PhotoDude.com

Let’s remember that DSL technology, at least as we understand it in the US, is a kludge to run an extra service over the analog telephone wire the Bells have already run into your house. The last time I saw a Bell run telephone wire in my neighborhood was in 1975, when the Bell’s name was Ma.

Other places haven’t had universal telephone service in private residences for nearly as long as we have. The bad part is generations without the convenience of your own phone; the good part is that when you do get it, you get much more modern technology.

I was under the impression that a very large percentage of the voice telephone system was digital, both in Western Europe and in Japan. That would explain why ISDN was comparatively widespread in Europe in the late ‘90s, while only the most irrational American nerds were willing to spend potloads of money for a digital phone line+minute charges to get only twice dialup speed.

But if the existing infrastructure is digital end-to-end, shouldn’t that dramatically lower the cost of delivering IP-based services of all types? You wouldn’t have to have converter modems at the consumer end, just dumber/cheaper terminal adapters; and it would eliminate some percentage of the ISP expense of fancy DSLAMs at the phone company central office.

Unless my impression is false…or maybe there’s some other major point I’m missing?

As long as it’s all TCP/IP, it’s all the same Internet; they just have faster (government subsidized?) pipes to it than we have. OTOH, if someone can get me 50Mb + VoIP for $40 a month in the SF Bay Area, I’ll take them out for barbecue.

Recent articles over the past few months have intriqued me. It wasn’t too long ago that “first world technology” was one of our biggest assets in the world economy. But it seems like the rest of the world figured that out too and wanted to know “whar’s mine?”.

One article over here
“One is a bandwidth glut. Some half-dozen fibre-optic submarine cables have been completed in the region over the last few years. One example, the Australia-Japan fibre cable, which connects the two countries and is interconnected to submarine cables heading to Europe, other Asia-Pacific points and the United States, has a design capacity of 640 Gbit/s, more than 500 times the capacity currently available from or to the East coast of Australia.

A second trend is liberalization. Most of the region’s advanced economies allow competition in the international Internet bandwidth market, which has driven down the costs of connectivity. For example, the Internet Initiative of Japan is paying one tenth of what it used to pay five years ago for Internet connectivity .

A third trend is that telecommunication operators got more comfortable with the Internet. Many had inherited academic networks and were new to the world of the Internet. Over time, they have begun to realise that their own traditional circuit-switched networks would evolve to Internet networks. These traditional telecommunication operators are now investing heavily in Internet infrastructure.”

It seems like maybe we haven’t invested in infrastructure like, say, the Asians have. Mebbe they’re on to something?

I’ve still got my $59/mo, 3 mbps (which actually tests out at around 2.2 mbps) DSL connection. But I tried out a 3 meg cable connection, since they were offering it for $22/mo for 6 months. It rarely pulls in download speeds of less than 3.3 mbps. And in a few months the basic service will go up to 4 mbps with 5 mbps for $10 more. A year from now, the offering will be 5/6 mbps. I already killed my 2 biz lines from Bellsouth for VOIP. I think bandwidth is coming here, more likely from the cable companies, who would be just as happy if they could kill the bells. But how good will it be if we’re still living in a 90’s infrastructure? A lot of huge linux distros I suck down from university mirrors still come in at ~80KB or so, no matter what I’m using.

I feel your pain, Richard. I’ve been testing the new debian-installer-rc2 this morning, and I’m averaging 80-90K from the Debian mirrors..sigh. But then again, lots of folks are hitting those mirrors, and they have to throttle. No biggie anyway, as I’m just setting up a testbed on an old K6 400.

As for the bandwidth discussion, there are huge differences in the situation for telecom between the U.S. and Japan. To start with, if you’re unhappy about your DSL offerings, thank your federal government, both the judicial branch and the Congresscritters, who think that market economics are the way to manage national infrastructure. That’s always been a dopey idea in my view, and the long, sad history of American telecom through the 1990’s is all the proof of my assertion I think any reasonable man should need. All telecom deregulation has really done for this country is create a huge amount of marketing garbage for our landfills, and kept some people working in the ad biz, at the expense of thousands of lost jobs in telecom. The rest of us are still stuck with quasi-implemented System 7 signaling designed 30 years ago, and a bunch of myopic Wharton biz types badly managing the Bell orphans, and playing “buy me” with other shell players created by the deregulation scam. Reminds me at times of the S&L meltdown of the ‘80s.

But in Asia, the population densities are higher, which facilitates high speed services at lower average per capita costs, because fiber isn’t run across vast expanses of places like Kansas and Utah. And NTT and the Korean telco operators didn’t spend most of the 1990’s plowing huge amounts of investment capital into major dark fiber projects that will never be lit, like their American biz whiz brethren did. The Asians built sensibly, and incrementally, but they built, and now they can still afford the switch toys and goodies to feed their upgrade fiber infrastructures, while we still think dialup access is OK for 40% of our population.

