Viewing profile — mrmoka

This is where DCCP or alternative protocols come in play.

In case of DCCP it can be implemented over UDP utilising existing firewall support for UDP traffic.

The main goal of this effort is to explore server-client communication scenarios.

You probably have not read the document. It's clearly stating - there is no need for reliability and ordered delivery of data. The only requirement from security point of view is c…

UDP over IPv4 and IPv6 has source IP and source Port in header. This can be used for sticky routing through Load Balancers.

You are absolutely right. But we need to remember - our motivation is to get a solution without most of TCP mechanisms to suit cases where it is not required. SCTP - is more of an …

Ability to connect to raw TCP or UDP from a browser is a major security hole. Due to that, WebSockets were created were handshake is handled by browser transparently from developer…

If there is alternative transport protocol to enable low-latency delivery as efficient as UDP, then it is worth exploring. UDP - is something been used by many industries for long …

> You can't compare WebSockets with WebRTC because the first one has a much larger implementation base across browsers than the later. This is exactly what I pointed out: WebSocket…

WebRTC is best option for media streams today for peer-to-peer cases. The goal of the topic is to explore simple option for server-client communication using low-latency communicat…

Simply look at present state of WebRTC adoption by back-end services for use in server-client cases. Fist of all it is not designed for server-client cases. Second, after so many y…

Worth mentioning again: this effort is to explore server-client low-latency, not peer-to-peer scenarios which WebRTC solves well. And this is collaborative effort, not personal. So…

It is indeed a bit like chicken-and-egg situation here. But SCTP implements reliability and ordered delivery making it more of an alternative to TCP, than a solution for low-latenc…

This is something that worth exploring further to identify if it is a requirement for a WebUDP implementation.

> I'm very[1] familiar with the IP family of protocols. Great, you have an expertise to contribute. And with this discussion you already are. > If you don't know what your requirem…

SCTP have been mentioned in the repo: https://github.com/Maksims/web-udp-public/issues/1#issuecomm...

This is valuable input, you are welcome to contribute in form of PR!

Just looking at the complexity of "minimal" implementation only highlights the need for different solution.

And with WebUDP life for those guys would be much easier.

This spec been there for very long time, and has been adopted by FirefoxOS (deprecated platform). Which exposes low-level access to establish pure TCP and UDP connections with perm…

This is collaborative effort. I act from my capabilities, but people with certain skillset are welcome to contribute with proper RFC.

As mentioned in the doc, one of the options is to simplify WebRTC by making some components optional to enable it's better adoption and easier to implement on the back-end as well …

Those implementations were created to be utilised within FirefoxOS where applications are granted permissions by user when installing them. And access to several API would be stric…

> There is a lot more to security than transport-layer encryption and authentication. You are welcome to PR. > Connection based This could be explored: starting from simple handsha…

If it is implemented with handshake mechanics same as WebSockets, then it would be no more harmful than WebSockets.