6 ms·
Far from it. It's quite likely going to be over 5 years if not a decade before it would be possible to run a pure-HTTP3 service without risking connectivity pro
by shockinglytrue 6y ago
Far from it. It's quite likely going to be over 5 years if not a decade before it would be possible to run a pure-HTTP3 service without risking connectivity problems
The problem is similar to the IPv6 transition, except thanks to the browser monopolies, it's possible at least for network providers to quickly feel significant pressure to fix their networks. But there will always be some networks that will never be fixed
edit: for those inexplicably downvoting this, please pay attention to the parent comment's question, and the Internet's long chequered history of adopting new protocols in any setting. TCP port 443 isn't going to magically disappear overnight, or indeed any time soon. This is evidently true because it has been true for all prior transitions. Mail still flows to many places unencrypted despite the standardization of STARTTLS 21 years ago. The long tail has only gotten much longer in those intervening 21 years.
- cryptonector 6y agoHTTP/2 and HTTP/3 do not change the semantics of HTTP. That means you can run reverse proxies to server HTTP/1 services as /2 and/or /3 and vice-versa. As a result the transition will be a lot easier than the transition to IPv6. I expect that the transition in corporate networks will be faster -- the opposite of the IPv6 case -- because there is a lot of appeal to HTTP/3.
- the_duke 6y ago> because there is a lot of appeal to HTTP/3 I thought the primary appeal of HTTP/3 is for mobile clients, and bad connections in general, because it circumvents TCP head of line blocking and connections can persist across networks. That doesn't feel terribly relevant in corporate networks. (not disputing that it's not comparable to the v6 transition)
- cptskippy 6y agoHis claim of corporate appeal is completely unsubstantiated and I think when he says corporate he isn't referring to enterprise. Enterprises largely won't give two shits about HTTP/3. Just last week I took ownership of another department's decade old app written in VB.NET WinForms. The former dev team was putting the finishing touches on the C# WebForms refactor. It's been interesting taking a step back in time to Dev practices from 2008.
- cryptonector 6y agoThe appeal of H3 is smaller memory footprint.
- user5994461 6y agoYou realize that HTTP/2 is still nowhere near to being adopted by corporations? It's really far fetched to plan HTTP/3 and expect any adoption.
- cryptonector 6y agoThe fact that UDP involves much smaller PCBs than TCP alone will drive adoption of HTTP/3 because it will free up a fair bit of memory. More availability of HTTP version gateways in load balancers and other reverse proxies is all that's needed, and that's coming along.
- microcolonel 6y agoMore specifically no PCB, for UDP itself.
- cryptonector 6y agoIt's not nil. For "connected" UDP sockets, it's smaller than TCP's, but not nil because, well, buffers. And for non-connected UDP sockets there's still buffers. The main thing is that you can have much less buffer space because you might always be willing to drop packets. Ultimately you can have much lower memory pressure from those buffers and the smaller PCBs.
- moreati 6y agoProtocol Control Block for anyone else wondering https://www.oreilly.com/library/view/tcpip-illustrated/020163354X/ch22.html https://www.oreilly.com/library/view/tcpip-illustrated/02016...
- the8472 6y agoYou still need to keep your unacknowledged data buffered somewhere. If the kernel isn't holding it then it's in userspace.
- tialaramex 6y agoQualys' "SSL Pulse" says 47.1% of surveyed sites offered HTTP/2 and about 30% of surveyed sites offer TLS 1.3 Increasingly "corporations" out-source this problem to specialists who are only too pleased to use newer technologies with better performance and collect the same money.
- npiit 6y agoThe transition will be faster because of that a big chunck of the internet is gatewayed through a few big players (e.g. Cloudflare, AWS, CDNs, the new wave of static deployment services like Netlify and Zeit.co, big websites like google, facebook, netflix, etc...)
- awirth 6y agoThis is only mostly true, they do subtly change some semantics. As a motivational example, it is possible to encode a colon in a header name in http2 but not in http1.1, and this does not violate the RFC which only blacklists "\0\r\n".
- 1_player 6y agoWhy would network providers have to fix their network? Why 5 years to adopt? HTTP/3 and QUIC are based on UDP. This is very different to the IPv6 transition.
- zozbot234 6y agoWe might want to have a SCTP-based HTTP/4 down the road. That would surely benefit from some fixes on the network side.
- jeffbee 6y agoSee, that will NEVER happen. Completely impractical. SCTP has a different protocol number in the IP datagram header and many devices will either drop or malfunction when faced with protocol numbers they don't understand. UDP and TCP (protocol numbers 6 and 17) are well-supported, by practically all devices.
- anticensor 6y agoWhy not package QUIC in IP directly without UDP in between?
- sagichmal 6y agoA protocol separate to UDP and TCP altogether would suffer from middlebox interference problems.
- anticensor 6y agoAaaa, you mean the problem of -smart- stupid pipes. These do and will exist all the time and this is an opportunity for them to realise how detrimental they do is to Internet.
- sagichmal 6y agoThey're inextricably woven into the fabric of the internet, and unfortunately can't be wished away.
- dathinab 6y ago> The problem is similar to the IPv6 transition, Maybe similar but much smaller then IpV6 with much less problems, because most web frameworks will transparently support HTTP/1, HTTP/2 and HTTP/3 for the large majority of use-cases. > Mail still flows to many places unencrypted Mainly because getting a TLS certificate wasn't that easier for a lot of people in the world until recent years and the standard being written in a way which can be easily (mis-) understood as you having to have support unencrypted sending/receiving of mails. (It requires it for sending for local, i.e. implicitly by OS user account authenticated same machine mail.).
- Matthias247 6y ago> Far from it. It's quite likely going to be over 5 years if not a decade before it would be possible to run a pure-HTTP3 service without risking connectivity problems There won't be a pure HTTP/3 service anyway. HTTP/3 requires negotiation. It is announced from a previous HTTP/1.1 or /2 request using a "Alt-Svc" header. Typical clients will not try to connect using HTTP/3 directly
- patrickmcmanus 6y agoI think we'll see some DNS version of alt-svc that doesn't require TCP to bootstrap.. see httpssvc and svcb