February 19, 2014

google fiber

Filed under:, , , — cwage @ 11:52 pm

Because I'm a contrarian (or should I say buzzkill. get it? buzz? kill? google buzz? get it? SEE WHAT I DID THERE), I'll play devil's advocate to the frenzied excitement over Google Fiber in Nashville. Some supposed factors driving the excitement:

Fiber is fast
Well, fiber is a physical medium that is quite capable of fast transfer. But a network has to be built to take advantage of that. This involves several points of expansion:

First, your own network infrastructure has to support these gigabit speeds. Maybe you're rollin on gig-e wired throughout your house, but if you're like me and 99.999% of internet users in Nashville, you're using a macbook on the same shitty $50 802.11n WAP that everyone else uses -- which maxes out at about 300Mbps and usually clocks in more realistically at 100-150Mbps. Yes, this technology will evolve and grow. When? Pro-tip: not till more than 3 trial cities support these next-generation speeds.

Second, all the services you use now that ostensibly would benefit from 1Gbps bandwidth -- a surprisingly short list, currently -- have to scale their own infrastructure to support all these new clients desperately wanting to saturate their shiny new broad pipes. Will these vendors -- Dropbox, say, for example -- have incentive to do so? Well, maybe. Are you willing to pay for it? Are enough people transferring big enough files that people are willing to pay a competitive rate? Maybe -- over-the-wire backups are a clear need here.

Third, google itself has to have built the infrastructure to support -- not everyone at 1Gbps at once of course -- but an infrastructure that will support the average number of users using their fiber service and some margin for bursting/excessive load/etc. This may sound like it goes without saying, but consider how often wireless carriers roll out a New Bigger Faster 8000G wireless platform but clearly don't expand the network capacity behind it (I'm looking at you, AT&T. and Sprint. and, well, everyone.) Did/will Google scale this appropriately? Maybe. Hopefully? Who knows. Minor but real concern.

That said, yes: there is an "if you build it they will come" effect wherein larger capacity spurs innovation. But it won't be some magical lightswitch that flips on once you have something with "fiber" in the name such that everything will be 10X faster. It will be an incremental, evolutionary path of growth -- which is good! -- but not magical or instant. Just like the transition from the days of ~56k/ISDN to broadband was not magical or instant, this transition too will take time. Maybe add an exponent or two here allowing for Moore's law, sure.

Video streaming with cable sucks -- fiber will fix that
Video streaming with cable (e.g. Comcast) sucks because Comcast is throttling services that threaten Comcast's dwindling legacy cable bread&butter. This gets into the territory of advocates for so-called net neutrality, but that's a debate for another day. So, it's great that Google can provide some competitive incentive for Comcast to evolve or die. Conversely, though, I see no indication that Google intends to do anything other than enter the fray of this very competition for the same media services and thus be incentivized to perform the same sort of throttling of their competitors. In short, streaming problems have long since ceased to be a problem of technology: Netflix's HD streaming (for example) maxes out at around 5Mbps. The remaining problems of shitty user experience now revolve entirely around competitive throttling. Google may or may not be just as bad.
I'll finally have the upstream/symmetric bandwidth to host personal video servers!
Sure. Explicitly against Google's ToS for the service, but hey. How aggressively will they enforce this? That remains to be seen.
Free internet!
TANSTAAFL. Ask yourself why Google is willing to offer a tier of free internet service. This is a company that has capitalized on the decreased privacy of its customers at every opportunity. The only conclusions you can draw are ... uncomfortable. How will they treat your data? How will they sell it? How will they treat traffic that ensures privacy (VPN/SSL/etc)? Will they even allow it? There's a good case to be made (sadly) that so few people even care (or are educated enough to care) about their privacy such that the few people who do will be a non-issue.

To sum up: competition is good. Even competition from Google is better than nothing. But you won't find me frothing at the mouth or fawning over the mere prospect. Google is a company that wants to compete with everyone by doing everything. This means they not only compete with Comcast for internet access, but they will (in all likelihood) compete with Comcast (or anyone else) for the same legacy content that has motivated Comcast to depriorize/throttle its content and generally be shitty at the one thing they're supposed to be good at. It also means (because Google isn't particularly specialized) that they're probably going to suck at it, and/or sacrifice innovation, support and progress for its multitudinous other endeavors (witness gmail). How many products with satisfied customers have already been sacrificed at the altar of Google+? Why do we believe this will be any different?

