On Tue, Feb 26, 2002 at 12:36:37PM +1300, Mark Foster wrote:

...

From my point of view bandwidth is very cheap now. Of course this is coming from first connecting to the Internet when we were being charged around $600/MB ($0.60 per kilobyte) for (international) traffic. Now I'm paying, what, about $0.20 per megabyte (less in some instances) for international traffic.

My problem with this is that you pay $0.20/meg for *traffic* regardless of its target.

No, you don't, unless you've chosen Telecom as a provider. You pay one tenth for domestic traffic, last I looked at my pricing plan. Ewen's talking about TelstraNameOfTheWeek cable modems.

Bandwidth has dropped in cost, but we download more, and faster, nowadays.

Oddly enough, you may find some relationship between these facts. -- Rodger Donaldson rodgerd(a)diaspora.gen.nz I've been bitten at least as badly. I poured 7 years into helping make SunOS 4.x a great operating system only to have it ripped away and replaced with that Solaris thing. -- Larry McVoy, in linux-kernel - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog

On Tuesday, February 26, 2002, at 03:28 , Mark Foster wrote:

At 07:38 27/02/02 +1300, Rodger Donaldson wrote:

...

On Tue, Feb 26, 2002 at 12:36:37PM +1300, Mark Foster wrote:

...

Bandwidth has dropped in cost, but we download more, and faster, nowadays.

Oddly enough, you may find some relationship between these facts.

I do see the relationship, and people who rant about how its all too expensive irritate me - but I find myself having to agree with them to at least some degree. Id really seriously like to know what kind of margins there are in the Jetstream/Jetstart packages - because I know from an ISP point of view, the overhead is pushing the boundaries of what is genuinely profitable.

The internal revenue characteristics of the Jetstream/Jetstart packages to Telecom are irrelevant. Telecom is not a charity. The problem is that competing with Telecom requires a fresh access network build. That's an expensive thing, particularly if your target market is spread out over a city the size of Auckland, the revenue per household is likely to be relatively low, your major competitor already has an access network which was funded by the taxpayer, and you are a carrier drowning in a culture of telephony engineering. I think that mandating unbundling of the local loop would result in more competition, leading to choice and cheaper alternatives to the Jet* packages that seem to be universally hated. I don't agree that it's too late to bother with, either; even if you optimistically assume that the telephony mindset in the non-incumbent carriers is long forgotten, building out to local exchanges and installing your own DSLAMs is a much more straightforward exercise to fund than a disruptive deployment like Richard has described. There are probably at least ten ISPs in Auckland who could start deploying DSL service straight away, if they had access to the copper at the local exchange. I don't suppose there are ten ISPs who are currently planning fibre-to-the-home. Joe - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog

At 11:34 a.m. 27/02/2002 +1300, Don Stokes wrote:

Joe Abley wrote:

...

I think that mandating unbundling of the local loop would result in more competition, leading to choice and cheaper alternatives to the Jet* packages that seem to be universally hated.

While others have suggested at least 10 ISPs would get into the action, I still don't like unbundling. In my (narrow) mind, the local loop is history. End of story. Its not just the cables (old construction, water in it etc, but the terminations, dodgy splices, etc. But also that at best xDSL will never deliver the sort of bandwidth that we as an industry want to see in the as soon as possible future.

Frankly, I don't see DSL over voicegrade copper as the long or even medium term solution to the problem of providing ubiquitous broadband service. At best, it's a stopgap.

At last !! someone agrees with me.

What's happened in Wellington in particular is that Citylink & Saturn built "next generation" local loops -- cable TV, fibre -- in parallel with the voicegrade copper links. As Richard noted, Wellington is suitable for this -- it's a small geographic area with a fairly high population density. Wellington has also been in the habit of cluttering the sky with overhead cables (the, uh, uneven terrain makes underground cabling difficult for everyone), and Saturn was able to get cable rolled cheaply on existing poles, whereas in other centres such services have had to have been put underground at greater cost.

Perhaps what's needed is assistance in building such local loop infrastructure.

No - we (NZ) just spent 10 years getting away from this. The economics of local loop roll out are very simple and the figures apply in both NZ and the US. - You have to get into a house for under $1200 - thats the build, the kit, connections, converters, etc (and often the set top box/router/whatever. This is not too hard. You have to take a total street or block approach not just one house. Then you look at your uptake figures and as I mentioned before you do well if you get 25 to 30%. I think Saturn got around 30% which internationally is good. So you work out your street build and divide the number of houses and then multiply by 4 and that is your return, so start applying your roi figures.

