Australian Government will offer 1Gbps this year to all NBN-connected homes | Hacker News Reader

Australian Government will offer 1Gbps this year to all NBN-connected homes

smh.com.au

Of course, we're having an election sometime this year, and the Liberal's — who are probably going to win — are planning on making the NBN slower and cheaper[1] (100Mbps by utilising the existing copper network, rather than fully replacing it with fiber).

1. http://www.zdnet.com/coalitions-nbn-to-cost-almost-30bn-7000...

Up to 100 Mbps. By they'll try to get at least 25 Mbps. Don't ask about upload speeds, those apparently don't matter.

Not to mention latency, which is the most significant factor for me. I'm on semi-NBN and now have a 4ms ping to work. Only have to leave the house once a week (Warning: side-effects may vary).

What's semi-NBN?

Opticomm or something similar?

Nah, it's South Brisbane. They knocked down the old exchange for the hospital extension, Telstra installed fibre, supposedly eventually it will be rolled into the official NBN network, but until then, we get rates way worse than ADSL and I'm forced to pay $30/mth for a "landline" which I never use... still, it's worth it for the latency.

I think Clarke and Dawe did a fairly decent video on the LNP alternative. http://www.youtube.com/watch?v=TptIs0k-spg

The ABC program "Inside Business" gave a neat summary of the differences between the Labor and Liberal plans: http://imgur.com/OtmOBE7

It's worth noting that the Liberals plan has 'Fibre To The Premises' going to 22% of households, while Labour's plan goes to 93%.

So we're spending an extra $17 billion to give 71% of households a boost from 100Mb to 1Gb per second? That's about $2,500 per household. That doesn't seem like good value to me. Shouldn't we delay the massive cost of installing fibre to existing homes until most households need more than 100Mb per second?

Can the current copper network handle 100Mbps? I think anything broken down into a per household cost is going to be large - health, transport, and Internet infrastructure.

There is a very significant difference between 93% coverage and 22% coverage with this issue. The value goes well beyond the percentage difference because of network effects[1].

Being able to count on near-ubiquitous FTTP and speeds of 100Mb - 1Gb p/s for a whole country completely changes the tech startup landscape.

It's hard to estimate how valuable that is to Australia's social, economic, and technological future.

The Coalition are expected to romp it in, so it's pretty disappointing that Malcolm Turnbull was convinced or corralled into playing politics on this one instead of just neutralising the issue by committing to FTTP. They happily make baseless blowout claims around the current NBN project, so I don't see why they couldn't have instead made baseless "we'll do the same, but cheaper" claims.

Their system design around a broken and aging copper network has few tech industry commentators impressed. I can't see how maintaining the copper network and delaying the inevitable FTTP rollout makes any sense. Being a first-mover in Western society brings a significant number of benefits for Australia's tech environment.

[1] http://en.wikipedia.org/wiki/Network_effect

> Being able to count on near-ubiquitous FTTP and speeds of 100Mb - 1Gb p/s for a whole country completely changes the tech startup landscape.

Let's not kid ourselves. The Australian market is tiny.

> so it's pretty disappointing that Malcolm Turnbull was convinced or corralled into playing politics on this one instead of just neutralising the issue by committing to FTTP.

There's also the possibility that Malcolm Turnbull, a former barrister, investment banker and venture investor in and director of Australia's first major ISP (Ozemail) has his own thoughts about balancing costs and benefits.

> They happily make baseless blowout claims around the current NBN project, so I don't see why they couldn't have instead made baseless "we'll do the same, but cheaper" claims.

It's pretty easy to demonstrate that paying someone to dig up every street in the country is more expensive than deciding not to dig up every street in the country.

> Being a first-mover in Western society brings a significant number of benefits for Australia's tech environment.

Japan and South Korea have shown that it's really not that big a deal. Really nice, and enables some new businesses. But you have to trade it off against alternative uses of the funds. The same money could build about 20 new hospitals, a new Sydney airport and a clutch of cross-city tunnels. It's possible these might have been more valuable in the long run.

