Abbott is going to not have his cake while eating it.

How much are you prepared to pay for a fibre connection to your home?

Our internet service provider iinet sent us its price list for accessing the NBN. It looked ok. Not cheap, but I don't expect cheap for an excellent service. We plan to transfer as soon as it is offered, which hopefully will be before 14 September.

All great, but that's a subsidised price, paid for by the taxes of people who neither want nor need such a service.

If it was unsubsidised you would be paying perhaps 3-5 times as much. would you regard that as a good deal?

That's a copout, Gavin. You will be offered a service that has a maximum data rate of 100Mbps. The same service could be delivered by a FTTN network at significantly lower cost to the taxpayer.

If you were to send your kids to private school, you would pay. You could choose a Catholic school at reasonable cost, or you could send your kids to Shore at great cost. Your choice, your cost to evaluate as to value. If you wanted private health insurance, you would pay. You could choose a basic plan or a comprehensive one. your choice, your cost to evaluate as to value. In both cases, the subsidy is justified because the cost of the subsidy is less than the cost of providing a public service. Roads are a common asset that all benefit from, even if they don't have a car. Without roads, you don't eat.

Why should I subsidise your internet? What's in it for me, when there is a cheaper alternative?

"Everyone benefits from national telecommunications just as everyone benefits from roads."

How?

I've already explained the argument for subsidising private schooling. The public service costs more to provide than the subsidy provided. The only argument is whether the subsidy is too high. Personally I think it is, so we're in agreement there. Now, what about the huge subsidy I'm paying for your FTTP? What's the benefit to me?

The network effect: the benefit that everyone gets from a network being accessible to a large number of users

http://en.wikipedia.org/wiki/Network_effect

A network is already available to a large number of users. What additional benefit accrues by having FTTP rather than a progressive replacement of POTS with FTTN?

Why is it better, Gavin? What benefit accrues?

You keep dodging that question.

Still dodging the question. I'm afraid you're looking increasingly out of your depth, Gavin. The issue of replacing deteriorated copper has been discussed above.

What benefit accrues compared to the alternative? Please avoid repeating the same irrelevancies, they're simply tiresome.

We agree it's wonderful, you still haven't addressed why it's needed.

My question is siimple: how much better? What benefit accrues?

What are you going to do with the NBN that you could not do with a FTTN network at much lower cost?

Not even a single thing you can point to that can be done using the NBN that couldn't be done using a FTTN alternative? Not one?

I thought not.

The simple fact is that replacing perfectly serviceable existing networks with fibre is just not justified. On any grounds.

Personally I am not concerned for home users' and although they determine the 'outcome' by their vote, it is actually our Universities, Medicine, Security and perhaps Business sectors' (to name but a few) that I want to see have Gigabit+ FTTP.

Selfish, yes; I rarely if ever watch video, mostly look at boards such as this and the ABC, so no real need for bandwidth, however as voters' determine the success or failure of the awe inspiring NBN it is important for FTTH to put new media and medical services there; that ARE the FUTURE.

But the idea is that universities can offer on line learning to students and health providers can provide health services to their patients in their own home, for which bandwidth to peoples' homes will be needed.

That's just jingoism, Henry. Why is the maximum cost solution justified, is the question that needs to be answered, sloganeering doesn't do that, it just stops you thinking about it.

Neville, this is more of the same sloganeering.

The medical service don't exist and may never exist. What exists and is likely to exist in the next 1 years or more is well served by a FTTN model at much lower cost, with a progressive upgrade reducing the cost further still.

This is all a lot of red herrings, John. When you were working, did you always specify the "best" solution, or did you specify a solution that fit the performance criteria for the problem presented?

To use a road analogy, you're suggesting that we should build a Hume Freeway to every door, while hoping that in future there will be a demand for trucks to every home, all the time and that the trucks will be bigger and faster than the ones that exist now or in the foreseeable future.

Would you have lasted long as an engineer if you'd exhibited a lack of pragmatism to that extent when specifying for clients?

A fibre to the node system would enable reticulation from that point by lots of different nodes - whatever was most suitable. A business with large bandwidth demands might want fibre, but a home with much smaller ones might be served by a wireless connection of some kind.

