[AusNOG] Gosford City Council and NBN RSP.

[Jake Anderson]

On 29/05/14 18:49, Mark Newton wrote:
> On 29 May 2014, at 13:00, jake anderson <yahoo at vapourforge.com> wrote:
>
>> The issue with achieving these things in practise is with the MTM you are in the same boat you are in now,
> Well, yes -- that's part of the point.
>
> I think there's a bit of woolly-headedness here that should be explored.
>
> Firstly:
>
> Let's say Australia rolled out a network that had multiple VCs delivered on multiple UNI ports on a complicated NTU.  And let's say Australian companies started designing services which depended on that infrastructure (e.g., medical monitoring which connected to a dedicated medical monitoring port)
>
> Australia, as the only country going down the road of delivering residential services under that model, would be a captive market: The devices/services developed under that model would only be usable here, because North America, Europe, Asia simply are not going down that path.
>
> So the only customers will be Australian customers;  All the R&D costs associated with developing the service offering would need to be recouped from Australian customers;  "Scale" would be impossible.  The service would thus be a commercial failure, because the Australian market isn't big enough to support that kind of innovation.
>
> An Australian company could develop an offering which wasn't a commercial failure by developing it to work on the kinds of networks that are available throughout the rest of the world, so they could shard their R&D costs across multiple markets.  What kind of networks are available throughout the rest of the world?  Internet access delivered on 1000baseT.
>
> If the Australian service developer needs to make it work under that model anyway, their offering will work on Australian internet connections too, and the unique-to-Australian-NBN aspect of the service then becomes an additional cost:  Its only purpose would be to increase the Australian retail cost of the service, and we'd be better off without it.
>
> Have a think about it:  Is "innovation" which makes Australian services uniquely expensive actually worthwhile, when the same services can be delivered globally without the unique expense?
>
> Think about how much more expensive Australian electrical appliances would be if we decided that we were going to deliver power at 290 volts at 85 Hz.  Same kind of problem, right?
Actually I would disagree with much of what you have said there.
There are 2 separate issues, one is bandwidth, and the other is the 
multi port NTU.
I don't see the multi-port NTU as allowing for "new services" that is 
the domain of bandwidth, anything can be delivered over IP as you suggest.
The multi-port NTU is just a mechanism for delivering those services and 
the key thing to me is it dramatically simplifies support and 
demarcation, and really if the upstream controls the NTD, firing a vlan 
at a port is not a great complexity burden.

>
>
>
> Secondly:
>
> Ever since the NBN project was first announced, part of the political rhetoric has been about all the whizzy new services it'd enable.
I'd suggest that is a part of the rhetoric of one side of politics which 
they then couple with basically suggesting the only use for it is 
watching tv.
>
> In parallel to that, we're invited to assume that Australia's infrastructure is inferior to everyone else's.
46th on the akami speed list last time I checked behind Mongolia, that's 
not exactly awesome.
>
> Which raises the question:  If someone else's infrastructure is so much better than ours, can we get a glimpse of the kinds of services we'd be able to utilize in Australia by looking at those markets, and seeing what they have that we currently can't use?
I'd suggest that you don't see these things because the network has 
become invisible.
You don't see people boasting about how they have created this brand new 
widget that overcomes the limitation of wan bandwidth in some way 
because that is just not an issue any more.

>
> Turns out we can't:  Other markets with better infrastructure than us simply have faster internet access than us.  The services are delivered "over the top," and are largely independent of the network infrastructure used to deliver them.
>
> Wind the clock back a bit to before ADSL in Australia:  Youtube was just getting started.  Unusable on dial-up, but broadband was an enabler.  But youtube works the same whether your broadband is 1.5Mbps ADSL or 1 Gbps fibre:  the over-the-top service doesn't care, it just wants a way of getting IP packets from source to sink.
>
> Are there youtube-like services out there which the South Koreans (100Mbps FTTH) and British (50 Mbps VDSL) and Singaporeans (100 Mbps cable) can enjoy right now, which we can't?  I can't think of any.  Can you?
>
> I don't think bandwidth is an innovation enabler in the way that the politicians think it is, and we need to be wary of buying into their rhetoric.  What I'd expect to see if bandwidth was a brake on innovation is that we'd be deploying services over the internet which are usable but crap in Australia, and usable and brilliant in markets with better infrastructure.  Think of what we'd have had if ADSL was capped at 512kbps in Australia:  Youtube world manage to work, but Australians would have a lot of buffering and would never be able to use HD.
>
> Are there any services knobbled like that for Australians?  I'm not sure that there are.
you kind of negated your own point regarding the future mystery services 
there.
Low bandwidth = no youtube.
High bandwidth = youtube.
If we didn't get adsl, then youtube would never have happened, it would 
have died you need the bandwidth before you can get the services.
Some small % of the world happens to have gbit now, I'd say its too 
small a sample to be drawing conclusions from.
You can't say because X doesn't exist yet, X must therefore not exist.

