Google’s fiber effect: Fuel for a broadband explosion | Hacker News Reader

Google’s fiber effect: Fuel for a broadband explosion

cnet.com

159 pointsprlin76 comments

> AT&T and local cable operator Grande Communications have beaten Google to market

Oh, gee, look at all this bandwidth I just found...

> AT&T maintains it has been planning this fiber upgrade for a long time, and that Google's announcement didn't affect the timing of its network.

No, no, we planned this all along...

> What Google recognized that others didn't is that Americans want to have the best communications infrastructure

Yes, as it turns out study after study showing Americans were fiercely proud of being in 15th place, may have suffered some sort of statistical vagaries. We're looking into it.

yeah, and it's just a total coincidence that AT&T's upgrades that they've been "planning all along" are only in the cities that google announces Fibre in.

PacBell installed a bunch of fiber in the Bay Area that suddenly went unused/projects cancelled when SBC bought them.

Competition they call it.

AT&T, maybe, but Time Warner recently doubled speeds in NYC for free. The announcement was rather silent; I noticed downloads being faster, did a speed test, and had to search for information about it. A couple months later, I got a little card in the mail about it.

100Mbps isn't 1000Mbps, but I'll take it.

TWC definitely did it because of Google Fiber, too. They started doing this right around the time Google made it clear it's not just an "experiment" anymore, and they were starting to get a lot of media attention.

I strongly urge people that if they have to choose between Google Fiber and AT&T or Comcast or any of these other guys who've frustrated you for years and couldn't care less about you until Google Fiber arrived, to dump them immediately and go with Google Fiber anyway, even they are matching speeds and prices by then.

They don't deserve you as a customer anymore, and they wouldn't have given you this speed were it not for Google Fiber. So Google Fiber deserves to be rewarded with customers because of that, and Comcast/TWC/AT&T does not.

If Google Fiber isn't rewarded with customers for its fiber expansion, then it might die, and then what do you think will happen then? Comcast/etc will go back to their shitty services once again and not care one bit about upgrading anymore, and we'll be in the same situation we are now 5-10 years from now, or perhaps much worse if there's no net neutrality anymore, and all the current telco cartel boys get to bully small start-ups without any fear of any real competition that might not do that, and without having to risk losing customers again. That's why you should quit them immediately.

What has Google done to deserve you as a customer. What in the data gathering they do now makes it a good deal to expand their view of you?

The only thing that makes Google different in this situation is they haven't pissed you off yet.

No I would not be getting Google Fiber when/if it showed up, but I don't live anywhere where it would any time soon, if ever.

> What has Google done to deserve you as a customer[?]

Introduced competition into a market that had none. That's worthy of a switch in my opinion.

There was a time that the cable companies did that. Then everything converged to a new normal.

I don't buy into the Google worship. They are a corporation like any other and act like one. When switching to Google Fiber means ubiquitous Google data collection, I don't see it as a positive thing.

I suppose if you have the memory of a goldfish Google fiber is great as huge speed bump.

Grab a VPN or two and forget about all of your data collection worries

Sadly, I would take data collection w/fiber over Comcast anyway. If I want protection than I can just purchase a VPN account.

They doubled it, but how much faster would Google Fibre have been? More than double, I suspect.

That is not inconsistent at all. They were planning it and part of that plan was to hold out extorting and defrauding America as long as they could before initiating the roll-out plan when a threat to their monopoly and collusion is detected. That was the whole plan.

This is also very consistent with America in general, we say we like competition, but our businesses don't like competition beyond the lip service of uttering the word. We have virtually zero competition in so many aspects of our economy and society, especially in those that matter in the current type of world.

What we are not realizing is that our corporations are actually strangling America like a psychopath strangling a hooker, repeating "I love you, mommy!"

This is super cynical but it sure seems correct.

The case-study with which the article starts isn't a particularly good example of a reason for upgrading to fibre-based connectivity: latency doesn't care whether your connection is 1Gbps or 1 Mbps.

There might be a slight improvement given that routers and switches aling the way have to operate faster to handle gigabit switching but in the UK the there have been complaints that VDSL interleaving ( which can't be turned-off ) actually raised latency above their ADSL levels.

I was wondering the same thing. I'm not much of a gamer these days, but back when I played a lot of LMCTF, it was all about latency (and the offhand grapple). Have game engines gotten so much worse that bandwidth is now an issue, or is this just artistic license?

I did some math back (a few years ago) when I was on U-Verse to determine if this was an issue. It's not, not even close.