It's possible that Google could surprise me by being focused and efficient in providing this one service, but they're a company that's spent the last 5 years doing the exact opposite.

High-speed fiber internet access provided by a competitive company with no interest in selling me anything other than reliable, fast, hands-off internet access? Now we're talking -- I'd be excited about that. Oh, Google? Yawn. How long will I be using that product before they decide to abruptly EOL it? (Yes, I'm still bitter.)

BOOM. Debbie downer out. *drops mic*

UPDATE 1: -- some good comments from facebook:

Scott:

I got called a "pessimist" for even hinting at the points you bring up.

You bring up a lot of technical points (and there is _some_ counter argument to them) but the bigger reason why I was non plussed about the announcement (again, more providers in a market is a good thing) is less about technical and more about economics and consumer empowerment.

The best thing that could happen is Metro opens up it's considerable fiber assets and leases them to service providers. Not only do the tax payers own those resources, but it keeps control and ownership here in town rather than Palo Alto. This can be done in a cost neutral model and in some examples, it even generates a profit for tax payers (shudder).

Having Google come to town is good but as you say, there's nothing indicating the they won't be the same oversubscribed and low bar customer service as the existing oligarchy already is. If Nashville really wanted to be respected as a "entrepreneurial tech leader" then it should be insistent on some things. Infrastructure issues you raised but also free access to the poor, ubiquitous coverage across the county (or MSA if we're really thinking ahead), and the ability to extend the wireline into wireless services for all citizens.

Let's have some self respect and stop slobbering all over ourselves because some popular girl looked our way.

cwage:

agreed, and well said. The technical aspects I bring up are not so much to dismiss the progress, of course -- I'm not a luddite -- but just to point out that this is nothing new. Google didn't invent fiber -- as you say, we've got shittons of it, which we paid for, sitting dark. So yeah, excuse me if I don't fall over myself faint with praise because google throws around the words "fiber" and "gigabit".

Knowing this and then reading the cryptic (and sinister in its ramifications, depending on your opinion of our current metro administration) language about how google is "evaluating" the cities and that they'll "work with our city leaders" is ominous. What exactly is there to "work with" them about? (hint: incentives). Which I guess we're supposed to be excited about and fall all over ourselves to support instead of asking ourselves "hey didn't we, like, already pay for a bunch of fiber? why are we paying more to convince google to do the same thing?"

Google is very clever indeed, using the aura of privilege and prestige to "select" cities to deign with their service instead of just launching it. This enables them to court their suitors and cherry-pick the very best of the state-subsidized incentives.

and I have no doubt our mayor would railroad this through as quickly as he has our other bread&circuses with nary a thought to what actually makes sense. So then we have a big-ass national company with government incentives serving us internet. Great. How is this any different, exactly? Except that google is also trying to do 23948234 other things? Meet the old boss, same but infinitely more distracted as the old boss.

But hey, at least we'll be really cool -- woo woo, #itcity.

To clarify succinctly: I don't care if Google wants to come compete. I have no problem with that. The excitement to me is ill-informed, and the minute that our political class attempts to turn that excitement into government-subsidized incentives for Google, you'll see me raise holy hell. (Or, more realistically, post a whiny blog entry and go back to work.)

  • Colin McCormack

    Someone needs to come up with a filter that converts 'fiber' to 'fibre' and sink that into the infrastructure (or infarstructure, if you use US-conventional spelling.) I will do it. I'll call it Filtre.

  • http://www.nashvilletechfeed.com/ David Beronja

    It's either Google Fiber or the city builds a public utility fiber network. Since we have crazy people who think anything government does is bad the second option will happen probably never in Nashville. The existing providers have shown no interest in upgrading their networks or anything close to decent customer support or even care beyond profits.

    I hope that if this happens Google Fiber comes in honey squares.

  • TheNuszAbides

    "This gets into the territory of advocates for so-called net neutrality, but that's a debate for another day."

    does that mean i won't find such a debate by digging back a year or few into your archives?