, and MoQ shares many of those properties."

[BA] Indeed, there are proposals (such as HESP<https://www.hespalliance.org/>) aimed at low-latency distribution that run over HTTP/3 and support caching.

This begs the question of "Why is MoQ better?"



If you want low latency, you need to support datagrams, and have some way to fragment long frames over series of datagrams. But then, if you require reassembly of fragmented frames at each step in the relaying chain, you can easily run into head-of-line blocking. That's something we can handle in MoQ. I don't know whether HESP does that -- the specs do not appear to be publicly available.

-- Christian Huitema