Aw, don’t get me finished. Reid’s tryin’ to run a respectable blog here, and I could go postal on the subject of American telecom, to no one’s advantage. ‘Cause it ain’t the donkeys or the elephants that screwed this up so bad it’s not gonna get fixed, it’s all of ‘em.

I just love it when I write an article mostly trying to be wry and witty riffing off a base topic, and along come three people to offer sound, expansive, and genuinely interesting reasoning about my catalyst for silliness.

It makes me feel so … dumb. But I’m glad I’ve got smart readers to make up for it.

Want to feel even dumber? With a little anger and frustration thrown in? Just wait on hold for an hour or so getting Earthlink to give you the free 3.0 upgrade.

On the other hand, these days with fast computers and big hard drives loaded with cache, I’m not sure I notice 3.0 over 1.5 for the average web page I look at. And I’m guessing my linux downloads from GA Tech aren’t coming in any faster, either. Honest question: what does one do with 50 mbps download speed? My 100 mbps LAN rarely gets over 20 mbps transfering files, due to hard and software limitations. Here’s my guesses: (anyone feel free to jump in)
1- video, music and virus swapping (none of which hold any interest for me)
2- you have around 30 mpbs upload, so you can realistically host a commercial or other website (even with apache I’d be concerned about security)
3- streaming video at 640×480. Uh, I think that’s why I have cable tv.

Did I miss any? Are we getting funny yet?

If I’d known Earthlink was going to boost their 1.5Mb customers to 3Mb for no extra money, I would have stayed; particularly given the bait-and-switch that Speakeasy put me through when I switched last spring.

Even though I STILL HATE EARTHLINK for the bogus “static IP address” $10/month upgrade, which is actually a static LAN address through PPPoE, and so is nonroutable, and so is utterly useless for a server.

Scoff if you will, Richard, at the IP/TV initiative, but it’s big in Asian planning. Directly, the “DSL Serving Prep for IP TV” article at the IPTV dictionary may answer your question about the utility of large pipes to the home, but it basically boils down to the bigger the pipe you have, the more individual TVs watching different channels you can have. Some major players have already rolled out IPTV services to millions of users worldwide, and this whole thing is gaining ground pretty quickly in the Far East for a several reasons.

First, they already have the infrastructure deployed to significant portions of their markets, as we’ve noted in this thread above. Second, most of that infrastructure is native on IPv6, which handles multicast much better than IPv4, and that’s a big enabler for IPTV on a mass scale. Classic case of being handed a lemon, and making lemonade here, as the Asians, who lacked IPv4 address space for years (while watching American committees debate American address block assignments, like the one where HP fiddled and fooled around for years trying to decide if it was “feasible” to turn back the Compaq/Digital Class A 16.0.0.0 net they weren’t really using anyway )went ahead with IPv6 deployment on a mass scale. Now look who’s laughing…

As an old broadcast engineer, I didn’t “get” IPTV for a long time, and I suppose I still don’t, like I don’t really “get” the Web either. Both look like fundementally silly mechanisms to me, but I do get that they look like big opportunities to millions of other people. Here at home, SBC has recently announced an initiative to do something along the same lines as the Asian telecom operators have done (but on the cheap, using a 12,000 foot service perimeter for their version of FTTN, given the geography of their subscriber base, and the costs of deployment and maintenance), but it looks like another salvo in the cable/telecom wars to me.

I predict the whole thing will shortly result in some big equity exchanges between SBC (and maybe Verizon, which is also aggressively pushing its new Fiber To The Home (FTTH) plans) and some cable operators, and few people really getting new services. Many industry analysts think it is just too expensive to the telecom operators to really do FTTH , and I tend to agree. FTTN and FTTC don’t really accomplish what FTTH offers, either, which is a big opportunity to dispense entirely with IP, in favor of the much, much more efficient ATM protocols .

That’s really what I want, FTTH/ATM to my home, and everywhere else, and IP only at the end points, if even there. Some Asian operators are actually looking at doing that, and it makes terrific sense, if you have real FTTH penetration. ATM is so far superior to any IP based scheme in terms of performance, security, manageability, robustness, etc., etc., that they’d be crazy not to do it, once they get to about 50% market penetration of FTTH, and in Seoul, they are already there. Once an ATM conversion happens, countries that have that infrastructure will automatically get huge network benefits, as ATM proponents have been touting for years, and it will become clear what a big dump truck full of stink IP is.

But not to worry, Richard. We’re Americans, and we’re old enough that we probably won’t live to see it happen here. No, my friend, we’ll go to our graves routing IP in little niggardly clumps, and waiting on Linux mirrors and ATM gateways far away to dole us out a piddling few packets a millisecond, while they stream millions of packets to better (ATM) connected folk in the same interval. And the really galling bit will be that most of the IP packets we do eventually get at our end points will have been transmitted back by American ATM networks, to be again reconverted to IP wherever the fiber stops near us. Most of us Americans will live within sight of the ATM network endpoints, but will never taste that water.

Because of our geography. Because of our decision to put telecom in the hands of market economics. Because the market we’ve so entrusted with that responsibility can’t tell a fiber trench from a dry hole, and will remember pouring a trillion dollars into one, thinking it was the other, through out the 1990’s, and will still be telling bitter stories about it in 2050.