(BTW youtube over a 100mbit link is a rather different experience to it 
over a 1mbit link)

>
>> you are at the mercy of the RSP to provide a level of service for over the top services and you are also dependant on them having enough bandwidth.
> Yes.  So ISPs (not RSPs - that terminology will have died with the ALP NBN) which can't deliver that will die commercially.  And good riddance to them:  My message to them is, "We're in a world of plentiful bandwidth, if you're in the business of selling internet capacity which you can't reliably supply, you should be forced to discover that the world doesn't owe you a living for terrible service."
Delivering crap bandwidth for cheap and dodging customer complaints 
seems to be a winning business model for a number of service providers 
out there today.
I don't see that changing in the future regardless of the mechanism for 
delivering those bits.
>
>> With FTTP and multi-port NTU's you can guarantee those things for a price.
> Nope.  Even ALP's NBN was never going to "guarantee" them, for any price consumers would want to pay.
Pricing is a separate issue, and one you will find I agree with you much 
more on.
>
> Remember:  Each RSP was going to have to buy NNI capacity at $20 per megabit.  Frickin' extortion.  I don't know why every single participant on this mailing list wasn't screaming from the rooftops in outrage over that.  If you had even the slightest understanding of your own industry you'd have been marching in the streets demanding change. Locking AVC charges into the NBNCo business model was one of the single worst things the ALP did with the project.
I agree with you wholeheartedly on that issue. NBNco should have tried 
to stay out of price plans and the like, and the whole idea of CVC was 
daft. All it serves to do is limit the use of the network.
I'd much rather they set the pricing to maximise the usage of the 
network and earn their revenue and let the RSP's work out what they want 
to do to deliver their bits.
>
> At the very least you'd have raised it with your MP and withheld any NBN boosterism until the issue was resolved.  How many of you did that?
I am pro NBN because it had more good than bad, if you wait until you 
have perfection before doing anything you are going to never have anything.
Getting glass into the ground, and *reliable* high speed services 
everywhere, seems to be a cause worth championing, I did in fact raise 
the issues of CVC and the pricing model with my MP, unfortunately he was 
given the ass fairly recently.
>
>> In the multiple port model, parliment could just post the user a laptop/whatever and a long cable and tell the user to plug it into port 2 on their NTD, They could provision ultra HD webconference software on the laptop and know that everything is going to just work.
> Why can't they do that now?
>
> Oh, that's right:  Quite a lot of ISPs are rubbish.
>
> If you're a network operator providing something that -isn't- the kind of service which would enable high-def videoconferencing from a Gosford council office to Parliament, you need to ask yourself some deep and meaningful questions about what the hell you're doing in this industry.
>
> Take some responsibility, drive change.  The NBN isn't going to do it for you.  It was never going to do it for you, under any model.
So I should avoid buying any internet at home because I can only get 
5mbit and that's kinda crap?
NBN solves the last mile being a heap of crap, all the other stuff is 
dressing on that really, some of its nice to be sure (multi port ntd) 
but really what it does is mean that for 90% of the country you can 
order a service and get it.
Using internode or dodgybros internet at the moment makes no difference 
if you get 1mbit with 30 dropouts an hour when it rains.

You can argue the implementation of it till the cows come home, but the 
fact is for the past 20 years there has been 1.3 last mile network 
operators in the country.
Telstra 1
optus .3

Everybody else banded together would be lucky to hit 1%.
>
> This is supposed to be a list of network operators, who understand this kind of stuff.  The rah-rah about the NBN lives on Whirlpool, right?
>
>    - mark
I would tend to agree, however my "understanding" and yours may well be 
different.