Or maybe not. We don't know because no cost-benefit analysis was ever done.

> But you have to trade it off against alternative uses of the funds.

It is an investment on borrowed money using our AAA rating, we expect to get a return on the invested funds higher than the interest rate paid. The alternative isn't hospitals or schools, it is not doing it.

It's possible to borrow for other purposes. Plus there's the little thing where borrowing by governments is simply deferred taxes or deferred budget cuts.

We don't know what the rate of return is because no proper CBA has been published. We're merely told that there was one and that the companies paid by NBN -- with instructions we haven't seen -- said it was just swell.

> Let's not kid ourselves. The Australian market is tiny.

I don't know if you're being willfully ignorant in your responses, but I can add to my comment: "...for a whole country completely changes the tech startup landscape for that country".

> There's also the possibility that Malcolm Turnbull, a former barrister, investment banker and venture investor

In this context appeal-to-authority isn't particularly interesting. The pundits are all credentialed in this debate, and the tech sector overwhelmingly prefers the current NBN plan to the Coalition's.

> It's pretty easy to demonstrate that paying someone to dig up every street in the country is more expensive than deciding not to dig up every street in the country.

The fact you're quoting a response of mine and responding with a completely different issue is classic trollish behaviour. However as it's an absurd statement that looks superficially like common sense, I'll respond. Would Australia have saved money by not building an electricity and telephone network?

> Japan and South Korea have shown that it's really not that big a deal. Really nice, and enables some new businesses. But you have to trade it off against alternative uses of the funds. The same money could build about 20 new hospitals, a new Sydney airport and a clutch of cross-city tunnels. It's possible these might have been more valuable in the long run.

Reducing a national FTTH network to 'really nice, and enables some new business' and proposing a back-of-the-napkin alternate spend of new hospitals, new airport, new cross-city tunnels (!) which 'possibly' might be more valuable is not an analysis that's worth refuting.

I think you're more interested in pushing a political agenda than having a tech + economics discussion, so best to park this!

I am having a tech & economics discussion. My sarcasm is a learned response from dealing with an unending parade of interlocutors who assume that the NBN is an unalloyed good. The main problem is conflating the technical characteristics of FTTP/FTTN with the total net benefit. They are not the same and should not be treated as the same.

Nothing in life is that simple. Nothing. There are always alternatives and it always pays to do a proper cost-benefit analysis.

Whenever I try to argue this case I am accused of all sorts of things. Trolling, being a stooge, a shill for tin-can-and-string manufacturers.

What I believe is that it's not enough just to want to build the right thing. You have to build it the right way and for the right reasons. 1/3 is not enough in matters of the public treasury and I consider that both major parties suck in this respect.

How much extra would it cost to upgrade the FTTN to FTTH afterwards? Surely it's more efficient to do it once and do it right with fibre?

And I have my doubts about the copper reaching 100Mbps.

My understanding is that copper will be replaced with FTTP when it wears out. That spreads the cost over time while bringing forward the 25Mbps minimum service level forward by a few years.

There's no doubt that FTTP is the ultimate long term solution. But that doesn't mean it has to happen first.

> My understanding is that copper will be replaced with FTTP when it wears out. That spreads the cost over time while bringing forward the 25Mbps minimum service level forward by a few years.

Much of it is wearing out. Users who already have issues with noise, interference and/or slow speeds on ADSL will be similarly disadvantaged by VDSL under the Coalition's FTTN plan. You also still need to worry about flooded pits, leaking conduit and so on. Having worked at a number of ISPs and watched all the faults roll in after even moderate storms, I know that these are real problems.

Further, how does the Government decide when your copper is "worn out", and what incentive do they have to spend more money (over their original plan) to upgrade you? How many people in your area need to suffer the same problem before they need to both a) completely replace your FTTN VDSL equipment with GPON gear, and b) pull fibre, replace conduit and wire small groups of homes on an ad-hoc basis?