Not just in terms of assistance in getting capital to do it raised, but in providing an environment where infrastructure can be built easily. The latter doesn't necessarily mean overhead cabling, but say co-ordinating infrastructure development; why do water supply, sewerage, power, gas, traditional telecomms and broadband need completely separate holes? If you're going to dig up a street to replace a water main, how about laying ductlines for some or all of the above? How about if local authorities were required to maintain a register of interest in all street works, and were required to notify all interested parties so that co-ordinated, shared development or maintenance could be done?

DON - they have been doing this for years ! remember that funny Palette based mapping system I used to run at the MED ? technology now called GIS....... However getting trench co sharing is as hard as interconnect....... even WCC find it hard agreeing to let us share a trench However, its not that bad. Remember we are DEREGULATED. That means YOU can dig up the road. Just fill in the form, comply with the specs and get out your shovel. Hook up 12 people and get an operators licence. your own letter from the GG. But in fact its even easier. In large parts of the US the utility reticulation is along the BACK fences. Thats where the poles are, not in the street. So what you do is run your cables along the back fences using a standard easement form for each home owner. They easy to train. Now you start using NID boxes like we build. They're basically a fiber media converter and an ethernet switch in a weatherproof box with psu and surge protection. Simple traditional engineering. You put boxes every 100m like good ethernet says, and you run cat-5 to the houses. If you can't do it bloody cheap you go broke. If your're real clever you even skip the fiber and just run gig copper between the boxes. THat saves you the media converter costs which are the big hurdle. And since you are running copper its about 30 cents per meter. So why is a decent gigabit connection to a home so hard ? Actually I only think we need 100mbps, but the gig switches are getting so cheap might as well go for gig...... rich ps if you want to see real fun look at http://www.r2.co.nz/20020219 its only wireless tho nothing too sexy, but it does show the services like voice and video running. richard.naylor(a)citylink.co.nz - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog

Richard Naylor wrote:

...

Perhaps what's needed is assistance in building such local loop infrastructure.

No - we (NZ) just spent 10 years getting away from this.

We spent 10 (well, 15) years getting away from a state operated and regulated monopoly, where it was *illegal* to compete with the Post Office. I'm not for a moment suggesting going back to that situation. Rather I'm suggesting that the things that make telecomms hard be eased as much as possible. Digging up the road is expensive for everybody. Last time I looked (which was a while ago), you were talking $100 / metre, at least if you had to dig through the road rather than a nice soft verge, which Wellington hill suburbs are rather short on.

DON - they have been doing this for years ! remember that funny Palette based mapping system I used to run at the MED ? technology now called GIS.......

Yes, I recall very well you showing this off. But it takes a bit more than just knowing where everything is. If I come to the WCC and say I want to lay a ductline in a given street, does anyone else get told? Do I get an opportunity to utilise existing ductlines and avoid digging up a busy thoroughfare at all? Do I get told that I should lay multiple ductlines, and someone else may be along sometime later to buy them off me?

But in fact its even easier. In large parts of the US the utility reticulation is along the BACK fences. Thats where the poles are, not in the street. So what you do is run your cables along the back fences using a standard easement form for each home owner. They easy to train.

I always get into trouble generalising about US infrastructure, cos there's a lot of it and they all do stuff differently -- someone always finds a counter-example. But it's very common in N. America to have back alleys; the (*ugly*) 110/220V reticulation is down the backs of the properties rather than on the street as in NZ. It looks better than cables up the front of the street (except were it goes over cross-streets), but it does take up space -- it's not "down the back fence" per se, but on separate land.

Now you start using NID boxes like we build. They're basically a fiber media converter and an ethernet switch in a weatherproof box with psu and surge protection. Simple traditional engineering. You put boxes every 100m like good ethernet says, and you run cat-5 to the houses. If you can't do it bloody cheap you go broke. If your're real clever you even skip the fiber and just run gig copper between the boxes. THat saves you the media converter costs which are the big hurdle. And since you are running copper its about 30 cents per meter.