Because we’re dopes.

Aw, don’t get me finished. It’s Reid’s house.

Very informative, Paul. Good read. I’m not sure of the use of IPTV either. And especially after renting “Lost in the Translation”, I think you’d have to drag me kicking and screaming to ever go to Japan, much less watch anything they had on tv. It’s the old “10000 channels and nothing to see” syndrome.

On the other hand, it does look big. When you look back at the inventors and technology that made us great, Edison, Henry Ford, etc., in the twentieth century, you’d like to see “American Ingenuity” pushing us forward today. Instead of turning our backs on the world and richly rewarding the purveyors of the status quo.

Methinks we’re the new Great Britain of the 21st Century, and whether we like it or not, a tremendous opportunity has been squandered.

“It’s Reid’s house”

Trust me, Reid doesn’t control as much in his house as he’d like you to believe. And you are welcome at this little podium any time, especially when you bring a boatload of knowledge on a topic where I was just pokin’ around.

Peter: “If I’d known Earthlink was going to boost their 1.5Mb customers to 3Mb for no extra money, I would have stayed”

Peter, shhhhh. It’s a secret. I’m an Earthlink DSL customer, and I only know about their speed bump because I hear rumors on the Internets. No one that I have a business relationship with has actually suggested to me they can offer me a vastly improved service.

Perhaps because then they might have to.

Secondly, when it comes to my Earthlink DSL connection, I am like a frozen rabbit caught in the high beams on the Information Highway. At its assigned speed, it works amazingly well (140kb per second on average). I have managed to keep my contact with Earthlink limited to my credit card and their emailed receipt for nearly eight months now, because when it increases above that minimal amount, things stop working.

I can’t explain it, it just is. It’s been that way for years. Asking Earthlink to improve my bandwidth for free seems like asking for more trouble than I’m willing to take on right now.

Richard: “Honest question: what does one do with 50 mbps download speed?”

I don’t know. But then, I don’t know what I’d do with $10 million, either. Once I had it, I’m sure I’d figure it out, though. The fact is, until you have that kind of quantity, your mind doesn’t even open up to the possibilities. It’s a horizon you can’t even see.

What kind of applications would programmers begin to work on if they knew the vast majority of their audience had big fat pipes, whereas they now start with the base assumption, “we must accomodate the 40% of users still on a 56k dialup modem”?

I think bandwidth is a lot like money and storage space. No matter how much you have, you’ll find a way to expand your usage to fill the new capacity. And soon thereafter wonder, “how did we ever make do with less?”

Do you remember what it was like to upgrade your browser in 1998? A full hour of the agony of 3 kb per second to collect the 7MB of Netscape 4.051501. Now we download nightly builds, and even over my measly 1.5Mbps, it’s effortless.

But it does bring to mind another question. When you can download at a rate of 250mb a minute (15 GB per hour), how big will your hard drive(s) need to be? It’s the same question I have to ask about upgrading cameras. Bigger files are great. Until you think about the fact your storage needs just increased by the same factor as your file size.

Paul: “Because the market we’ve so entrusted with that responsibility can’t tell a fiber trench from a dry hole, and will remember pouring a trillion dollars into one, thinking it was the other”

I have to admit, I’ve wondered about that. There was about a two year span where you could not make a three mile trip in Atlanta without encountering some crew pouring money into a hole in lengthy insulated pieces. I figured all this alleged fiber would soon cuase me to get an “upgrade notice” when they replaced my copper. Now I have to ask … does anybody really know what was encased in that heavy insulation? They claim it was fiber…

...but the only improvement I’ve seen in my DSL connection since I got it in September, 1999, was caused by a lightning strike on my line. Sort of. It screwed up the line enough that the BS tech simply switched me to a new line on the nearby remote DSLAM across the street. My “copper run” went from 13,000 feet to about 1,500, nearly doubling my effective speed.

Somehow I view Earthlink safely doubling my speed again about like safely getting hit by lightning twice.

I don’t want 50mb download speeds, but I do want 3mb uploads for one reason only…so my wife can run her little internet radio station at a decent audio rate for her 20 loyal listeners, and do it out of my home. And that’s where the providers are clueless. They think people are dying to download feature-length movies to watch on their PC or rebroadcast to their $5000 digital plasma bigscreens (and some do, I’m sure). I think people want just as much to be able to do their own A/V broadcasting / multicasting, but upload limits prevent a lot of it.

By the way, if you think provider support sucks now, imagine an ATM backbone that runs up to the NAT. You can’t find 10 guys in the support center that understand a 30 year old technology like Ethernet and TCP/IP. There probably aren’t 10 in the whole of Earthlink that understand this new-fangled ATM.

PS Most of that fiber is unlit, Reid. The Great Tech Bust of 2000 has at least one lasting legacy and that’s miles and miles of fiber sitting in the ground waiting for the photon router. Boy, we were sure dreamers back then.