One size does not fit all, as you must be well aware. Why are you advocating such an approach in this case?

You've written a lot of comments in the last day, I'll just stick to addressing this one.

Yes I've played with the connection profiles, it doesn't make any difference. I live at least 3km from the exchange. Nothing is going to improve that. On my parent-in-laws connection, I actually had to forcibly reduce the syncing speed (by forcing the profile to ADSL2 rather than ADSL2+) to stop it from dropping out. It still drops out, but not quite as often. They now sync at around 7.5mbps.

You mention that FTTN will require building new "Exchanges" so that the copper distance is less than 500m. Isn't this going to cost a lot of money? How is it beneficial in the long run, given that those "exchanges" will be made redundant by a future FTTH network? What about replacing degraded copper between the exchange and the house. You left that out of your comment. Are you proposing a progressive rollout to FTTH? If that is the case, surely you admit rolling it out as the copper degrades will cost more in labour than doing it all at once. Or do you suggest replacing the "last mile" of copper in the ground with new copper, or just letting it rot?

As for the "killer app", this is the mindset a lot of people fall into before dismissing the NBN. In the past, people had an internet connection to one computer. People had to take turns using this computer to use the internet. Then, people started sharing the connection to multiple family members. In the average household, it is unlikely that everyone is using the internet to its full potential. Maybe the parents just use it for some facebook or something. But we are entering an age where both parents and children will have grown up with the internet. This is going to put much more strain on a household connection. You might have 3 or 4 people wanting to stream a 1080p video, whether that be skyping with a relative or watching an educational video from a university.

Regardless of that concept, we already have a few killer apps that will utilise the bandwidth. Cloud computing/storage is one. Yes it is a buzzword, but that hasn't stopped companies like Apple and Microsoft investing in it. Have you tried office 2013? Heavily integrated with Sky Drive. Now office isn't going to utilise the high bandwidth, but other applications will. Applications are going to switch to a subscription model more and more, once internet connections improve. And they won't just wait for Australia to catch up to the rest of the world. Imagine photoshop running in a web browser. Or how about video editing software? Digital storage of music and video? Games streamed from the cloud? All of these things will chip away at your bandwidth.

Now lets think about the possibility of working from home. An RDP connection on a 4mbps connection is horrible. It is much much better at 100mbps. There are quite a few areas which could switch to a work from home system, perhaps even the company will pay for the NBN connection as part of the salary. Call centres are one. People prefer talking to an Australian rather than someone in another country, perhaps the NBN would allow businesses to cut costs and keep them in Australia. Call centre operators could work when they like, with their hourly rate dynamically adjusted for working in peak times.

And then we come to the uplink speed. This is the real problem with ADSL, and I see no evidence that FTTN will improve it. In fact Malcolm Turnbull seems oblivious to the concept of upload speed. Our connections are basically one way. Small businesses dealing with large files are screwed, unless they can somehow afford to pay for a fibre connection. Digital content is growing in size, why isn't the upload speed?

And finally we come to the concept of ubiquity. If we have FTTH covering most of australia, we have the potential to decentralise. You no longer have to live close to a CBD to get good internet, especially if you are a business. Developing an application that uses the internet? 20 million people in Australia are going to be able to use it if they want. You don't have to worry that a large portion of your potential customer base is on ADSL 1 speeds, or worse, stuck trying to use Telstra NextG in a conjested area.

The Coalition FTTN plan is not going to address many of these points, because they want a quick and dirty solution that makes them look good. But ultimately it is just money down the drain.

You suggest that new emergent technologies are going to orphan a cable to the home. This is rubbish, and the answer lies in physics. Fundamentally wireless and fibre are the same. They both use electromagnetic radiation to transmit a signal, one is just much closer to visible than the other. So why are they so different? Because one travels through a medium we control, and they other does not. SO much effort is put into combating the harsh environment wireless signals have to deal with, but it will never beat fibre because we can't control the weather and we can't control the other wireless signals in the area. Te medium is also segregated with fibre, you can run 100 fibres in a small bundle and they won't talk to each other. Each fibre can carry multiple independent signals. Wireless capacity will never reach the density fibre can.