The average shooter, based on the highly scientific study of Counter Strike Source and Urban Terror (a quake mod), the average bandwidth used was under 128kbps. Latency, especially the variance of latency and packet loss are the biggest issues at hand.

I find it absolutely hilarious whenever ISPs market their larger packages for the "hardcore gamer", when in fact a real hardcore gamer would probably opt for the smallest package.

It’s all assets. For example, an average Team Fortress 2 map is 50MB, and servers can push new maps to clients when they connect.

Lag compensation [1] does not compensate for lost kills, unfortunately, and I bet people still uniformly complain about bad net code nowadays.

[1] https://developer.valvesoftware.com/wiki/Lag_compensation

How common is it to need to load a new map, as apposed to downloading the map once then caching it for future use?

Not frequently. Competitive players loathe new maps ;)

The one thing I can think of is bufferbloat in low-bandwidth networks, though I was under the impression that it was mostly cellular, not land-line, networks that used the kind of packet-drop policies (i.e. not dropping packets, just queueing them for up to tens of seconds) that led to that problem.

VDSL latency seems to be a bit of an odd one; when VDSL launched in Ireland, latencies were the same as, or a little worse than, ADSL, but in the last few weeks it seems to have, for at least some cabs, fallen sharply, to around 5ms. Not entirely clear what has changed.

My guess is that the cited problems are almost completely the fault of bufferbloat at or near (on the ISP side) the home router level. Pretty much all consumer hardware in existence seems to be configured with way too high buffer sizes.

Switching to fiber might help somewhat though, because a higher data rate translates the same number of bytes of packet queue into a shorter actual packet delay. So it slightly mitigates the pain of bufferbloat (at least until people start configuring the buffers to be even larger to "optimize for our faster networks").

Not quite true that VDSL interleaving can't be turned off. BT's line management software will adjust the interleaving depth, which can vary from quite a lot to not much at all depending on line quality and noise.

> customers don't need that kind of speed.

Sure they don't, they also don't need bluray or terabyte hdds either... Should be rephrased to 'We hope customers don't realize they can use that type of speed.'

Yup. That argument tends to bug me a bit. It's not that users don't need it, but more that they can't use it because they don't have it. I'd love to backup all my computers to the secure encrypted cloud with something like Backblaze. But, with 2mbps upload, I'm looking at like a month of straight upload time. If I had gig or even 100Mbps upstream, there would be no question here, I'd get to it. For now though, someone can argue "Oh look, you don't need gig because you're not even using your 2 meg all the time". Umm, yea, there's a reason I'm not using it all the time.

Verizon never gets enough credit for their fiber roll out. Sure they don't offer 1Gbps, but they laid fiber so they can hit those speeds when the time is right. And until then 50 or 100Mbps over fiber is a lot better than 20Mbps over DSL. Not just in the speed department but in terms of latency and consistency. DSL does amazing things but is just at greater risk of interference. My DSL seems to trane at different speeds each day. Fiber is just a much better medium, things don't get too weird until you want greater than 100 Gbps per wavelength.

Disclosure: I wrote some of the software for the early hardware behind FiOS back 10 years ago or so. Wow FiOS has been out there for a long time. Also I am old.

...what fiber roll out? As far as I know, they did some token fiber installations, but the vast, vast majority will never get it. In fact, from what I've read, they took the subsidies for fiber, and used it to build out their metered wireless network.

http://www.techdirt.com/articles/20140424/06185027014/verizo...

http://www.nj.com/cumberland/index.ssf/2014/03/hopewell_town...

Depends how you define token? They have about 5 million subscribers on FTTH. Supposedly passing 18M houses.

http://en.wikipedia.org/wiki/Verizon_FiOS#Stable_footprint_....

Note: I was focused on them being technical leaders. Which they were. I make no endorsement of their business practices. Also they were not the first to FTTH in the US or not the world, just the first big mover in the US.

Sorry, I was talking about the huge about-face they did once smartphones became popular and they realized there were much larger profit margins to be had in the wireless industry, so basically started shutting down expansion of fiber internet.

I think Ivan Seidenberg was fired because of short term returns on FiOS being low.

Verizon never gets enough credit for their fiber roll out.

Check the amount of subsidies that they have received over the years for promising fiber to the home. That is why, IMO.

Again, my statement was focused on the technical side. They were ahead a bit there. I am not endorsing their business or policts although they did indirectly pay my bills for a few years.

I actually think part of the problem is that no one in the bay area seems to know about FiOS because it doesn't exist here. They do offer the service to 15% or so of US household, but not me, or other folks by the bay.

