My New FiberHouse in Kansas City
feld.com
feld.com
It seems to me like most bandwidth-critical businesses also require a datacenter anyway, and while fast to-the-home internet may enable the delivery of innovative technology (like high-definition gaming in the cloud) a fast connection to the home isn't necessary for the creation.
With that being said I do appreciate the concept: if building high-speed fiber can attract entrepreneurs (regardless of if it actually helps them), the incentive for municipalities to encourage high-speed buildout grows.
Having a 1gbps network would let you run more of your hardware at home where it's cheaper. This makes sense for certain specific cases.
For startups that need this power, hosting at home can save money. Also you can't beat the security of having physical access and control of your hardware.
For my app streamjoy.tv I run all the intensive tasks on a core i7 at home and then just push the rendered results to a $5 a month digital ocean vps.
For the kind of computation that I am doing it is much less expensive and faster to use my local hardware and push static pages to my nginx server at digital ocean.
If I had a 1gbps connection I would run even more of my setup at home. Its not that important for the features I have coded so far but in the future it would be helpful.
Can anybody comment on whether or not this any different with Google Fiber? Are they fine with you running whatever you want on that connection?
http://support.google.com/fiber/bin/answer.py?hl=en&answ...
(Whether they will enforce this remains to be seen.)
Really? You can't beat the security of hosting servers in your basement with a top-tier datacenter with armed guards, mantraps, insurance, fire suppression, redundant power and internet, etc.?
My AWS instances are likely orders of magnitude more secure than a server sitting in my house.
It seems like the grandparent poster really should rent part of a co-located rack instead.
I don't think there's much sense in hosting locally in a home environment with Google Fiber for almost any use case regardless of their ToS or the service's quality: for a system serving user requests a home doesn't deliver reliability or redundancy, for a lightly-utilized offline processing system a cloud provider is almost universally cheaper, and for a heavily-utilized offline processing system power and cooling costs become competitive with a datacenter relatively quickly (not to mention the cost of downtime - if you're at high utilization, downtime will cause your job queue to build rapidly and possibly insurmountably when you can't supply burst capacity).
For my use I need the fastest CPU + ram + IO I can get for as little as possible. The network does not need to be the best because this is not serving user requests.
I do not want to take the time to make my code run in parallel. An Intel Core i7 3770k with SSDs is the cheapest scaling I can get.
I use the same machine as a workstation for coding.
This setup saves me $180 a month at the cheapest dedicated host I could find + I have a blazing fast work station.
Yes, a data center has armed guards. Those armed guards are not loyal to you and they are not observable to you.
If you are a startup with a trade secret written in code that is incredibly valuable and absolutely can not get out, self hosting gives you a kind of physical security that is hard to get running your code on rented machines.
I am not saying that this is my case, but I could see a case where it would apply.
Lets say you have developed extremely profitable algo trading strategies. If they are the kind of strategies where millisecond execution is paramount, of course you have to co-locate.
If your strategy is not under that kind of time constraint I would want my algos on machines only I could access.
Production servers really don't belong in houses for many reasons besides connectivity and Google Fiber sure doesn't change that.
Very few entrepreneurs are moving here exclusively for Fiber, the real attraction is the super inclusive community and the extensive resources available to startup founders. Every single person I have met here is so unbelievably welcoming and will go out of their way to help you out. The Fiber is just icing on the cake.
Another huge reason people have been moving here is cost. My team just moved into a 1800sqft house for $1100/month and fiber ($79/month) gets installed in a few weeks. So much cheaper than Boston..
Oh yeah, and the Kansas Angel Investor Tax Credits [http://www.ksrevenue.org/taxcredits-angel.html] are great.
And I am paying less in rent than my friends in DC would pay for a broom closet.
But I think the idea of "arbitrary rallying flag" is a good one. This could be a really great way to evolve the KC tech community. They now have some validation (both Google and Brad Feld have invested there), and there is some excitement.
I think it's a good thing, and I'm excited to see what happens
Is not a qualitative change, but is a really important quantitative change. New applications that today we don't really know about will be made possible.
Living in that environment will make them able to "see the future" and design something for where the "puck is going to be."
As an example, Mr Steve Jobs "saw the future" with the network system of Apple, and Microsoft used a supercomputer to simulate their future programs when computer power were ready.
As for the network speed, plenty of colleges and research institutions have hundreds-of-megabits and higher Internet connections straight to the terminal, not to mention datacenters, Korea, and Scandinavia: I'm pretty sure we know what's possible over a gigabit network and have been able to imagine the "what if everyone had this at home" scenario for a long time.
To your analogy, it's like Microsoft using a supercomputer to imagine a future when consumers own powerful computers, except the supercomputer is a $50 gigabit Ethernet switch that's been available at your local Best Buy for years.