I agree, lots of of the existing infrastructure is worn down. My understanding is that Turnbull proposes that some of it will be replaced immediately and the rest would continue to be replaced on a rolling basis.

The main difference between the coalition plan and the business-as-usual plan is that old copper will be replaced with fibre, not with new copper. It's easy to forget that waaaay back when, this is what Telstra and the government were negotiating to do anyhow.

In Turnbull's position I wouldn't have promised anything except to review once in government. For one thing, a lot of work will already be contracted and it's not plausible to renege on the contracts. So there'll be an uneven distribution of fibre/copper which will lead to some distortion in the housing market. Not huge, but it'd be nice if it wasn't there.

> I agree, lots of of the existing infrastructure is worn down. My understanding is that Turnbull proposes that some of it will be replaced immediately and the rest would continue to be replaced on a rolling basis.

The problem is that these costs are most certainly not encapsulated in Turnbull's $20B costing.

The problem is that neither the government nor the opposition have ever produced their original costings nor any audit of those costs. In the government's case we're promised that audits were done. But we can't see what was audited and we can't see what the instructions are.

My understanding is that the major cost is to pay people to do the physical work of drilling holes and digging trenches, so the cost of upgrading to FTTH in future (adjusted for inflation) would be about the same as creating a FTTH network now.

I wish the US government would get its act together on this. The Constitution includes a grant of power to Congress to create a national system of post roads (Article 1, Section 8.7) and my understanding is that the founders included this because they recognized the importance of a public communications infrastructure rather than a privately held one [1]. Considering the criticality of our communications infrastructure in an era of cyber-attacks, I find it disappointing that Congress seems to have abandoned this important role to the private sector.

1. http://en.wikipedia.org/wiki/Thurn_und_Taxis

Do you really want the government to be in charge of data transport with its track record of blatantly ignoring the right to privacy? Aren't you the least bit concerned that we'll end up with something like in the UK or China, where government control over internet infrastructure has lead to extensive monitoring and censorship?

I'd rather take (possibly) slower and/or more expensive internet that the government has to put a lot of elbow grease in to censor or monitor than a (possibly!) cheaper and/or faster internet that the government has extensive control over.

If you've worked in telecom, you might know how extensive the government's data monitoring infrastructure already is. I sure as hell don't want to make it any bigger.

I'm concerned about the creation of the infrastructure, rather than the operation of it. And no, I don't lose a lot of sleep over this, any more than I lose sleep over having my mail opened.

>any more than I lose sleep over having my mail opened.

You know, that does bother some people.

Are you bothered with having your incoming and outgoing mail checked for un-approved messages and then having those thrown away?

I'm not bothered by it because it isn't and hasn't been a problem in this country. We have laws against such abuses that work quite well and have, if anything, strengthened over the lifetime of this country.

With the NBN, the government is just a wholesaler. They just provide a physical medium and end users will still get a retail service from a private company. Besides the government has the same responsibility to adhere to privacy laws as any private company, in fact government have their own privacy requirements with additional channels to resolve complaints. Ethnically I don't see why a government that you vote for would have any less reason to respect privacy than a private company (who you pay money to). Finally, the government could just pass legislation to monitor/censor within a private network, so it doesn't really matter whether it is public or private from this perspective. It's more a matter of voting for the right representatives.

> With the NBN, the government is just a wholesaler.

The Minister responsible for the NBN -- Senator Conroy -- is also the guy who wants ISP-level internet filtering based on a secret blacklist.

The whole NBN has been a joke from the beginning. The roll-out delays for what was originally a 100mbps fibre network that is now supposedly going to offer 1gbps speeds conveniently announced right around the time Quigley is about to be grilled in parliament over the delays and so close to an election it looks like Labor will be losing in a landslide.

While the prospect of 1gbps speeds sounds great as a developer, the very fact those speeds will only be affordable to businesses at a cost of what looks like could be close to the AUD $200 per month mark initially and no doubt rise with minuscule bandwidth provided for the cost. The reason current Internet offerings in Australia are so expensive is because of the small number of cables that connect us to the rest of the world and monopoly of our limited selection of ISP's, are they planning on laying more deep-sea cables to handle the increased load?