You might argue that FTTN will be enough, or that new wireless technologies will be enough. Unfortunately, current trends in bandwidth consumption disagree with you, and until we see a bandwidth consumption trend that has plateaued, we cannot assume that it suddenly will, especially given that so many people in Australia are stuck with an extremely low limit (remember how you are in the top 4% of internet connections in AUS).

And that, is why I'm supporting FTTH and not FTTN.

Phillip, I said it would involve replacing exchange lines with fibre, not building new exchanges. In other words, copper that is in those big bundles you referred to. The connection to the house from the node might be done in all sorts of ways.

On the concept of replacing cable with wireless you're overlooking two things. Firstly, the fact that it is trivial to create a directional beam that can be very narrow indeed between two points. Every telephone or streetlight pole could house a transponder array, while houses had tiny transponders of their own. Second, using such directional beams allows all sorts of other frequencies than are feasible with broadbeam transmission. For example, even visible light could work, just like it does in an optical fibre, but with no need for the fibre over the short ranges needed. Modern LED and CCD tech makes this trivial. It might even be possible to do this with light fittings in the premises, it seems to me, freeing up the spectrum currently used by wifi.The flicker rate would be so high as to be indetectable for occupiers of the room.

The big cost for telcos is in the installation and maintenance of the cable to the home. It makes MDU access a PITA, it involves truck rolls for all sorts of issues relating to cable damage. There are property intrusion and duct installation issues, the list goes on.

It seems to me that rolling out fibre to the node and then distributing fro there could use the existing copper from there to the premises for as long as it lasts, then upgrade to whatever connection is most appropriate when it degrades beyond service specs, taking available of the technology that becomes available over time.

On the subject of ADSL and upload, I can choose the option of Annex M if I wish. which allows higher upstream bandwidth. I don't need it, but I'm sure some businesses could do with it.

Nobody is arguing that slow is less useful than fast. The only question is a cost-benefit one, for which we need to work out how much bandwidth is required and how much it can be delivered for.

The "ugly street cabinets", if that's the way you want to define them, will still exist regardless, for all sorts of services. That's a pretty precious sort of argument. Have you noticed the Optus cabinets, by the way?

Why will it be more costly to install and maintain? You keep making this assertion, but you haven't backed it up. There will be no need to rollout cable to individual homes, it will already exist. It will enable advances in technology to be incorporated. It will be cheaper at every point in the process. It will still enable those who insist on fibre to have it, if they pay for the installation work from the node to their home, or perhaps that could be included in the contract cost over time as was done with cable.

Anyway, the point's moot, the NBN and all the boondoggle it implies is what we have.

Ok, maybe exchanges wasn't the right word. New nodes will need to be built, and a lot of them. They will need power, unlike the passive optical network nodes planned in the NBN. The only was to get faster speeds out of copper is to shorten the length of copper. This is the whole point of FTTN and FTTH just takes it further.

Using a wireless protocol for that very short distance you suggest is silly, especially if it is directional. You open yourself up to even more problems with the uncontrolled medium. Eg, a large truck driving past, or a tree growing taller. Maybe your neighbour gets angry with you and blocks your internet with a piece of paper (in the case of visible light). You effectively want to remove the fibre from the equation, but make the termination hardware the same. How is this a sensible idea? The fibre is the one thing that won't need upgrading for a long time! It is the one thing that gives us control of the transmission medium, and allows us to achieve the fast speeds we do.

Furthermore, not everyone has access to Annex M, and even if they did it only increases upload speeds by a factor of 2. Any increase in upload speed is paid for with a decrease in download speed. Stop gap measures are not the answer, and we really need to start thinking in orders of magnitude if we want to get anywhere. Increasing upload speeds by a factor of 10 would be more appropriate, especially if you want to do video conferencing or you work with large digital files.

You suggest upgrading to FTTN and then to something else later (I put forward this will be FTTH) and yet you claim in the next sentence we need to look at the cost benefit. How is spending billions of dollars on something we can already see isn't going to be enough, a good idea? FTTN is the end point of copper. In contrast, A FTTH network can be upgraded by replacing terminating hardware, much like copper has over the last few decades.

