[Rpm] [LibreQoS] benton's consumer broadband label prototype

Wed Oct 26 16:08:11 EDT 2022

Hi Dan,

> On Oct 26, 2022, at 17:42, dan <dandenson at gmail.com> wrote:
> 
> My argument here is more about government inadequacies.

	[SM] Sure regulatory capture is a reality and we can/should always strive for better government, but existing government is IMHO better than no government ;) (not a fan of anarchy)

>  Without government warning labels for internet service we have a massive streaming industry, gaming industry, home office workers, and so on.

	[SM] That is somewhat orthogonal, as we also have a massive organized crime "industry" in spite of government warning/acting against it ;)

>  It's imperfect, but the fact that many of us complain about how kids are stuck in front of an xbox playing fortnite... well I guess it's working to some degree.  

	[SM] Not saying that all internet service is bad, just that enough of it is sub-promise that I personally consider regulatory action desirable. We can discuss how that action should look like. (I also want to add, that e.g. in Germany many ISPs already delivered on their promises even before the changes in law).

> Government's ability to regulate and protect consumers that already exists just isn't utilized, ie if you say you sell X, you give X else void contract and potential refunds of up-front costs.  

	[SM] Hence me bringing the example from the German regulatory agency that (acting based on a change in the relevant telecommunication law) does exactly that, make sure both parties know what an internet access link contract entails.

> In the US at least, the government's decision making process has fed the 'bad' companies 3000-4000 per subscriber to build out services that don't address the underserved anyway.  All of this money thrown out there by the government making their decisions and they've not moved the needle at all.

	[SM] There is some inefficiency in the system of how to getting those links built/updated to broadband where a pure profit motive with the established prices will not work by itself. However it needs "free-market" players to abuse the government's willingness to pay for such build-outs (and gorvernments willingness to not look close enugh to the details of the implementation).

> IMO, very simple rules on advertising based on delivering what is claimed will force industry innovation.

	[SM] What kind of innovation do you envision here? After all "basic internet access" is a commodity, or should be, no?

>  If a company says 'fast' and that term is too broad, then any attack on 'my latency is crap so this isn't fast' could lead to refunds.  This is capitalism after all.  

	[SM] In the UK ISPs are, as far as I know, only allowed to advertise with rates that are somehow empirically evidenced. In Germany they can advertize with the PIB's "maximal rate", but then are held accountable to actually deliver something close to that rate (the rules how the regulator controls whether ISPs hold their contracts are, in good German fashion, somewhat baroque and convoluted/complicated, I am sure once the US commits to its own system it will be likely better/simpler to use).

> Maybe companies should have to say 'best throughput' if they are trying to hyper their 1G or 2G service, they can't say 'fast' because that's not really a valid measure.  'great low latency services' instead of fast.  

	[SM] All of these terms like "fast" have no bearing over here, ISPs need to reveal actual rates and then need to deliver these (again there are rules how to interpret the availability of these rates end users can expect). Over here ISPs advertize typically with their maximal download rate.

>  We have the language to market properly but companies are allowed to market vaguely and it IMO contributes to consumer ignorance and misunderstanding.
> 
> For our 'fiber like' services we literally pitch latencies to various gaming services and zoom relays. 

	[SM] Which IMHO is fine, assuming you actually deliver robust and reliable low latencies as far as that is under your control. Actually if you deliver that, it well above just fine ;). Bit of a sign of competence for any network operator not to blow up latency under load beyond reason.

Regards
	Sebastian

>   
> 
> On Wed, Oct 26, 2022 at 9:27 AM Sebastian Moeller <moeller0 at gmx.de> wrote:
> Hi Dan,
> 
> 
> > On Oct 26, 2022, at 17:09, dan via Rpm <rpm at lists.bufferbloat.net> wrote:
> > 
> > Complete fail on a marketing perspective though, this would have to be legislated and then handled by a third party.
> 
>         [SM] Well this is not intended to be a marketing tool, but a regulatory tool to make sure the market works to the benefit of society (I understand that market members are incentivized to skew the market mechanism to their advantage, this is why working markets need equally working regulation, just like competitive sports require umpires/referees). So yes legislation might well be required, but that would not be a sign of failure, no?
> 
> >  No one is going to put out essentially a warning label that says 'poor' or 'marginal' in any category for a product they sell.  I wouldn't, and we have LTE services to get to people with no other option and they are quite happy, it would be detrimental to hand them a sheet that says that the service is actually 'poor'.
> 
>         [SM] True, but e.g. in Germany ISPs are required by law to publish their contracted rates in a pre-described fashion pre-sale and are actually held responsible to some degree to actually deliver the promised rates. (Well, not really, but consumers can get a cost-free right to immediately cancel their contract or reduce their payments commensurate to the under-delivery of the contracted speed*). What happens here is not that ISPs need to disclose shitty service but that the need to declare what they intend to deliver and they are simply held responsible to actually do so**.
> 
> 
> *) The first option is already well established and works, the payment reduction part is ATM still being worked out.
> **) Unfortunately, the required numbers currently do not include latency under load or even idle latency... there is still work ahead to convince the regulatory agency of that.
> 
> > Also, my trust in the government to decide what's good or bad... laughable.
> 
>         [SM] Compared to bigger cooperations operating in "free-market" capitalism? Really there is no alternative to government for that purpose...
> 
> >    You'd get things like on the example page.  903.5Mbps Median download speed, 811.8 Median upload speed, gaming rating poor and video conferencing marginal, on Fios service.  I know that's an example, but it's so spot on what the government might do...
> 
>         [SM] See e.g.:
> https://www.bundesnetzagentur.de/SharedDocs/Downloads/DE/Sachgebiete/Telekommunikation/Unternehmen_Institutionen/Anbieterpflichten/Kundenschutz/Transparenzmaßnahmen/templates_for_information_sheets.pdf;jsessionid=0868AE15965FB584C81008C96BA15E4B?__blob=publicationFile&v=1
> and
> https://www.bundesnetzagentur.de/SharedDocs/Downloads/DE/Sachgebiete/Telekommunikation/Unternehmen_Institutionen/Anbieterpflichten/Kundenschutz/Transparenzmaßnahmen/Instruction_for_drawing_up_PIS.pdf;jsessionid=0868AE15965FB584C81008C96BA15E4B?__blob=publicationFile&v=1
> 
> for how something similar might look in practice.
> 
> > 
> > As an operator, I will not implement this unless forced to and then I'll support lobby efforts to get it removed.
> 
>         [SM] Ad that is why we can't have nice things... ;) No really, I agree this needs legislative/regulatory backing/teeth to work, but that is not a failure but simply how our system developed.
> 
> Regards
>         Sebastian
> 
> 
> > 
> > On Tue, Oct 25, 2022 at 8:30 PM Dave Taht via LibreQoS <libreqos at lists.bufferbloat.net> wrote:
> >  is actually... not bad.
> > 
> > https://www.benton.org/blog/consumer-driven-broadband-label-design
> > 
> > -- 
> > This song goes out to all the folk that thought Stadia would work:
> > https://www.linkedin.com/posts/dtaht_the-mushroom-song-activity-6981366665607352320-FXtz
> > Dave Täht CEO, TekLibre, LLC
> > _______________________________________________
> > LibreQoS mailing list
> > LibreQoS at lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/libreqos
> > _______________________________________________
> > Rpm mailing list
> > Rpm at lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/rpm
> 