Re: [Bloat] Germany mentions latency

[Sebastian Moeller <moeller0@gmx.de>]

HI Michael,


> On May 22, 2022, at 18:10, Michael Richardson <mcr@sandelman.ca> wrote:
> 
> 
> https://www.thelocal.de/20220504/explained-how-germany-is-trying-to-tackle-its-slow-internet-problem/
> 
> Near the end of the article:
>   "In addition to the minimum download and upload speeds, the government
>   says the latency (reaction time) should also be no more than 150
>   milliseconds."
> 
> It would be nice if they used the (Apple) RPM metric instead, as more people
> would be be able to immediately measure that from a phone/laptop.

	This is IMHO a slightly different kettle of fish, alas. This is the legal reference of what is/will be considered to be the minimum internet grade people in Germany are entitled to (for some measure of entitled). As is all measurements are against the nationally certified measurement infrastructure (breitbandmessung.de) operated for the national regulatory agency (Bundesnetzagentur: short BNetzA) by zafaco. I am willing to bet the 150ms are a direct result from old ITU acceptable mouth to ear one-way delay data.
	I agree that it would be helpful to convince the BNetzA to switch their latency measurement from idle latency to latency under load (especially since the ITU numbers really only make sense if interpreted as latency-under-load/working-conditions). Personally I think an IETF RFC (currently in the making thanks to great folks at apple and other places) will have more sway than (just) pointing to apple's Macos/iOS internal/proprietary networkQuality tool.

> I wonder if some Apple PR person might be willing to respond to this effort
> in a complementary way and suggest this.
> (And I'm not an apple user)

	That could help, if someone from apple would talk to BNetzA* pointing at the IETF effort and explain why they implemented the networkQuality tool; not sure though how such a contact would happen (full disclosure I know exactly 0 person inside the BNetzA or politics that would need to be convinced).


Regards
	Sebastian

*) I had a discussion about reporting MTU per direction which span multiple years between me trying to explain why this is worthwhile (and apparently failing), and them implementing something, albeit half-arsed: for their servers they report the servers configured MTU and not the path MTU in download direction... I tried to also explain why latency under load would be immensely intersting, but so far crickets....


> 
> --
> ]               Never tell me the odds!                 | ipv6 mesh networks [
> ]   Michael Richardson, Sandelman Software Works        |    IoT architect   [
> ]     mcr@sandelman.ca  http://www.sandelman.ca/        |   ruby on rails    [
> 
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat