Recent searches
Search options
Apparently HTTP/3 is going to be a thing in the near future.
And like HTTP/2, it's built on protocols designed by Google.
Meh.
@zatnosk And Google is a surveillance capitalist. It’s like having medical protocols designed by Philip Morris. The protocol is political and the infection runs deep.
@aral Exactly. While the technical benefits from the protocols Google design are attracting, I don't know what the weaknesses are - and I definitely don't have the resources to investigate.
Will HTTP/3 aka HTTP/QUIC make it easier for Google to track anything they want? Will the complexity of the protocol lead to implementer monopoly, giving them control of all of our network usage?
I don't know, I don't know how to find out, and I don't want to risk it.
@zatnosk @aral is it being standardized by the #IETF? There was a time I would trust without question that IETF processes keep protocol development vendor neutral. Now, after the whole EME fiasco at the W3C I'm not so confident. But I'll still trust the IETF until presented with evidence that it's compromised.
@strypey @aral Yes, they have a working group on QUIC:
https://quicwg.org/
And Mark Nottingham (chair of HTTPWG and QUICWG) has requested that HTTP over QUIC be renamed to HTTP/3.
https://mailarchive.ietf.org/arch/msg/quic/RLRs4nB1lwFCZ_7k0iuz0ZBa35s
@zatnosk @aral IMO the main risk posed by HTTP/3 - which is/was also a problem with HTTP/2 - is that of complexity.
It's getting harder and harder to implement a web server or web browser from scratch.
Just like the constant ratcheting up of security requirements, this puts pressure on and squeezes out the little guys.
This impacts me (a little guy), because https://pagekite.net interacts with HTTP(S) at a low level. I haven't had the cycles to implement support for /2 and /3.
Beatrice Martini and Niels ten Oever wrote a report on this:
"Will HTTP/3 aka HTTP/QUIC make it easier for Google to track anything they want?"
If anything it will make it harder, the QUIC working group seriously advocates for increased user privacy
"Will the complexity of the protocol lead to implementer monopoly, giving them control of all of our network usage?"
Nah, there are many implementations, most of which are open source.
QUIC is the next thing because people are behind it, if it were just google it would be the 'chrome' protocol
𝚜𝚎𝚕𝚎𝚊 
@zatnosk @aral When it comes to implementor monopolies, I'm more concerned elsewhere in the webstandards.
There's just so many! And because of that it's unlikely for any browsers other than Edge, Firefox, Safari, or (do they belong in this list?) Chrome will ever be developed. Anything else including my work is just a repackaging of their software. With plenty of features added around the edges.