I don't dispute that this is a productive idea for entrepreneurs, that revitalizing the KC tech scene is a great idea, or that Brad Feld's hacker-house offering is really neat, but I still can't see how moving to an environment that's already existed in other niche areas and is very trivially simulated is going to create revolutionary ideas.
That said, the hope is that this new, fat internet pipe will yield creativity and new services/technologies that we just can't predict yet. We have to push these technical boundaries first, because that stretch yields new products and groundbreaking thinking.
History would seem to suggest that that's one way to help lay the groundwork for new and powerful inventions. There are literally innumerable products and technologies we use every day that are many iterations away from the enabling-technology, but would have never been discovered or applied without the initial technical push.
However, I don't see why making it in KC and having Google Fiber makes any difference? Will I be able to do anything I previously couldn't with my 50/25 FiOS connection?
I do know that hosting a server is prohibited by Google Fiber TOS, so I'm not sure how this is going to work.
Down the street is another house shared by several more startups working on things like real-time search [2], retinal authentication [3], etc. There are other similar endeavors scattered nearby, plus some web-dev shops, etc.
We just rebranded threed.ee to gethandprint.com and moved our entire team of four from Boston to KC!
@gethandprint
Also, the city is wonderful. Beautiful, amazing food, great schools, cheap, clean, and great transportation options for vacations (out of a great airport!). As someone who lived there for 18 months before moving to San Francisco, there is a lot about Kansas City that I miss.
Here are some downsides (I agree/disagree with you in some things):
1. Crime rate is astonishing. There are deadly shooting nearly every other day. In fact, last one i heard involved 2 deaths at Westport, which is right where Google Fiber planted its roots.
2. City is beautiful in early spring. Summers are incredibly hot and humid and burn the entire city to the ground. Winters are freezing-cold.
3. Schools are no different than anywhere else. Some good. Some bad.
4. Public transportation is non-existent. Unless you consider waiting for a bus for 40 minutes a good transportation system. Any attempt to put rails to use failed miserably. KC is ultra-conservative. Nothing will ever happen here unless it has something to do with the Chiefs & Arrowheads Stadium. Bottom line is if you don't own a car you will have very difficult time getting around. Some neighborhoods don't even have sidewalks. Walking to a store 2 miles away would mean walking on public roads and dodging cars until you find a sidewalk.
5. This is something that everyone who lives here for more than a year complains about. There is nothing to do in KC. It's the same people and same places. Summer time is more event-friendly. Winters are dead. Nothing to do but go out to drink and eat somewhere.
6. Majority of people are overweight... and don't care. It's actually scary how much of the population is overweight. That's because you're forced to drive everywhere.
And some upsides:
1. It is VERY clean. Almost pristine.
2. Living is VERY cheap even in the center.
3. Food is amazing.
4. Airport is very convenient and conveniently placed.
5. Great place for sports. Chiefs will always have fans here. Regardless of how bad they are. Tailgating is amazing. Unfortunately, every team except for Sporting KC has been miserable for decades with no signs of recovery, but still have very solid fan bases.
6. You're 2/3 hours away from everywhere in the country.
Take what you want from this. If i were in my early/mid twenties and single I would never move here. If i had a family or looking for a place to retire, I would definitely consider KC. IMHO, if KC wants to attract young talent like Boston, New York, San Francisco or Seattle it'd have to be a bit more engaging than getting fiber installed in every house.
[1]http://www.nytimes.com/2011/09/21/us/kansas-city-mo-school-d...
As a northern Kansas City resident I am patiently waiting for Google fiber expansion to my area (Currently in "Coming Next"). Having an extra bedroom I am excited to make my home available to the Homes for Hackers project as soon as the fiber is available. It will be great to see a startup progress from my couch.
It's not even that easy to get 1gb traffic to flow over the lan, though of course its pretty easy to get 70%-80% or so. To get sustained gigabit it pretty much has to be an ssd or array on both sides. And this is from the perspective of someone that runs three 3x1gb port channel bonds at home between a multi-tier san and a few virtualization hosts.
I could upgrade my home internet from 85 to 150 or 300 but i have no desire to - it'd pretty much be impossible to tell the difference 99% of the time.
Will this always be the case? Nope. 10 years ago it was hard to push 100mbit datacenter to datacenter. As computers & io speeds catch up and content targets higher speeds something like 1gb at the home will become much more useful.
But I sure wouldn't move to kansas city just get 1gb right now.
I know that the offered amount of bandwidth is drool-worthy, but realistically how much of it can be used without drawing the ire of Google and getting shut down?
The reason is that I live in a rural area about 20-30 miles from major cities. If you go another 10 miles your only option is satellite or dial-up (if you have phone lines to your house).
I'm considering switching to cable, but we will see.
[1]. http://www.pts.se/sv/Dokument/Rapporter/Internet/20122/PTS-b...
I live in average sized town in the Netherlands. I pay Euro 50 per month ($65) for 100 Mbit up & down. That price includes unlimited bandwidth (placing servers is allowed), full cable, and a HD DVR.