Re: [NNagain] broadband measurement summit in DC March 7th

[Dave Taht <dave.taht@gmail.com>]

On Fri, Mar 1, 2024 at 8:50 AM Sascha Meinrath <sascha@thexlab.org> wrote:
>
> Hi Dave,
>
> Never got it unbricked,

Worst fat fingering bug of my life. I so wish you had had that box as
a reference while you developed other measurement tools.

> alas, but we did finish up the RadarToolkit.com
> firmware, and are currently winding our way through getting permissions to open
> source that software stack (the federal agency has given the o.k., and now we
> just need the lead NGO to get sign-off from its board).

I have grave difficulty trusting any measurement software that is not
open sourced. I look over your website and it makes me really twitchy
- I point people at validated tools like irtt and flent and there is
no, there, there, over there...

Please tell me you are measuring latency properly? I just took apart
some data from someone else that foolishly let me see the code and ...
pesky NDAs... they are a license to lie... I am never signing another
NDA again so long as I can help it.

Over here is a classic example of why measuring "average latency" is a
function of the measurement interval.

http://www.taht.net/~d/Misunderstanding_Residential_Bandwidth_Latency.pdf


Radar is a fairly
> extensible Pi-based (and mobile-app-based) measurement tool that enables
> longitudinal data collection to document network reliability.

The site is devoid of useful scientific data.

 We've already
> deployed hundreds of these devices across 4 states, and have documented
> real-time "rolling blackouts" on specific ISPs up in Alaska. Radar is also
> fairly seamless in terms of collecting info needed to make BEAD challenges
> (which is this month's hot issue) -- though I'm far more interested in
> collecting baseline intel prior to BEAD interventions (so that we can document
> meaningful impact).
>
> I remain quite perplexed on how NTIA plans to actually document the efficacy of
> their public investments. One of the biggest (rightful) critiques of the BTOP
> program was that there was no empirically-based assessment of what worked, and
> what didn't. 12 years later and NTIA appears hell-bent on making exactly the
> same mistake.
>
> Meanwhile, I'll be very keen to see how the Broadband Nutrition Label rollout
> goes -- starting next month, we *should* be able to do the first-ever,
> systematic comparative pricing analyses on a national scale (at least among the
> large-scale providers).
>
> --Sascha
>
> On 3/1/24 08:34, Dave Taht wrote:
> > On Fri, Mar 1, 2024 at 8:32 AM Sascha Meinrath <sascha@thexlab.org> wrote:
> >>
> >> Hi all,
> >>
> >> I'll be there!
> >
> > Did you ever get your wndr3800s to work? I could send you a box of a
> > few 25 dollar fq_codeled gl.inet boxes to anyone that needs ´em,
> > particularly those still stuck with DSL. I have also been encouraging
> > folk to run the waveform or speedtest.net tests on the conference
> > wifi....
> >>
> >> --Sascha
> >>
> >> On 3/1/24 07:29, Dave Taht via Nnagain wrote:
> >>> Anyone going to this?
> >>>
> >>> https://broadbandbreakfast.com/broadband-measurement-summit/
> >>>
> >>>
> >
> >
> >



-- 
https://blog.cerowrt.org/post/2024_predictions/
Dave Täht CSO, LibreQos