Slashdot Mirror
10Gbit to the Home by 2010
womby writes "Nihon Keizai Shinbun report (Japanese) that NTT, Fujitsu and the Japanese Government are forming a working group to develop internet technologies that will hopefully allow homes to receve 10 gigabit internet connections by 2010.
'The Japanese government (the Ministry of Public Management, Home Affairs, Post and Telecommunication) are going to start a development plan next year that will increase the speed of the internet in Japan to 100 times faster than the current 100MB fibre internet, with partner companies it is aiming for completion by 2010.' A complete Translation is here, if my blog gets beaten into the ground try the Coral Cache Link."
'The Japanese government (the Ministry of Public Management, Home Affairs, Post and Telecommunication) are going to start a development plan next year that will increase the speed of the internet in Japan to 100 times faster than the current 100MB fibre internet, with partner companies it is aiming for completion by 2010.' A complete Translation is here, if my blog gets beaten into the ground try the Coral Cache Link."
First, we need hard drives and system buses that can get the data moving at this speed.
When you look at the state of the world, how can you not become a radical, liberal anarchist?
People said that about 14.4, 28.8, 56k, and DSL.
By the time we have 10GBits in the home, porn, warez, and Linux distributions will hit a size large enough to make that not the worlds greatest connection.
It's always been that way.
Vonal Declosion
Put it out there, and people will find a use for it. Let's not fall into the trap of thinking that because we can't imagine how someone would use it, that means that nobody will find a use for it.
ABSURDITY, n.: A statement or belief manifestly inconsistent with one's own opinion.
Each human eye has about 4K x 3K retinal receptors triggering the optic nerve at about 40Hz. Assigning 2x2 32bit pixels to each, at 60Hz, is 2*8K*6K*60*4 bytes per person, per second. That's under 24GBps, with hifi audio channels and metadata, it's still under 25GBps per person, before our senses can't tell the difference from more data. 2:1 compression means 12.5GBps, or 100Gbps - only 10x more than these plans. The end of multimedia data networking might be just over the horizon, at least for one person at a time.
--
make install -not war
Take a stopwatch and try to see what 1/100 of a second is before complaining. BTW, but you do know that nothing can go faster than the speed of light, right ?
If you think about it more carefully. The architecture of the internet is likely to mirror the rise in home network speeds. If every user has a 10 Gigabit connection, then it is likely to be capped, at least until servers can handle the greater loads. A lot of what you wrote sounded like fear, uncertainty, doubt.
People said similar things about adoption of adsl use and its growing popularity. It didn't suddenly make the internet a series of dead servers.
A determined person can always cause havok on the internet. However it is not likely to crumble around your ears. Problems exist for connections today due to the holes in operating systems and the increase in speed as to which a service pack can be downloaded (in the case of Windows) and a virus getting onto the machine, I believe will always be there. Therein, lies the greater problem.If server speed increases at the same rate home connections do, then the risks will be less.
People may not need a faster internet connection, not for the size of data transfer, but for speed. Also may give people the option to host their own servers, which would interesting for most people not just nerds. Video on demand could become a feasible reality, possibly even generating a new generation of amateur 'tv shows'.
Like the internet has opened up computing for people who are particularly interested in pc's this could herald a new type of user or social networking and a new age of the internet where more people are involved in its infrastructure. Therein may lie the risk and the benefits.
Personally I think there are risks inherent, spam being the greatest problem if greater upstream is available. Not that I am implying damage, I am talking about wasted time deleting or reading through increasingly sophisticated spam messages. Even if servers do increase in bandwidth enough to combat DDOS attacks (which I think you are aiming at), the increase of spam messages, is not always something that can be easily ignored, since sometimes messages do get through. With bandwidth, DDOS can be ignored, spam cannot always be ignored. However, despite this I think this is a good advance and can only benefit. Limiting these services to the chosen few (students I believe are popular candidates for creating 'bot armies'), is not likely to improve matters for the determined few.
We can all build nukes! Forcast weather for the whole planet!
If we all build nukes, I'd say the forecast is cold and overcast, for the next few hundred years.
Lost: Sig, white with black letters. No collar. Reward if found!
Sure some local hardware could be improved, but the bottom line is that if you have a decent broadband connection, you're already getting far better latency for many local servers than you'll EVER get for ones across the globe, or even across the US, just due to the speed of light.
Even if cables transmitted at the speed of light (which as you say, they don't, it's generally at least a little slower) the time for a photon to get from here to the other side of the planet along the surface is something like 2E7meters/3E8 meters/s, or 67ms. So to ping a server there you'd HAVE to get something over ~135ms as an absolute limit. Throw in actual speed of light in the cable and the fact that the cables probably aren't "as the crow flies" and even without finite latency in the routers you're going to do worse than that.
I just pinged www.hinet.net (a taiwanese ISP) with my cable modem and got an average of 165ms and mantraonline.com (an indian ISP) and got 290ms. It's just not gonna get that much better than that. Sure there are probably plenty of podunk ISPs in Asia that could use beefed up connections and better routing or whatever, but for many cases, we're already doing shockingly well.
Getting bigger pipes to combat bigger DDOSses is not a valid solution. If a regular, run of the mill website needs a 100Tbit pipe, 1% of which is for legitimate requests, and 99% of it is to make sure the legitimate requests get through, there is a problem
History will be kind to me, for I intend to write it - Sir Winston Churchill
No, of course it's not necessary. We as a species have survived without it for 1.3 million years. For that matter, cars aren't necessary, hell, horses aren't necessary. But all of that isn't the point. The point is we have an opportunity to be able to transfer more data, do things better. We could keep doing things the old way, but if that was all we had ever done we'd still be sitting around a fire happy to be eating rabbits. Improvements aren't a bad thing, just because you don't see the immediate gain doesn't mean there wont be one.
WANNAWIKI Wannawiki WannaWiki WANNAWIKI!
You must have been using Suse. Debian Sarge (the upcoming release) is 1 CD for the basic install, and 11 CDs more if you want to do more than basic things with it. These values are for ia32-CPUs, the coming 64-bit CPUs might see a twofold increase in binary size... (I don't have any data on that, though...)
Regards, Ulli
Simple things should be simple, complex things should be possible.