Please expand upon your point. Why exactly is 25/3 reasonable for most people? I'm not necessarily disagreeing; indeed, this would be a god-level speed compared to what's currently standard for most of Latin America. However, your point about Netflix is irrelevant because not every use case can be buffered. Also, many use cases in our new work-from-home reality require better up speed than 3Mbps with bad ping and high jitter.
nothing special about netflix, just a service that most people use. even their highest quality offering is streamable if we actually got advertised speeds. just trying to say that 25 isnt a meaningful cap for most people.
> require better up speed than 3Mbps with bad ping and high jitter
rejectfinite makes this point downthread and i agree with both of you. i think theres a lot more bang for your buck to be had by reducing quality problems and widening the upload pipe.
100Mbps download so that multiple people can simultaneously stream 4k video is not a sane default.
Remote work has become very much A Thing since the COVID lockdown. Solid upload/download bandwidth isn't about streaming multiple 4K movies. It's about productivity where a sketchy 3Mbps connection is a hurdle.
Yes it is because this is how many people already live. If you’re a typical family of four everyone has at least one device they stream content on.
It is if you don't want to update the minimum every 5 years. 100 isn't even high enough, IMO. It should be at least 500 down and at least 100 up to be safe for another decade, maybe 2.
If the pandemic has shown most of the world anything, it's that people need internet with high speeds. With the increasing gap between rich and poor, there are more people in flat shares too (per flat and total flat shares). Houses are bought later, if at all.
Even ignoring the "more than one person" issue, 3Mbps makes it very difficult to do things like file transfers and backups and accessing home data from elsewhere.
25/10 per person is a much better minimum bar than 25/3
> all in all, this reeks of regulatory capture to me.
Why? How does this favor entrenched players over newcomers?
The companies I see being disadvantaged by this are incumbent cable companies that need to update equipment to let more frequencies be shifted from download to upload. But that's not regulatory capture by ISPs and should have good effects. And this doesn't seem like some kind of incumbent ISP vs. incumbent ISP regulatory capture either.
> all in all, this reeks of regulatory capture to me. if we're gonna have a serious discussion about internet infrastructure, i think the starting point is how we classify ISPs and where in the city-state-federal hierarchy the responsibility lies. changing a 26 to a 100 is a cool bullet on a resume, but i dont think this is the low handing fruit here.
The distributed regulation is the biggest problem IMO. ISPs are generally granted explicit or effective monopolies by local governments. Local competition is difficult enough for a utility business, but city governments make it effectively impossible in many places. Then there's a few big players who are able to claim competition and get lenient regulation at the federal level.
Granting ISPs explicit monopolies has been illegal since the 1992 cable act. Show me a franchise agreement that has an exclusivity provision. They’re public documents.
On the other side of the scale, Google Fiber has had the carpet rolled out for it in many places, for example being given for example free leases to public property for installing fiber huts: https://techliberation.com/2012/08/07/what-google-fiber-says...
> Google received stunning regulatory concessions and incentives from local governments, including free access to virtually everything the city owns or controls: rights of way, central office space, power, interconnections with anchor institutions, marketing and direct mail, and office space for Google employees. City officials also expedited the permitting process and assigned staff specifically to help Google. One county even offered to allow Google to hang its wires on parts of utility poles – for free – that are usually off-limits to communications companies.
See also: https://hbr.org/2018/09/why-google-fiber-is-high-speed-inter...
> But we believe Google Fiber’s most significant impact was to change the nature of relations between infrastructure providers and local authorities. Even after substantial deregulation in the 1980’s and 1990’s, and even as separate networks and technologies converged on a single internet-based standard, local governments continued to treat network providers as quasi-governmental public utilities, regulating their construction efforts and access to public rights of way with cumbersome procedures developed decades earlier.
One of the biggest examples of special treatment Google has negotiated in almost every Fiber city is exemption from build-out requirements: https://www.techrepublic.com/article/how-google-fiber-launch...
> Google Fiber has started with three neighborhoods in Louisville--Portland, Newburg, and Strathmoor (better known as the Highlands, where all the young professionals hang out). Residents of those neighborhoods can start ordering the service today by going to google.com/fiber/louisville. There are two packages to choose from:
While focusing on higher income neighborhoods where more people are likely to sign up for service seems like a no brainer in fact it’s an explosive political in most cities. That was the exact reason Verizon wasn’t allowed to deploy FiOS to Baltimore, and was a major reason for delays in FiOS deployment for NYC.
Looking at the details of how Google approached Fiber in various cities is a roadmap for what’s holding back deployment across the country. The problem was not incumbent lobbying—cities like Baltimore were falling over themselves to get Google to build Fiber. But what Google demanded in every fiber city was the sort of thing it got in Kansas City: elimination of the red tape that drives up costs. It’s no surprise that Google ended up mainly building Fiber in a bunch of red state cities.
Yes, for one person.