I'm sorry, but I see no better option than FTTH at this time. We can put it off like you say, but from here it just looks like wasting billions of dollars. Current estimates put the coalition NBN plan at $10-$20billion. Malcolm Turnbull was quoted as saying $15billion once I think. So whatever happens, we'll be spending at least that much. The question then becomes, do we spend the extra $12.5 billion (based on government prediction of $27.5 billion investment of public funds to NBNco) on going straight to FTTH. Given that the ongoing maintenance costs of FTTN will be higher (due to the copper) and the ongoing power requirements for FTTN will be higher, my vote would be for direct to FTTH...

power's trivial, Phillip. The Optus network reticulates power at 90V to power the CAU on the side of the house via a voltage carried with the coax already. Power could be easily reticulated along a power pair attached to the coax. If it was low voltage it wouldn't require the installers to be electricians, either.

The idea of someone scaling my house or flat to put a piece of paper over my transponder seems a bit faretched and the problem with trees already exists. A wireless transponder would give more flexibility of placement and not have any more issues with trucks than cable does. Less, because it doesn't sag, especially in summer.

Government estimate of the cost of the NBN is >$37B. The difference is double. What do we get for spending double right now, that we couldn't get by spending less on a FTTN system that could be upgraded when needed? Especially given that technology doesn't stand still.

Thanks for that link, Henry, it's interesting. The thing is, there actually WAS an incumbent, with cable in the ground and it's being paid to take it out and for the use of its ducts and cable right of way.

Why not buy the cable as well and use it till it breaks. then replace it? Spread the cost over a longer time frame and take advantage of technological improvements into the bargain.

Craig, the point with power is that you need more nodes than exist now. The power usage is predicted to be higher. Also, you were the one suggesting that optical transmitters could be used. Now you are switching back to wireless transmitters. Are you still suggesting using a directional wireless transmitter, one for every house? Or are we back to a shared wireless signal, that isn't directional.

Yes the cost of the NBN is predicted to be $37B, but the government contribution is only $27.5B. The rest is made up by private investment. This is because the project is slated to provide a return on investment. As far as I know, the coalition FTTN plan is not expected to receive private funding, which might tell you something about a predicted ROI.

And as has been pointed out by commenters here (and many others including people in the industry) FTTN is not really on the upgrade path to FTTH. You waste billions of dollars on FTTN cabinents. The suggestion that we can spend $15B on FTTN and then be able to spend $15B less on going to FTTH is laughable. It isn't going to work that way, and you know it. SO again, why waste money now?

I'm suggesting that technology is not static. Wireless modes are among the most heavily researched areas of tech there is at present, whatever frequency is used. Optical spectrum is simply one potential part of that for specific applications, not something conceptually different.different. Why do you struggle with that? The short range and directionality makes optical spectrum a potential candidate for the local connection, with or without an encasing fibre.

I have no knowledge of what the LNP plan might be, I'm simply arguing for a general case. No doubt it's possible that their plan is worse than the NBN, I haven't looked at it in detail.

I don't see private investment as an issue either way, except insofar as it ups the cost for the sake of ensuring a profit for a private operator. or otherwise distorts the technical argument, which it seems to do for you.

FTTN is not on the upgrade path for a specific FTTP model. On the other hand, it provides lots of interconnect capacity with other forms of technology that a "pure" FTTP model doesn't, including low-power wireless of many different forms that might be useful for improved services of all kinds, not just for streaming porn in the nation's bedrooms or giving kids the best possible gaming connection.

The question I ask is why are we spending so much on an all or nothing approach, rather than taking the time for a phased approach that allows uptake of technology as it arises?

In future, will all new premises have to be connected via fibre, even if there are cheaper options that will serve the same purpose?

All we've done is lock ourselves into a single model for all time. As a physicist you should know better.

Craig, I don't know how to make it clearer than I already have. The medium of transfer is the important point I'm trying to make. Base load broadband needs to be robust and it needs to be fast. Wireless cannot compete with fibre in this regard, EVER (this is from a physical perspective).

