[{"author": "Victor Pascual", "text": "
I assume more than one solution means 2?
", "time": "2024-03-19T04:13:05Z"}, {"author": "John Preu\u00df Mattsson", "text": "More than one solution very likely means B and C
", "time": "2024-03-19T04:25:49Z"}, {"author": "J\u00f6rg Deutschmann", "text": "quiche implementation was done by Ana Custura :+1:
", "time": "2024-03-19T04:30:40Z"}, {"author": "J\u00f6rg Deutschmann", "text": "picoquic implementation is here
\nhttps://github.com/hfstco/picoquic/tree/careful_resume
\ndone by Matthias H.
My problem is that these state transitions assume a RENO-like congestion control. This would not necessarily match other CC algorithms, like BBR.
", "time": "2024-03-19T04:33:33Z"}, {"author": "Christian Huitema", "text": "But really, the specification should be in the scope of the congestion algorithm, not generic for the transport.
", "time": "2024-03-19T04:34:11Z"}, {"author": "J\u00f6rg Deutschmann", "text": "@Christian Could you open a issue on the tsvwg GitHub to discuss this?
", "time": "2024-03-19T04:34:35Z"}, {"author": "J\u00f6rg Deutschmann", "text": "https://github.com/tsvwg/careful-resume
", "time": "2024-03-19T04:36:43Z"}, {"author": "Jason Livingood", "text": "How might this be affected when streamers move to less bursty & more consistent chunk downloads? I think there was an IRTF paper at 118 where Netflix was experimenting with that.
", "time": "2024-03-19T04:40:59Z"}, {"author": "Eduard V", "text": "I did believe that \"pacing\" was already widely adopted. Hence, no bursts.
", "time": "2024-03-19T04:45:26Z"}, {"author": "Tom Herbert", "text": "DSCP-like is a good something we've seen in several proposals liek this. IMO, this motivates a common solution
", "time": "2024-03-19T04:45:38Z"}, {"author": "Kirill Pugin", "text": "hm... tolerance to delay is not necessarily sender decision...
", "time": "2024-03-19T04:54:26Z"}, {"author": "Gorry Fairhurst", "text": "I'd be interested on understanding if we can pull-out some general principles for this type of interaction...
", "time": "2024-03-19T04:57:18Z"}, {"author": "Tom Herbert", "text": "@Gorry, IMO yes. We can describe the base requirements of host to network signaling. At the highest level we can described the carrier and content of signals hopefully independently.
", "time": "2024-03-19T05:00:16Z"}, {"author": "Tom Herbert", "text": "See draft-herbert-host2netsig for attempt to define requirements for host to network signaling
", "time": "2024-03-19T05:01:47Z"}]