For the graphs, it would be great for f they were using a normalize output that allows for easy comparisons between runs. Especially the y axis for the “all” graph.<br><div class="gmail_quote"><div dir="ltr">On Mon, Nov 27, 2017 at 15:55 Dave Taht <<a href="mailto:dave.taht@gmail.com">dave.taht@gmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On Mon, Nov 27, 2017 at 3:16 PM, Martin Geddes <<a href="mailto:mail@martingeddes.com" target="_blank">mail@martingeddes.com</a>> wrote:<br>
> Hi Toke,<br>
><br>
> The two critical references are this paper and this PhD thesis. The former<br>
> describes "cherish-urgency" multiplexing. The "cherish" is what is different<br>
> to today's scheduling. It is used to create a new class of algorithm whose<br>
> goal is global optimisation, not local optimisation (and global<br>
> pessimisation).<br>
><br>
> The latter describes a paradigm change from "build it and then reason about<br>
> emergent performance" to "reason about engineered performance and then build<br>
> it". It works in practise, so whether it works in theory is left as an<br>
> exercise to the reader.<br>
><br>
> The first step is to get the measurement right. I'm running a public<br>
> workshop in London on 8th Dec, and I am happy to accommodate anyone from<br>
> this list at our internal cost.<br>
><br>
> Everyone working on AQM has done the best possible within the paradigm they<br>
> are operating. There is a bigger box of possibilities available, but it<br>
> needs you to engage with a paradigm change.<br>
<br>
We are currently benchmarking the known alternatives vs everything<br>
else via a dozen methods we understand.<br>
<br>
fq_codel v "cake":<br>
<br>
<a href="http://www.drhleny.cz/bufferbloat/cake/round1/eg_csrt_rrulbe_eg_fq_codel_200mbit/index.html" rel="noreferrer" target="_blank">http://www.drhleny.cz/bufferbloat/cake/round1/eg_csrt_rrulbe_eg_fq_codel_200mbit/index.html</a><br>
<br>
<a href="http://www.drhleny.cz/bufferbloat/cake/round1/eg_csrt_rrulbe_eg_cakeeth_200mbit/index.html" rel="noreferrer" target="_blank">http://www.drhleny.cz/bufferbloat/cake/round1/eg_csrt_rrulbe_eg_cakeeth_200mbit/index.html</a><br>
<br>
> Martin<br>
><br>
> About me Free newsletter Company website Twitter Zoom My new start-up Not<br>
> LinkedIn Martin Geddes Consulting Ltd, Incorporated in Scotland, number<br>
> SC275827 VAT Nu<a href="https://maps.google.com/?q=mber:+859+5634+72+Registered+office:+17-1&entry=gmail&source=g">mber: 859 5634 72 Registered office: 17-1</a>9 East London<br>
> Street, Edinburgh, EH7 4BN<br>
><br>
> On 26 November 2017 at 12:20, Toke Høiland-Jørgensen <<a href="mailto:toke@toke.dk" target="_blank">toke@toke.dk</a>> wrote:<br>
>><br>
>> Martin Geddes <<a href="mailto:mail@martingeddes.com" target="_blank">mail@martingeddes.com</a>> writes:<br>
>><br>
>> > It doesn't matter what scheduling algorithm you build if it creates<br>
>> > arbitrage or denial-of-service attacks that can arm a systemic<br>
>> > collapse hazard. The good news is we have a new class of scheduling<br>
>> > technology (that works on a different paradigm) that can fully address<br>
>> > all of the requirements. We are currently deploying it to enable the<br>
>> > world's first commercial quality-assured broadband service.<br>
>><br>
>> Could you point to any research papers describing this technology? Would<br>
>> be interesting to read up on...<br>
>><br>
>> -Toke<br>
><br>
><br>
><br>
> _______________________________________________<br>
> Bloat mailing list<br>
> <a href="mailto:Bloat@lists.bufferbloat.net" target="_blank">Bloat@lists.bufferbloat.net</a><br>
> <a href="https://lists.bufferbloat.net/listinfo/bloat" rel="noreferrer" target="_blank">https://lists.bufferbloat.net/listinfo/bloat</a><br>
><br>
<br>
<br>
<br>
--<br>
<br>
Dave Täht<br>
CEO, TekLibre, LLC<br>
<a href="http://www.teklibre.com" rel="noreferrer" target="_blank">http://www.teklibre.com</a><br>
Tel: 1-669-226-2619<br>
_______________________________________________<br>
Bloat mailing list<br>
<a href="mailto:Bloat@lists.bufferbloat.net" target="_blank">Bloat@lists.bufferbloat.net</a><br>
<a href="https://lists.bufferbloat.net/listinfo/bloat" rel="noreferrer" target="_blank">https://lists.bufferbloat.net/listinfo/bloat</a><br>
</blockquote></div>