Having lived with a family fo 4 using ADSL, it was not fun when 3 people watched Netflix on their own screens, and having to update a game.
Or with the pandemic and WFH, you have maybe 4 people using DSL for Teams/Skype/Zoom meetings and trying to work/study at the same time.
streaming netflix in HD instead of UHD is 5mbps. 5x3=15, leaving 10 free to download a game. pretty reasonable. the pain was likely that you didnt have a real 25meg pipe, or that there was latency or jitter. a true 25Mbps would have been enough.
> the pandemic and WFH
I think you could make a strong argument that 3meg up is no longer sufficient. we live in an era where duo, facetime, and conference call programs and ubiquitous.
i do not, however, think you could make the argument that changing the download Mbps target from 25 to 100 is the best way to spend billions of dollars in a market that no one would accuse of being saturated and highly competitive.
Except most household have more than 1 person. A family of four sharing 100Mbps isn't a lot.
I am not sure if 100Mbps upload is a realistic option in US. But I would at least expect 50Mbps to be minimum per household. DOSCIS 4.0 should be able to handle that.
DOCSIS 4.0 was ( finally ) finalised last year. And could be pushed for upgrading. But that is just the last mile, Cable companies will still have to ensure fibre roll out. Look at UK. They had a half decent infrastructure with DSL and could have pushed G.Fast, instead they just bite the bullet and pushed FTTP to maximum.
Although technically I was only upgraded from 100 to 200 for free a few years ago because a competitor came in since then with fiber who offers 200/200 at the same price. I haven't switched over because they cut me a deal and truthfully I don't upload enough for the extra amount to matter and I get single digit ping times out of my provider which I'm happy with.
I have 50/20 in Australia on VDSL2, and it is just barely good enough for this. You're right about it not being consistent/maintained though. Some days I could collab remotely no worries, other days it was simply impossible.
No, 3 megabits upload is not reasonable. If people have to work from home or school from home the 3 megabits upload quickly becomes too little for a google meet/teams video chat or screensharing.
It becomes even worse if multiple family members have to do it at the same time.
starlink is changing that, as these articles and letters are being written.
> So will Starlink be a good option for anyone in the United States? Not necessarily. Musk said there will be plenty of bandwidth in areas with low population densities and that there will be some customers in big cities. But he cautioned against expecting that everyone in a big city would be able to use Starlink.
...
> "I want to be clear, it's not like Starlink is some huge threat to telcos. I want to be super clear it is not," Musk said. "In fact, it will be helpful to telcos because Starlink will serve the hardest-to-serve customers that telcos otherwise have trouble doing with landlines or even with... cell towers."
> Starlink will likely serve the "3 or 4 percent hardest-to-reach customers for telcos" and "people who simply have no connectivity right now, or the connectivity is really bad," Musk said. "So I think it will be actually helpful and take a significant load off the traditional telcos."
From https://arstechnica.com/information-technology/2020/03/musk-...
Beaming internet from space solves a ton of problems for anyone out at the ends of rural copper, or using p2p systems, or where there simply isn't another alternative.
But using myself as an example, I live in a part of NYC where there is fiber internet in the street in front of my building - and my landlord has zero obligation to allow the 60-ish apartments in the building to access it. We're stuck on legacy cable, which has abysmal upload speeds and congestion issues.
The dense urban or suburban environments theoretically have more options, but thats not always reality... and I'm not sure that tens of millions of people replacing their local ISP with satellite is desirable, or even technically viable. New provider options are important, but making sure access to the existing options is equitable and reasonable (across all the various local governments, providers, middlemen, etc) is important too.
What I hope to see from Starlink is some freedom for more knowledge workers to work remotely, as in more rural communities.
-regulating ISPs as a free market is, to be hyperbolic, a disaster. there is effectively a monopoly in most parts of the country.
-25/3 Mbps is a very reasonable speed for most people. the issue is that its provided on a best effort basis that is never maintained for long, if at all. netflix says thats enough to stream UHD[0]
-FTA: "Those data points likely undercount the number of unserved Americans because the FCC lets ISPs count an entire census block as served even if it can serve just one home in the block" hilarious - what else is there to say.
-FTA: "The senators are also frustrated by differing standards across agencies. [...] the FCC defines [...] Alternatively, the US Department of Agriculture (USDA) defines" why is this the federal government's job again? the internet is clearly not an enumerated power. is this the interstate commerce clause that keeps on giving in action again? i think decentralizing federal power would help alleviate a lot of the existential dread vis a vi corruption that many americans are feeling.
all in all, this reeks of regulatory capture to me. if we're gonna have a serious discussion about internet infrastructure, i think the starting point is how we classify ISPs and where in the city-state-federal hierarchy the responsibility lies. changing a 25 to a 100 is a cool bullet on a resume, but i dont think this is the low handing fruit here.
[0]https://help.netflix.com/en/node/306