Will the high-speeds only apply to Australian content like Optus's fraudulent high-speed 100mbps addon pack promises for cable subscribers which only could promise close to those speeds for Australian content, not content from overseas.

On the outside this looks like a Godsend for Australia and while it undoubtedly is in some aspects as we saw with the rise of ADSL in the early 2000's, the project has been poorly managed from the start and this sounds like a stunt more-so than a reality based on what we've seen so far.

Having said that, I would be more than happy with 100mbps speeds compared to the 20mbps connection I currently have which costs me $75 per month through Optus Cable. And overall the premise and vision for the NBN is still better than the archaic coalition "high-speed" network plans they have of their own.

The odds that a project of this magnitude and complexity would be done on time and budget were basically zero. But it's very hard to tell what could have been better because of the Commercial-in-Confidence veil.

Quigley is not an idiot; he's a smart, well-educated bloke with decades of telco experience. He had no control over the schedule -- Treasury (remember those guys? they're the ones who said that the European carbon price would continue rising to hit $29 by 2015) pretty much drew up an estimate for a press conference and that's what got chucked over the wall.

In Quigley's position I'm pretty sure anyone would have struggled to make it work.

That's kind of a tautology ... all the homes in the US connected to 1Gbps fibre "this year" will also be connected to 1Gbps fibre "this year". The salient missing fact is that the NBN is going to reach only a tiny tiny miniscule minority of homes "this year" (my house is not even scheduled on the plan that stretches out past 2017).

True, just that "XXX,XXX homes currently connected to the Australian Government's NBN initiative, will have 1Gbps speed enabled by the end of the year" is a bit long for a title. I've made a small title edit :)

Here's the NBNCo targets:

June 2013 - 661,000 "passed or covered" and 92,000 premises with active service

June 2014 - 1,681,000 "passed or covered" and 551,000 premises with active service

http://www.nbnco.com.au/assets/documents/nbn-co-corporate-pl...

They'll find some way to cripple it for sure, like sticking a 200GB/month cap on it. Australia can't have nice things :/

> They'll find some way to cripple it for sure, like sticking a 200GB/month cap on it. Australia can't have nice things :/

If you look at many of the plans already offered by ISP's, you'll find that this isn't the case.[1]

Once more homes are connected, these limits are only likely to increase (as they have each year prior). Perhaps we shouldn't have limits at all, and just pay for a pipe of a certain bandwidth; but from an ISP's perspective they implement these caps to protect themselves from abuse by a small % of users.

Personally I'd rather know my cap is there, as opposed to the odd behavior of many US ISPs in targeting "high usage" users with what seems like little transparency.

[1]: http://www.iinet.net.au/nbn/nbn-plan-residential.html

> If you look at many of the plans already offered by ISP's, you'll find that this isn't the case.

The largest plan offered by iiNet has a quota of 1000GB. If you're on the 100Mbps plan, this can be consumed in less than 1 day. How is this not the case?

Because you won't consume it in less than a day.

For the present time, 1TB is enough for almost all home users.

As history has shown us, these limits will increase as demand does.

And iiNet overage at $0.10/GB is pretty favourable as well… what does AWS charge again?

> Because you won't consume it in less than a day.

I would if I could…

> And iiNet overage at $0.10/GB is pretty favourable as well…

Well, I suppose it's reasonable. I don't think it's fair to compare it to business-grade services like AWS though; it would be more appropriate to compare the NBN to Google Fiber or one of the Japanese/Korean ISPs.

Or Labor will muck up this election and the Libs will muck up the NBN.

You have universal healthcare

How is that crippling it? From the point of view of subsidized public infrastructure, it's much better to keep it fast for everyone than to let it get overloaded with people torrenting movies.

As an Aussie who left to live in the USA, let me assure you Australia has many, many nice things.

