What are the most resource onerous parts of Cloudflare TV? What are the ways in which the resource demands can be alleviated?
Are there changes independent of the network that can optimise the livestream (eg codecs)? Are there any explicit efforts made to remediate the poor connection of some participants in the livestream? This situation happened recently when Matthew Prince interviewed the Atlassian cofounder. There were at least six disconnections. Does the participant get notified somehow?
At which PoP do the streams get mixed together to produce a livestream? Is a middle point between both geographic locations chosen? Or does it depend on the status of the network?
A lot more questions but unfortunately that post addresses very little from a technical standpoint. If it is celebrated as a technical feat, there should at least be some technical information.
The article points to this older post that has more technical info (https://blog.cloudflare.com/building-cloudflare-tv-from-scra...). But the post is nearly a year old. It would have been more interesting to see a technical lessons learned post in comparison with that older original post. Hopefully, that is what is in store and hinted at.
Thank you for all the efforts you put in towards keeping the standards of conversation high. I am a long time reader even though a first time contributor. I have seen your thoughtful measured nudges for better quality on more than one occasion.
If my comments ever fall short, I would welcome your messages. Thanks again.