I'm not suggesting no one ever use wireless, far from it. Wireless is a good complementary service, but it can't handle the base load any time soon. I've also outlined above reasons why a ubiquitous broadband network are desirable. You can go back and read them if you like.

Furthermore, to suggest we have "locked ourselves in" to a single model "for all time" is ridiculous. If a better technology emerges, of course we will use that. But no such technology exists now. Doing a half-assed job of upgrading our communications network now because "something" else might come along later is idiotic.

Fibre is the best solution we have now. It does not have a limited upgrade path (there are already plans to provide up to 1gbps connections in the future). Rolling out a ubiquitous FTTH NBN in Australia will provide a ROI.

Finally, you mention that wireless is a heavily researched area. Do you know why? Because it is that much harder to make breakthroughs in wireless technology because you can't control the medium of propagation. You have to spend so much more time correcting for that.

Anyway, Unless you get my point about the transfer medium, I think I've had enough debating.

Phillip, what is important is the bandwidth needed and the bandwidth that is able to be delivered. The medium doesn't matter a bean.

If you need water available at the tap, you don't really care whether it is delivered in a pipe or by the truckload, what matters is the cost, quality and convenience. There are many homes that rely on water being trucked in at their own expense: would you be prepared to pay for them to have pipelines installed - hang the expense, or do you recognise that it is simply not viable to do so on a cost-effective basis?

If you don't, then there really is no point discussing much at all with you - you're living in a dream world.

I'm also fascinated that a PhD candidate in physics would be so certain about future technological developments. Don't they cover the history of mechanics these days? Ever heard of a chap called Newton - had some interesting ideas about how the world works, some people thought he had it nailed...

On the subject of wireless, you're still not thinking clearly. All you say about spectrum is correct, for broadcast, not for narrowcast directional, low-power local links that don't contaminate spectrum by their very nature. The lack of capacity you cite is a consequence of trying to use narrow channels that have to be spectrally separated to avoid cross-channel interference and have a limited capacity based on the narrow channel width. Optical fibre is a special case of what I'm suggesting, but it requires a wave guide, mostly because that maximises the potential transmission range.

There is no reason a dedicated fixed, directional - using a phased array RF or some other means,such as discussed above- wireless connection can't provide 1Gbps. Power demand is a furphy, since the low power required could be provided by all sorts of means, given the transponders would be mounted on poles that already carry power. Possibly even inductively.

In the meantime, lots of existing infrastructure is available to be used and is going to be pulled up and destroyed.

I do think it's funny that you demand fibre because you think future apps will come along, but you think that suggesting future hardware technology will do so is "idiotic".

I'll try one last time...

The medium matters because the medium ultimately limits the maximum achievable bandwidth. As I tried to make clear, wireless, fibre, and any of the other "new technologies" you are trying to talk up are just transmitting or manipulating certain frequencies in the electromagnetic spectrum. The medium thus defines which frequencies are going to be useful and also defines the type of technical noise that will impact the systems performance (when I use the term technical noise I mean ANYTHING that will negatively affect the measured bandwidth in a real world test when compared with a theoretical prediction. That covers discrepancies between theoretical predictions and experimental results due to the environment, etc. I hope you understand this and don't just associate the term with analog noise).

So again, my point is that the medium is what is important.

Your water analogy is also flawed. If you could store the Internet in a tank, and only refill it when it was getting low, I might take it seriously. If you want to stick with your analogy, a truckload of water (internet) is not feasible in any situation. What we are debating is the medium of continuous water transfer. Do you pipe it direct to the house (fibre), or do you have a hose up on a pole which shoots it across the road into a funnel on the other end (wireless/pipeless)?

I see fibre as being the preferred solution because it gives you complete control of the medium, thus giving higher maximum bandwidths. You see wireless as better because you think it is cheaper? I would question whether you have thought that through. Infrastructure to set up an antenna for every house will be costly (I assume you are still going with point to point wireless, or have you changed your mind again?)

