• Great articles on other sites
  • RSS Great articles on other sites

  • News, Telecommunications - Written by on Tuesday, November 27, 2012 11:19 - 118 Comments

    Copper network not “rooted”, says Telstra

    news The nation’s largest telco Telstra has rejected unsubstantiated claims that its copper network was suffering a fault rate of up to 30 percent, pointing out that in October this year, only about 1.3 percent of its telephone lines nationally suffered any kind of fault, and that it invests hundreds of millions of dollars each year keeping it that way.

    Late last week the ABC’s Technology & Games site, which has taken a broadly critical line with the Coalition’s rival National Broadband Network policy, published an article citing claims by an anonymous industry insider that Telstra’s copper network was “rooted”. The site wrote:

    “Our insider explained how three years ago, when Telstra tendered the process to switch to their own Fibre to the Node scheme it was found that the average network pillar had a 30 per cent failure rate in terms of line faults and lines being listed as connected to the wrong number.” The ABC’s article linked this issue to the Coalition’s NBN policy, which focuses on re-using portions of Telstra’s copper network in a so-called fibre to the node strategy, as opposed to Labor’s fibre to the home-based project.

    However, in a statement responding to the ABC’s article, a Telstra spokesperson pointed out that the company’s latest network reliability data showed that one 1.3 percent of its customers had experienced a fault during the most recent reporting period — October this year. In some areas, such as the Sydney central business district, that figure was even lower — as low as 0.58 percent. In addition, Telstra’s statistics also showed that services were available, on average, 99.92 percent of the time.

    “Telstra is constantly installing, replacing, repairing and upgrading copper lines throughout its network,” the spokesperson said. “It’s a never ending job and we’ve been at it more or less non-stop for 100 years. Every year we invest hundreds of millions of dollars operating and maintaining our network to ensure our customers are able to enjoy the quality of service they expect. The work installing, replacing, repairing and upgrading copper lines hasn’t stopped as we prepare to progressively migrate our customers onto NBN Co’s Fibre to the Premises Network.”

    The Australian Communications and Media Authority, which is responsible for monitoring the performance of telecommunications companies in Australia, praised Telstra’s reliability in its most recent telecommunications performance report for the 2010/11 year (PDF). The regulator wrote: “The monthly percentage of Telstra’s CSG-eligible services that did not experience a fault remains consistently high. In 2010–11, 98.49 per cent of services were fault free on average per month, marginally lower than the previous year.”

    “In 2010–11, at a national level, services were available on average for 99.90 per cent of the time each month. Most areas across Australia experienced a very high level of service availability on Telstra’s fixed-line network.”

    In addition, NBN Co itself has also publicly commented on the issue. In an article published by iTNews in September last year, Peter Ferris, general manager of design and planning at NBN Co, said Telstra’s copper network suffered more than ten faults per 100 services — a figure far short of the 30 percent rate claimed by the ABC.

    The ABC’s new round of criticism of the Coalition’s FTTN-based NBN policy comes just a week after the latest stoush between the broadcaster and Shadow Communications Minister Malcolm Turnbull on the issue. Again quoting industry sources who didn’t want to go on the record with their claims, the ABC’s Technology & Games editor Nick Ross concluded in an article entitled “Huge doubts cast over Coalition’s ‘cheaper’ NBN alternative” that deploying a FTTN network in Australia would be ” an expensive, time-consuming hindrance”, compared with the current FTTH deployment.

    Speaking to ABC Radio last week, Turnbull rejected the criticism. “Well that is, with respect to whoever wrote that, complete and utter nonsense,” he said, pointing out that BT in the UK and AT&T in the US were using the FTTN deployment model, and that the FTTN streetside cabinets didn’t need to be the size of “fridges” as they were “getting smaller all the time. The speed of rollout was also demonstrably faster than with FTTH rollouts, Turnbull said, pointing out that in the UK, BT had deployed its FTTN infrastructure to seven million premises in one year alone. Turnbull added that the cost was about a quarter of the cost of the fibre to the home model.

    opinion/analysis
    So who’s right here? Well, it’s a bit hard to say. To start with, I’m not entirely sure that we’re comparing apples with apples here, given that “fault rate” is a bit of a subjective term. I’m sure Telstra has hundreds of different types of fault rates and hundreds more of other types of measures for tracking the reliability of its copper network; and let’s not forget that it delivers both voice and data over that infrastructure.