OK. Now run video and telephones over it. This is where it gets thornier, but would mack such an approach a hell of a lot more attractive to the punters. By "video" I mean something I can (a) plug into the TV, and (b) that I might actually watch. Saturn ended up running separate POTS and broadband cables. I get TV and Internet over the cable, but a telephone is still stuck with Dark Ages technology. -- don - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog

I go to Auckland for two days and find the list has exploded with things that I want to talk about. Oh well. At 9:05 PM -0500 26/2/02, Joe Abley wrote:

On Tuesday, February 26, 2002, at 07:58 , Don Stokes wrote:

[snip]

...

OK. Now run video and telephones over it.

Why cripple the IP infrastructure with requirements for legacy services? Leave the voice to the cellular operators and the existing copper pair people, and the video to the satellite and terrestrial broadcast people.

Convergence can come later, once a fast, cheap IP infrastructure is reliably available.

We (TelstraClear) are running a converged infrastructure now. Because it is cheaper. At least, where we have the new build (and that includes AK CBD). It cost us some more to build the IP infrastructure to accommodate the needs of telephony (and I'm NOT talking VoIP here by the way). But the overall cost went down. Personally I was surprised a bit that we managed to do it so fast, but we have. We've still got lots of 'legacy' stuff (and I do find it ironic that SDH gets relegated to legacy when it was cutting edge not so very long ago. Internet time I guess) but it's cheaper to do all new build based on an IP infrastructure. I recall a presentation from some telco types back at VUW when they were sneering at the reliability of data equipment vs telephone exchanges. Neil James made the comment that if he had a telco budget he could make the data network that reliable in a couple of years. Well, since rather less money was available, it's taken more than a couple of years, but Neil's vision is pretty much fulfilled. -- Michael Newbery Technical Specialist TelstraClear Limited Tel: +64-4-939 5102 Mobile: +64-29-939 5102 Fax: +64-4-922 8401 - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog

At 4:41 PM -0500 27/2/02, Joe Abley wrote:

On Wednesday, February 27, 2002, at 04:11 , Michael Newbery wrote:

...

I go to Auckland for two days and find the list has exploded with things that I want to talk about. Oh well.

At 9:05 PM -0500 26/2/02, Joe Abley wrote:

...

On Tuesday, February 26, 2002, at 07:58 , Don Stokes wrote:

[snip]

...

...

OK. Now run video and telephones over it.

Why cripple the IP infrastructure with requirements for legacy services? Leave the voice to the cellular operators and the existing copper pair people, and the video to the satellite and terrestrial broadcast people.

Convergence can come later, once a fast, cheap IP infrastructure is reliably available.

We (TelstraClear) are running a converged infrastructure now. Because it is cheaper.

You're talking about a converged core, right? Not delivering voice, video and internet to customers over ethernet.

No, I AM talking about delivering voice, video and Internet to customers over ethernet. Right now. If you call Christchurch City Council or people in the WestPacTrust tower in AK (not all of it, just some are our customers), or a numver of other customers, that POTS call runs over enet and has done for some months. In fact, the transport between exchanges up and down the country currently runs on a traditional telco SDH core, but the access is on IP. And it is NOT VoIP. Its a POTS line into a standard telco cabinet that happens to backend over a protected ethernet ring. No ATM. No SDH. It re-emerges back onto a traditional telco switch. We've also got the converged core, but the cool thing was integrating the access plot as well. Now, this is only where we have the new build at the moment, but all the legacy stuff integrates---at least for voice. The challenge is how to deliver high speed IP over existing infrastructure at reasonable price. -- Michael Newbery Technical Specialist TelstraClear Limited Tel: +64-4-939 5102 Mobile: +64-29-939 5102 Fax: +64-4-922 8401 - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog

On Mon, 25 Feb 2002, Chris Wedgwood wrote:

Part of the reason New Zealand is such a bountiful and rich place is that has a low population density and it's comparatively remote compared to the rest of the world; as such various commodities such as bandwidth is more expensive.

Err... I need to disabuse you of the notion that NZ is a rich place. It isn't. NZ has slipped way back in national wealth charts (from 3rd place to 20th, IIRC). It's official, the government is finally recognising that it cannot let the slide continue. No doubt this is one reason why Internet access costs a lot in NZ.

Deal with it.

Pay for it, you mean? I am, I am, don't worry about that. -- Juha Take off every sig! - To unsubscribe from nznog, send email to majordomo(a)list.waikato.ac.nz where the body of your message reads: unsubscribe nznog