Anyway to address your final points...you suggested that:
A) I'm a PhD student who is certain of future developments
B) I think that suggesting future hardware technology is idiotic

So:
A) Nowhere have I suggested I have any idea about all future developments. The point I am trying to make is that said future developments are not ready for commercial deployment, nor will they be for a decade or two. Some may not even have been imagined yet. So do we wait for them to appear, or use the best we have now? I would prefer we use the best we have now, and when it comes time to upgrade again, use whatever the best is at the time. Since most communications methods are EM based, it seems a reasonably safe bet that the fibre will be resuable, and future improvements for the next few decades will revolve around improving existing technologies.

If you can name a technology that:
1) is already conceived, AND
2) doesn't use the EM spectrum OR is limited to a part of the EM spectrum that isn't within the range of transmission of the fibre we are installing, AND
3) is predicted to provide more bandwidth than existing fibre technologies
Then I'm all ears. I'd also appreciate an approximate (+/- 5 years) guess on when the technology will be available for commercial employment.

Also, I'll channel my quantum mechanics lecturer here....Newton was right. As he puts it, the new theory (quantum mechanics) reduces to the old (Newtonian mechanics) in the domain of validity of the latter. Lots of people still use Newtonian mechanics today. It is still perfectly valid in most situations. Quantum mechanics didn't make it obsolete, one of the reasons why Newtonian mechanics is still taught. Would you have told Newton "oh we can't rely on your theory that is experimentally sound within our finite set of knowledge, because at some point in the future someone might discover a more complete theory that incorporates yours"?

B) Yes I think suggesting "future" technology (that is technology not yet commercially available, especially technology not going to be commercially available with sufficient speed within the next decade or two) for a project that is happening now, is idiotic. We can wait until the cows come home for the next big thing around the corner. Or we can build the best we can now, and tackle the new technology when it is ready.

Concluding remarks:
Fibre seems the best bet for now, especially in terms of maximum achievable bandwidth. I believe we should run with that plan now, and worry about future technologies when and if they eventuate.

Phillip, you're simply not addressing the topic.

I couldn't care less if fibre is "best", there are lots of other solutions that are near enough to best for the local link and an easily meet or exceed the capacity of the fibre backbone to deliver to the local point of distribution. All the stuff about medium deosn't matter a damn, because transmission distances are short and the signal can be directional. A short-range, low-power directional signal can be quite large in bandwidth and bandwidth is what determines the capacity to carry data, all things being equal. A high-bandwidth signal in air can carry every bit as much data as a lower-bandwidth signal in a "more perfect" medium.

Moreover, when it hits the premises, the signal will be translated and re-transmitted using some other mode, whether wifi, twisted pair or whatever, all I'm suggesting is that the point of translation should be at the earliest rather than the latest reasonable point, thus dodging the high cost of cable installation and maintenance at the premises.

Try to move down from your "this is the theoretical best performance" hobbyhorse and have a think about what's actually being said to you. I agree that a signal in a low-loss wave-guide is an ideal situation, but it is not necessarily the most appropriate technology for the purpose.

How much bandwidth do you think you can reasonably consume? Are there cheaper ways than fibre to deliver that?

I say there are and that few consumers require fibre.

I would have liked to see an NBN arrangement that made best use of all the existing cabling. A sort of Super HFC. As was mentioned earlier, an incumbent has a good reason to do so and so it's a no-brainer for Australia, which has already spent $20 billion buying the cables back so they can be scrapped.

The only factor limiting bandwidth in existing HFC networks is contention at the node level. A fibre backbone that allowed local nodes to devolve to block size would enable much higher available bandwidth through lowering contention rate.It could be distributed to the home via the existing means first, with upgrades or cross-grades to another existing mode (ADSL/POTS to RG6) as the closest fit deteriorates and eventually installation of whatever technology is available, subsidised by the removal and sale of the copper in the cable. LTE or some other future broadcast wifi standard.

DOCSIS networks are already achieving 200+K Mbps down/15+ up and DOCSIS 3.1 is specified at 10Gbps down/1Gbps up.

We already have the cable in the ground and on the poles.

Phillip, you're getting quite hot under the collar. Take a chill pill, dude.

