3 ms·
The sane development path would have been to build QUIC as a thin layer on top of SCTP, with a fallback to UDP implemented only if it proved necessary. I don't
by wtallis 8y ago
The sane development path would have been to build QUIC as a thin layer on top of SCTP, with a fallback to UDP implemented only if it proved necessary. I don't think the middlebox problem is a showstopper for a limited-audience experimental protocol like QUIC, and a successful QUIC-over-SCTP experiment could have gotten rid of the middlebox worries (by either motivating people to get rid of the broken middleboxes, or by proving they weren't common enough to care about), in much the same way that Apple's limited deployment of ECN has eliminated the worries about middleboxes that kill ECN flows.
As for WebSockets, WebRTC, and WebAssembly, the fact that they've been standardized within their insular niche says nothing about the extent to which they duplicate functionality that has long existed outside the browser as part of the host OS or as libraries that use the low-level networking constructs that any native application has access to (but JavaScript didn't).
- petters 8y ago> limited-audience experimental protocol like QUIC That is not true. QUIC is currently used by a large portion of all traffic on the internet. How? Precisely because it uses UDP, which is supported everywhere (and that Google controls both Chrome and YouTube).
- wtallis 8y agoGoing from Google's servers to home and cellular users is a limited audience in this context, because it avoids all the legacy middleboxes that tend to plague corporate networks. It's the same effect that's responsible for IPv6 traffic being higher on weekends. Consumers may not be competent at upgrading or administering their network equipment, but at least they throw it away every few years and buy something newer (or their ISP swaps out the rental modem+router box).
- dagenix 8y agoPeople use Chrome at work too.
- petters 8y agoTraffic from people's homes and cell phones to Google really isn't a limited audience.
- wtallis 8y agoI explained the sense in which it is limited compared to the entire Internet. Do you have any specific refutation of my explanation, or are you just rejecting the general idea that anything involving Google could be regarded as not all-encompassingly huge?
- dagenix 8y agoThe "sane" thing to do is to complicate the protocol by building it on top of both SCTP _and_ UDP? How is it the "sane" thing to do to make the protocol more complicated with a fallback that easily allows the middleboxes to just ignore the SCTP option knowing that UDP will work? That wouldn't have accomplished anything, except re-affirming that middleboxes tend to reject SCTP at rates that make it unrealistic to adopt it at scale. What they did accomplish, however, is getting an open protocol to the IETF for adoption that works on the actual internet. I fail to see how that is some sort of conspiracy or incompetence. And that isn't even taking into account that building on top of SCTP would have made some of QUIC's goals hard / impossible such as eliminating the penalty for establishing encrypted connections - something which benefits the internet at large. What open protocols / standards duplicate WebSockets, WebRTC, or WebAssembly? TCP is kinda like WebSockets - but, giving any random JavaScript the ability to open TCP connections is a bad idea - nevermind that WebSockets provides are variety of useful abstractions such as messages. I'm not sure what you are saying duplicates WebRTC - there are a bunch of proprietary protocols, I guess, but, I'm not aware of anything open that its competing with. As for WebAssembly, it explicitly works with JavaScript providing significant efficiencies for certain types of code. Google, Microsoft, etc have their faults. But, picking out instances of when they have worked together, in the open, to create free, open standards that benefit everyone to complain about is just super bizarre.
- wtallis 8y ago> What open protocols / standards duplicate WebSockets, WebRTC, or WebAssembly? Those aren't duplicating open protocols, they're just re-inventing the operating system inside the web browser, but with more overhead and fewer features and generally not even working as well as Java. The Web-* standards just make it possible to do things inside the browser that are already possible outside the browser, and are a bad idea to do inside the browser. Their only advantage is that it makes it a bit easier for the developer to deliver the "application", but it usually also results in the developer leaving out a lot of functionality and refinement that a native application would have.
- dagenix 8y agoWebRTC isn't even working as well as Java? I guess thats hard to argue with because it makes no sense.