Verizon's fiber rollout was disingenous. Installer's were told to disconnect the old copper connections to the house. The old copper connections in many states had open access requirements. The new fibre connections meant you were now locked into Verizon's fibre with no open access requirement.

Disclosure: I was a Verizon vendor and worked closely with several of their staff. At an individual level they were nice folks. But, their corporate culture only understood competing as a monopoly and didn't know how to play on a level playing field.

As awesome as it is to see Google kick the regional monopolies into upgrading, it's really a lack of competition that's stifling both modern connection speeds and reasonable prices. Prices for the ridiculously named "Gigapower" are undoubtedly going to go increase repeatedly over time if Fiber backs out since it has achieved what it wants. Competition would have the opposite effect.

Or better yet, why not treat ISPs as utilities. Could anyone offer insight as to why internet in this day and age is not regulated like electricity or other utilities?

Could anyone offer insight as to why internet in this day and age is not regulated like electricity or other utilities?

This was the obvious solution after the telecom act of '96. Many experts in the field said so much at the time. But the telecoms dug their heels, broke the law by locking out people from the DSLAMS, and used disinformation and lobbying to get the FCC to say it was O.K. to have a telecom-cable duopoly. This was also around the time that telecoms did everything they could, at the state level, to prevent cities from offering free city-wide internet [1].

[1] http://en.wikipedia.org/wiki/Municipal_broadband#Controversy

because utilities suck? Innovation, price, customer service, take your pick on which dimension they suck. Governments have had a century of experience dealing with utilities. I'm glad they are at least trying new models before going with the known to suck model.

Utilities would kill any chance of competition. Playing devils advocate here but if your city was in charge of upgrading your internet, you would be stuck with their plans and budget and red tape. No hope for google fiber or otherwise to attempt to make things better.

> The mere promise of Google Fiber seems to be enough to send rivals scrambling to deliver ultrafast Internet service at a reasonable price.

We have a TWC connection 2000m outside of Google Fiber availability that's been disconnecting up to 10-20 times a day for months that says some rivals just don't give a shit. The disconnections are only mildly more annoying than the constant spiking and pl that had been going on for a year or so..

Google is extremely good at commoditizing their complement. A whole lot of their strategy makes sense in this light: Android, "Project Ara"/Phoneblocks, Chromebooks, Google Fiber.

Good read on the subject: http://www.joelonsoftware.com/articles/StrategyLetterV.html

This strategy makes sense when we are talking about actual commodities that are interchangeable, like gasoline, but the effects are a lot more ambiguous when we are talking about unique inputs to a business like technology.

As an example, consider the current topic of broadband. In theory net neutrality is beneficial in terms of keeping costs down. This is beneficial for web companies. However, consider that without net neutrality, broadband providers have additional business models available that increase overall revenue. This additional revenue can help drive broadband R&D, which in turn could actually benefit web companies in the long run.

Making technology cheaper is good in the short run, but in the long run it may interfere with innovation, which benefits from things not being cheap, but actually profitable to produce.

He was spending $208/month for 20Mbps? What the hell is going on in Austin?

U-Verse is "triple play" so he probably had phone(s), TV(s), and Internet.

This is stll ridiculous pricing. I pay something like 120 a month for tv, hbo included, and 30 mbps internet. Unless he has every station on earth, he's being overcharged,

And there are people renting phones - yes, phones, really cheap ones too, those with only the 12 buttons and no other functions - for over $50 a month, from their telephone companies. These customers are mostly elderlies. It was just sickening.

(Worked at a call center)

What? I Can't believe those prices you're forced to pay. I have Triple Play, 120 Mb/s and a iPad App to view shows/sports separately for 55 euro's a month.

I would be paying the same except for the charge for two extra cables boxes and the charges for HD on three cable boxes and the charge for the DVR listing. Not to mention all of the taxes and fees. My bill ends up being over $200 with no phone.

Nickel and dimeing is off by a factor of 1,000.

I'm paying ~$300/mo for TV (HBO etc), phone (no long distance charges) and ~$35/mbit Internet connection. It's a rip-off, of course. But since there's no real competition (U-Verse?!) I have no choice. Meanwhile, the cable industry is reducing potential for competition (Time-Warner merging with Comcast).

There's always a choice. You can always choose to disconnect one or two of those.

That's a non-choice. A choice would mean being able to choose between a variety of competing services with good offerings. Disconnecting is not a realistic choice.

Talk to retention. You don't even have to threaten to cancel, just talk to them. I have to do this every year. The price I just quoted was set last week by TWC retention.

Thanks for the advice. I need to do this, obviously.