Healthcare, education, safety are very high on the list of things you should be thankful for, and are high on the list of reasons I'll be coming back (at some point)

Pardon my ignorance, but aren't the pipes that actually go to Australia relatively small? So, wouldn't this just increase burden on those interconnects and maybe not help unless connecting to somebody else in Australia?

I don't know much about the pipes actually, but I do believe a new one is being constructed. That said, I am curious to know the answer to this question too.

I guess torrenting within the borders will be a speedy affair however.

Allow me some time to find the statistics again on this, but from what I've gathered following the NBN debate for a few years now: No. The pipes are more than double our current requirements and growing faster than requirements do due to some funny incentive scheme.

Total capacity: ~2TB/s [1], but the PPC-1 line has potential to add another 2.5TB/s. The others are being upgraded too [2]. "Currently, Australia has a theoretical 5637734.4Mbit/s of transpacific bandwidth, however lit capacity is much less."[3]

"The total international capacity in use for the Australian market in 2009 is estimated to be around 300 gigabits per second."[4]

References: [1] http://en.wikipedia.org/wiki/Telecommunications_in_Australia... [2] http://www.itnews.com.au/News/157753,ppc-1-delivers-more-spe... [3] http://en.wikipedia.org/wiki/Internet_in_Australia#Internati... [4] http://whirlpool.net.au/wiki/nbn#nbn_faq (See section 5.7 - "Does Australia have the international capacity to provide every customer with 100 megabit internet?")

The days of this being an issue are long gone, especially since there is now good competition in the submarine cable space. Last mile is starting to be the bottleneck.

Could you elaborate? I don't see prices for bandwidth dropping at Australian hosting facilities.

Quick eyeball of Wikipedia suggests there are five major cables out of Australia: Telstra Endeavour, Southern Cross, AJC and Pipe Pacific.

Telstra Endeavour is currently 80Gbps with capacity to 1.28Tbps (http://en.wikipedia.org/wiki/Telstra_Endeavour)

AJC is currently 240Gbps with capacity to 1000Gbps (http://en.wikipedia.org/wiki/Australia%E2%80%93Japan_Cable), and probably tons from Japan to the US.

PPC-1 can deliver up to 3Tbps (http://en.wikipedia.org/wiki/Pipe_Pacific_Cable)

SXC has 800Gbps * 2 lit, with capacity up to 6Tbps * 2 (http://en.wikipedia.org/wiki/Southern_Cross_Cable)

There's also SEA-ME-WE and probably other smaller ones, but they are the main links out of Australia.

I hate to be the guy who says that this looks like an election-year stunt to try and "save the furniture", but ... well, it does.

Edit 2: A mate of mine pointed out that they made the same announcement before the 2010 election:

http://www.theaustralian.com.au/national-affairs/nbn-downloa...

Part of the problem with the NBN is the highly selective veil of secrecy. It's a public project and it's being backed by the Commonwealth. The lenders who are loaning the funds are not doing it because they think NBNCo would be allowed to go broke; everyone knows that these are really Australian Government bonds in a dress.

But the government still uses "Commercial-in-Confidence" to stymie all oversight. This would be more reasonable in dealing with a company at arm's length, but NBNCo is 100% government owned. They would be well within their rights to waive the requirement for confidence.

Consequently, nobody really knows how the rollout is going or what it costs. NBNCo claim they can't give Parliament information about particular streets because they don't track to that level of granularity, yet mysteriously they are able to claim in aggregate to be on track.

Leaving to the side of whether this is the right thing to build, there's still the important question of building it the right way. And IMO that's not been happening, and it really bugs me.

edit: I know people don't like this point of view, but if you disagree, the done thing on HN is to reply, not downvote.

The upgrade to 1Gbps was always in the roadmap: http://www.nbnco.com.au/assets/documents/initial-roadmap.pdf

Frankly, I think they should've launched with 1Gbps to begin with -- customers don't have to pay for the higher speed tiers if they don't want to.

Ah, I hadn't realised that. So it's a stock standard re-announcement.

Reply on news.ycombinator.com