    On the one hand, many who work in Australia’s technology sector would suspect that there is some basis in fact to what the ABC’s un-named source is claiming. You only need to move house a few times or check out Delimiter’s gallery of “worst of the worst” examples of cabling in Telstra’s copper network to know that there are definitely problems in various places with the network; it is ageing, it is being used for purposes which its original designers never intended it to be used for, and Telstra does need to constantly fix issues. I’ve heard quite a few nightmarish stories from Telstra contractors about the state of the network over the years.

    However, to pull out one un-detailed, unsubstantiated fault rate on one area of Telstra’s network and extrapolate that into a claim that Telstra’s copper network is so “rooted” that it would be “completely impractical” to deploy fibre to the node in Australia is just, in my opinion, horse twaddle. The situation is a lot more complex than that in reality, and I think most within Australia’s telecommunications industry would accept that it is indeed possible to roll out a fibre to the node network in Australia and that such a network would indeed deliver a dramatic improvement in basic broadband service delivery to most Australians. Perhaps such a deployment wouldn’t be as easy, quick and cheap as Malcolm Turnbull constantly claims it would be, and perhaps it would be a lot messier than Labor’s elegant FTTH model but it could be done.

    If the ABC, or anyone else, wants to make an argument that it would be impractical to upgrade Telstra’s copper network to fibre to the node, they’re going to need to present a whole lot more evidence to prove that claim. Telstra itself still appears to believe that its network could be upgraded to support FTTN; one unsubstantiated claim about fault rates isn’t going to be enough to change many people’s mind that Telstra is wrong about its own network.

    Image credit: Telstra

    submit to reddit

    118 Comments

    You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.

    1. Posted 27/11/2012 at 11:21 am | Permalink |

      Rooted? No.
      Degraded? Maybe.
      Old? Yes.
      Expensive to maintain? Clearly.
      Good enough for VDSL? Bwhahahaha!

      • Posted 27/11/2012 at 11:26 am | Permalink |

        “Good enough for VDSL? Bwhahahaha!”

        Evidence?

        • Posted 27/11/2012 at 11:45 am | Permalink |

          I’ve attempted to have many locations provisioned for ADSL that have returned “Not Viable” results – and if they can’t do ADSL, they sure aren’t going to do VDSL.

          • CMOTDibbler
            Posted 27/11/2012 at 12:31 pm | Permalink |

            Was that on FTTN?

            • Brendan
              Posted 27/11/2012 at 1:40 pm | Permalink |

              We already have FTTN (or a very near equivalent) in many places. Typically this is thanks to a CMUX, or similar box, located on on near a curb, serving a number of homes using ADSL, with connectivity back to an exchange (sometimes fibre, often something else).

              However that’s not really the point. The point is the last mile of FTTN, is currently dependant on copper.

              Copper that has a service life, that if left to rot, will indeed do so. THe same copper we use today. The same copper that provides wildly different outcomes, depending on where you are.

          • Dave
            Posted 27/11/2012 at 3:15 pm | Permalink |

            Yeah, Telstra only count faults as issues affecting voice services. Our line has a fault which means we get 4Mbps less bandwidth than our neighbours but Telstra won’t look at it because our voice service is working.

            It wouldn’t surprise me if both figures are right; that there is a 30% fault rate for data and a 1.3% fault rate on voice services.

            • NBNAlex
              Posted 27/11/2012 at 3:55 pm | Permalink |

              Interesting thought!

            • Gwyntaglaw
              Posted 28/11/2012 at 8:14 am | Permalink |

              Our phone line was so badly corroded that ADSL speeds sank to below 512kbps, but that didn’t concern Telstra at all – their brief was to porvide voice service, no more.

              It was only when the crackle on the line grew louder than the voices speaking that they conceded a line failure, replaced the wiring to the house, and there followed an immediate boost of ADSL sync to >14Mbps.