Just out of curiosity, I checked Virgin's prices in the UK. In my area (Kentucky, US) my ISP options are AT&T DSL and Comcast cable internet. I'm paying for the cheapest internet plan offered by comcast $45 for 25Mb down. The MOST EXPENSIVE plan offered by Virgin for £37.50 gets 152Mb down.

UK is much more dense than Kentucky. Really hard to compare prices like that.

For those curious:

UK: 660 people per sq mile Kentucky: 101 people per sq mile

I disagree. The US has already built all this cable and telecom infrastructure (only ~50% of the UK has access to cable, which is way lower than the US).

The expensive part - digging the road and laying the physical cable - is already done, and was done 30+ years ago in many cases. These assets have been paid for.

The high internet prices in the US are down to terrible regulation. While there are many things that the EU hasn't done well, it has really nailed telecom (both fixed and cellular) regulation.

Living in Austin, I've been closely following news related to Google Fiber's rollout throughout the city. The two big providers here, TWC and AT&T, are still doing business as usual. TWC has promised increased speeds by summer 2014, but as a TWC customer myself I cannot help but view their announcement with a fair bit of skepticism [1].

Interestingly enough, Grande Communications has actually beat Google with regards to being the first company to actually start offering/installing fiber services within the city limits [2]. Granted, their fiber offerings are limited to a small subset of people living in downtown/west Austin, but it's progress nonetheless.

Google has been incredibly tight-lipped about their fiber rollout. To this day, there have not been any official announcements since their initial unveiling 1 year ago. But from small news pieces I can find online, it seems their service launch has been delayed to 'later this year' (compared to the summer 2014 date they originally gave) [3]. Last, performing a search of permits files by Google for fiber construction reveals that their initial rollout/offering will be exclusively to south Austin residents (south of Lake Austin) [4].

[1] http://www.pcworld.com/article/2099908/aha-time-warner-cable...

[2] http://mygrande.com/austin/1-gig-fiber-internet/

[3] https://gigaom.com/2014/04/08/google-delays-austin-fiber-lau...

[4] https://mapsengine.google.com/map/u/0/edit?mid=zKGIh6VixM7Q....

Great, can Google Fiber take a serious look at our market in Australia: the current chumps running the country seem to be taking backward steps with the previous government's (admittedly ambitious) plans for a National Broadband Network every day.

I just visited my sister in Australia and much worse than the speed was the bandwidth caps. Instead of unlimited, she bought credit for a specific amount of data. Is that how it works all over?

Almost completely. There's a few ISPs that offer unlimited data but they're usually fairly oversubscribed as a result. Quotas have always been the major cost point here, not speed.

As much as I have problems with the current govt. here in NZ. We're getting Fibre to the Home rolled out and it seems to be working. My suburb (admittedly one of the nicest in the country...) has got it already and it's pretty great. We just had two of the biggest ISP's announce unlimited fibre for $110 per month with a phone line, ~$80 Naked.

Things are getting better here slowly, can't believe how retarded the liberal party is over the NBN.

A millions times, yes! I'm so mad every time I heard that the people who voted "Liberal" thought they'd fix, not jeopardise the NBN.

Please help us, Google!

Yep, broadband speed here in Australia is pretty bad overall, and the contracts are very expensive, and most are capped for bandwidth as well. I am on an unmetered contract with TPG, but they're the exception. I was holding out for the NBN, but with the recent change in government, this has all but fallen apart.

You do realise the previous government of chumps passed a law making it illegal for Google to create Fiber in Australia even if they decided they wanted to?

The NBN is the worst policy idea to hit the Australian tech industry, ever. I doubt there will be one worse. It has already wreaked a trail of destruction and slowed down adoption of high speeds, it will continue to do so for at least the next 5 years, until some future government finally sells the scraps and turns it over to private companies, as it should have done from the get-go.

We had a grandiose government scheme announced 7 years ago. 7 Years down the track, private investment is forbidden, none of the Telcos have made any additional investment, existing infrastructure is being torn up. And the number of people passed by the new system is less than what a decent football stadium will hold, but they are $15 billion spent already. The forecast cast has progressed beyond $65 billion, but any number you hear is just finger-in-the-wind anyway.

It's fashionable - very fashionable- to love on the NBN, but it is a terrible, terrible idea. Fibre? great. Government owned communications monopoly with grandiose promises and an army of people and advertising, but little actual rollout.

Worst. Idea. Ever.

You think I'm being partisan? No. I would have opposed the idea no matter which party put it up. You've only got to look at various transport projects which are endlessly promised, re-promised, and then re-re-promised when a PR opportunity is ready. The Redcliffe rail link has been promised for 100 years now.