The simple fact is that fibre is NOT necessary for the local link. You still haven't told me what you think your own bandwidth demands might actually be, which means you have no idea of what the actual size of the pipe to the premises needs to be. Without that basic piece of data (I'd say 100Mbps would provide more than adequate bandwidth to the vast majority of homes and businesses for the next 5 years at least, until much larger screens become the norm, with the additional bandwidth they demand).

You're simply not making a lot of sense. If the cost of fibre to the node is half that of FTTP then FTTN should be the automatic first choice, in the absence of a compelling demand-based argument to the contrary.

In a nutshell, you're very happy to throw other people's money around but you haven't provided a strong argument, despite several days and many hundreds of words of opportunity.

I'll ask again:

1. What is the projected bandwidth demand?
2 What drivers are behind that demand and are they real or "potential"?
3. What modes can supply that projected demand?
4 At what cost?
5 What other technologies are either existent or in development to deliver the final link to the home?
6 What other forms of interconnect could occur if there were convenient points to do so at local level, rather than merely at the 121 new exchanges the NBN will use?

Your argument is simply not thought through, it consists of a repetition of "It's the best ever, I want it, so there" in various forms, with no justification. You haven't even tried to address the cost/benefit argument, nor the demand argument, nor the flexibility argument, nor the translation within the home argument, nor the alternative technogies argument. you seem to fail to grasp the difference between broadband narrowcast signals and narrowband, broadcast signals altogether. What branch of physics is your PhD in?

I say to you that the arguments you have ignored are ignored because you know they are strong and you don't like the implications they have when you insist that you require fibre (just because you do).

A good scientist doesn't think that way, Phillip.

Oh, and yes, the user could happily align their own receiver. It's no harder than lining up a TV antenna and could be much easier, given that the device will be attached to a computer that can provide feedback in regard to field strength and other parameters.

However, I would suggest that it is pretty straightforward for a tech to climb a ladder and do it at the same time as the transponder on the pole is installed. A pair of techs, one at the house end and one at the pole end would do a whole street in a morning.

My record for installing cable broadband to a home was 35 minutes for a new installation, including premises cabling and a drop from the street, as well as provisioning and operational checks on the modem. With no cable to install and a simple piece of gear to screw to a wall I reckon 10 minutes - tops. Add another 20 if cable has to be run from the transponder to some other point, which need not be the case.

Phillip, here's a real world application of a directional wireless (over LTE in this case) link.

We don't have TV or Radio out here (no reception) so the internet and phone are our only means of communication with the outside world. All Telstra, of course, we don't muck around with any other carriers. It's Telstra's network and they maintain it after the fires, floods, storms etc which play havoc with our lines.

"There's no chance of any fixed line internet out here. We tried the Government subsidised satellite and that was only marginally faster than dial-up so after a lot of trial and error and frustration with Telstra call centres, we found a fantastic solution – wireless internet via 4G modem and fixed booster (yagi) aerial on the roof. Briliant."

http://www.onlineopinion.com.au/view.asp?article=14767

Craig, I am far from "hot under the collar" despite your comments such as:
"A good scientist doesn't think that way, Philip" (Attempting to imply I'm a bad scientist?)
"you're very happy to throw other people's money around" (Attempting to imply that I have less ownership over government funding than someone else, despite the fact that the current NBN is projected to provide an ROI of 7%, a point you continue to conveniently ignore)
"Try to move down from your "this is the theoretical best performance" hobbyhorse and have a think about what's actually being said to you"
"I'm also fascinated that a PhD candidate in physics would be so certain about future technological developments."
Etc.

But anyway, to address your questions:
I'll direct 1,2 and 3 to this article: http://www.abc.net.au/technology/articles/2012/06/14/3524848.htm
Particularly the first set of dot points.

You could perhaps say that the article is in support of point 5) though I maintain it is better to leave it up to the end user to decide what wireless technology to use. People spend the majority of their time at home or work, so it makes sense that people in charge of those places provide the required solution than hoping a national communications company can provide for a wide variety of situations across the nation.

As for 4) We've already established the cost. You apparently have a problem with $37 billion ($27billion of public money) but not with $15 billion. I will add yet again that this is $37 billion that is expected to generate a 7% ROI and we have no evidence that $15bn spent on a FTTN network will provide a ROI.

