On Thu Apr 04, 2024 at 01:09:46PM +0300, Aleksi wrote:
Just because prior proposals/methods/ideas didn't work, doesn't mean a new one wouldn't or that all avenues have been explored.
People not wanting to discuss it endlessly on mailing lists are not preventing you from doing the research to create an RFC and two interoperable code bases demonstrating it works. As with Edison the only way to disprove the doubters is to do it and see if people adopt it.
So what's the complete unworkable thing on this? It requires zero network upgrades, only end points need to understand this. Therefore, adept coder makes a patch, it gets included in Linux kernel after a while of testing, and just couple years down the line you have what 90-95% of servers supporting it already?
Anyone can submit those patches to Linux any time. brandon