If you want fast speeds, then you need competition. Just like this article says. Not a quasi government department full of paper shuffling. Please kill the NBN and invite Google over. Give them tax money if necessary, but dismantle a nationalised communications network, stat.

It's interesting to see posts on this topic about turning the Internet into a utility, then see your post about how Australia seemed to try that and failed.

Utilities can be privately owned. Indeed, many power stations and transmission lines are privately owned. It should be a utility, it just shouldn't be owned by the government. Australia would have to be the only country in the 21st century trying to actively undo Telecommunications privatization by re-nationalizing the telecoms infrastructure. It's like all the 'baby bells' getting reformed into the Mama Bell, and all the new entrants being told they are now only retailers of a fixed price plan. It's nuts. But you'll never hear criticism from most Australian tech people because they fervently hold out hope that it's all going to happen and they'll get affordable high speed broadband. Just like people grow old and die hoping their highway will get upgraded or their town will finally get connected by a train line.

The problem with government ownership (and particularly with the NBN) is that the focus of the organization transfers from cost-effective rollout to political points and favors. Those regions which supported the previous government were given priority for the rollout, despite them being areas of poor commercial return. The lack of return from early customers compromises the cashflow of the organization. Because the technology was selected with a one-size-fits-all approach - madness in a country the size of the USA with the population of the Greater NY area - commercial consideration was again ignored, right down to paying $11 billion to an existing Telco to pull up working HFC in existing areas.

Definitely a plan is needed - the original case was for the government to lay down the backbone in Fiber and let the Telcos complete the costly last mile in whatever worked best. But this didn't fit the grand 5-year plan so beloved of committed central planners so it was expanded to a FTTN vision - famously undercosted, underestimated and underperforming.

The result is a disgusting mess that nobody wants to own up to, and meanwhile, I'm sitting here with the exact same crappy connection that was here 10 years ago, and no progress because committee this and study that. Politically I'm in the wrong area, and the existing Telcos aren't going to upgrade my line because it's only going to get torn up and replaced (supposedly) sometime soon. They aren't even allowed to advertise 4G mobile as a competitor to the NBN because the government needs every single person on board to have a hope of making a $50 or 60 billion white elephant look remotely commercial.

I live in Austin.

I have Grande and AT&T fiber to my house. I have service from Grande. AT&T won't sell me their 300Mbps service.

When Google comes, Grande is dropped like a hot rock.

I can't seem to find the article right now, but Google said their intent with Google Fiber was to prove it is possible. Further, to then force other companies to improve their service.

They seem to be on that second point and it seems to be working well.

What I want to know is when does everyone start to realize that if one can get service of 1Gbps for around $70 and not just from Google, why does existing much slower service cost almost as much or more than that?

I suppose this just goes to show that internet infrastructure is a natural monopoly and it takes a company with $400 billion market cap to disrupt the market and create actual competition.

I'll believe it when there is any better available service. I even live in a location where google fiber was announced, but so far it is just nothing.

I used to have a 3Mbit DSL connection from AT&T, I have now switched to a 6Mbit DSL connection from a local provider (who is using same wires). Sure, I'd love more speed, although I've grown to cope with the lack of it. I am just appalled at the prices. And now all of a sudden they are offering 1Gbps connections; if they like tiers so much, why not offer something "slower" for a cheaper price, I'd be more than happy to have a 50Mbit connection, do consumers REALLY need 1Gbps?

And what's up with San Francisco disliking fiber so much?

NIMBYism. No-one wants the street cabinets which contain the routers and what not on their street corner and are doing everything in their power to block them. Result = terrible internet speeds.

This is definitely a thing. Here's a link from my friend in Kansas, and how Comcast is preparing for it already:

http://www.speedtest.net/my-result/3479092379

Incredible what happens when you get a little competition!

I'm due to get fibre installed later this month. Can't wait to go from 10mbit/0.8mbit ADSL to 100/50.

A reasonable article on how cable companies can respond.

http://arstechnica.com/information-technology/2013/12/why-co...

tl;dr: by using more spectrum in the coax cable can get pretty fast... So pressure from Google should get them to crank up the speed sooner than later.

As mentioned in the article, this requires a DOCSIS 3.0+ modem, and there are costs to rolling out equipment. Cranking up the speed isn't simply a matter of turning the dial on their end.

However you are correct that coaxial cable has plenty of room for improvement with a cost that pales in comparison to rolling out FTTH.

Reply on news.ycombinator.com