I actually don't have an opinion on 6). Is that going to be a problem?

The fact that you have boiled my argument down to "it's the best over, I want it, so there" is pretty childish in my view. I've tried to explain why I believe a fibre based NBN is the way we should be heading. I tried to give significant justification in my second post, which was largely ignored (though you did suggest using Annex M...so helpful of you).

I'm going to end my contribution (actually really this time. I'll even let you have the last word) by addressing this statement:
"You still haven't told me what you think your own bandwidth demands might actually be, which means you have no idea of what the actual size of the pipe to the premises needs to be. Without that basic piece of data (I'd say 100Mbps would provide more than adequate bandwidth to the vast majority of homes and businesses for the next 5 years at least, until much larger screens become the norm, with the additional bandwidth they demand)."

My bandwidth requirements are largely irrelevant to the debate (my use combination is unique to me, but that doesn't mean no one else will want to use a high bandwidth connection either), but since you care so much...
I'd start streaming TV/Movies instead of buying them on bluray.
I'd transfer large datasets from my research. SInce you also seem to care a lot about what I do, I work on an experimental apparatus which makes Bose-Einstein condensates. This experiment can run continuously, under computer control, without human intervention for long periods of time. It produces a file that ranges from 20MB to 100MB every 30 seconds. It is likely that I will want to analyse large collections of these files (at least 500) at a time, a significant time after they have been created. It would still take an hour to download all of these files on a 100mb/s connection (pretty sure I did the maths right there).
I would set up a VPN to create my own, relatively secure cloud storage system, with offsite-backups to other family homes.
And that's just what I know I'd do now. I'm not fool enough to predict all the applications that will arise in the future.

But do you know what the kicker is? You said in the above quote that 100mbps (what a fibre NBN will provide) will be more than adequate for the next 5 years. Did you miss my last post where I said we need some long term planning? The NBN isn't slated to be completed for another 7-8 years. FTTN will be slightly quicker, but the coalition is unlikely to get it completed sooner than 5 years as they'll have to deal with undoing the current NBN agreements first (especially with Telstra). So even with what we are planning, we'll (according to you) have an adequate network by the time the NBN is completed. It'll be even less adequate if we do FTTN.

So no Craig, it is you that doesn't make any sense. You refuse to consider long term implications. You want us to build an FTTN network (I think...sometimes you swap to wireless or "future technologies") knowing full well we'll need to do something better in 5 years. You haven't addressed the potential cost of doing two upgrades instead of one, even though most experts claim that FTTN doesn't upgrade to FTTH (or any directional wireless technology for that matter).

And I suppose I can't go without addressing this wonderful line:
"You haven't even tried to address the cost/benefit argument, nor the demand argument, nor the flexibility argument, nor the translation within the home argument, nor the alternative technogies argument"
I've attempted to address the cost/benefit. FTTN vs FTTH is a factor of 2 at most. One has better upgrade potential than the other. Other options haven't been costed, and speculation on costs by you or me will only be uninformed. FTTH is predicted to provide a 7% ROI (maybe if I mention this enough times you will finally read it?) but we have no information on an ROI form the coalition FTTN plan.
Companies such as cisco (see the above link to the ABC story) are predicting we will need increasing bandwidth in the future. They believe this should be provided by Fibre with wifi end points. A FTTH network achieves this, while leaving the end user the freedom to choose their preferred wireless option.
The only thing more flexible than a FTTH network would be an omnidirectional wireless network (aka 4G) because it provides a mobile connection. But 4G can't provide the baseload bandwidth, so will remain a complementary service to fixed line connections.
I still don't understand the translation in the home argument. Whatever the mechanism of getting the signal to the home, as long as it is directional, the user will have to translate to whatever they want. The only advantage to translating between mediums before the home is if it saves money, and you'll forgive me if I don't take your anecdotal word for it.
And I've addressed the alternative technologies argument several times, indicating why I don't believe they compare to fibre. This links back to the cost (again at best a factor of 2).

Adios!

Well done, Michael, your dog-whistle attracted a whole pack.