              The Telstra techie sighed and confessed that they were going out all over, but hardly anyone was left to replace them.

            • TechinBris
              Posted 28/11/2012 at 12:20 pm | Permalink |

              Hey! A nice point to note in that Telstra will only counts fault issues affecting voice services. When in today’s communications requiring the communications pathways to provide more than just the 3.1Khz voice pathway in the old PSTN product, and that “Voice” Services can technically be carried (VoIP) in the data part of the “extra” bandwidth that Telstra charges extra for ADSL products, it raises the horrible nightmare for a Corporation who sells a product to explain why it will not support a product it actually charges people for.
              I wonder what the TPA says about no support for a product that is not working in a way it was marketed and sold to you for ongoing charges. I’d fathom to say it is a breach of the Act when it does not work.
              Any excuse at all in legal mumbo-jumbo to justify supporting such a model shows why the NBN is required.
              Be careful in your objection to the NBN, as everything currently that is our Communications market is a “disappear up your own ass” trick. The trouble is, how does that last part actually disappear? The Liebrals (well that is what they currently are) have not explained that part yet.

              • Posted 28/11/2012 at 2:57 pm | Permalink |

                As a former telstra support tech who did thousands of ADSL & Line Tests in my time.
                Yep
                Totally true.
                For telstra, its not a ‘fault’ till your having trouble talking to people.
                If they are claiming <2.5% its got to be voice only.

                I wish someone could subpoena a statistically valid size, randomized sample of full cable data & line tests and dump the data online for analysis. That would show things to be very different.

          • Nich
            Posted 27/11/2012 at 5:20 pm | Permalink |

            They might not do ADSL for a handful of reasons that may disappear if a sub-eschange type cabinet was installed closer to you for VDSL. I’d have thought the more pressing concern would be just what speed they’d guarantee you getting, depending on distance from exchange/quality of copper along the path :p

        • Posted 28/11/2012 at 9:26 am | Permalink |

          As someone who programmed ADSL I can assure you that the network barely supports ADSL, let alone is it capable of VDSL.

          From gel filled joints corroding copper, to 50 pair sections being faulty due to no OpEx being allocated to proactive/reactive maintenance, to low gauge copper from the 50’s, through to excessive line lengths even in metro areas, VDSL is going to be more of an uphill battle than it’s worth.

          The fault reporting over a period is hardly an indication of how faulty the network is, most of the faulty network is dead copper that Telstra refuse to replace as they don’t see it as economically viable.

          I’ve had lineys have to cut into pipe & CV joint different pairs just to get services running, I’ve had streets with 100 pair cables yet only 3 or 4 pairs are functioning. They put their requests into the (at the time) WMC, only to have them rejected.

          There is no way Australia can support a VDSL network without spending billions more than the (fibre) NBN on both the initial set up & the maintenance of the network.

          • Posted 28/11/2012 at 9:49 am | Permalink |

            “As someone who programmed ADSL I can assure you that the network barely supports ADSL”

            Seems to work pretty well for millions of Australians at the moment.

            “There is no way Australia can support a VDSL network without spending billions more than the (fibre) NBN on both the initial set up & the maintenance of the network.”

            Evidence … ?

            • Posted 28/11/2012 at 10:18 am | Permalink |

              “Seems to work pretty well for millions of Australians at the moment.”

              Uhh, no, it doesn’t. Spend 5 minutes with gamers from all over Australia & you’ll see this is patently false.

              “Evidence … ?”

              BIS Shrapnel’s recent report, the cost of copper per metre vs polymer fibre, & the requirements for VDSL2. All three of these things mean continued use & upgrades of the copper network is financially retarded. Not only that, with the speeds the opposition are claiming they’ll get & line lengths (sub 1000m), you’ll need at least 20x the cabinets a PON would require. Yet more costs there, not just small costs, massive costs. 32x 20km of fibre optic cable is far less expensive than 20 cabinets.

              I can see you want to give balance to the opposition’s plan, but there is such a thing as false balance.

              • Posted 28/11/2012 at 10:56 am | Permalink |

                “Uhh, no, it doesn’t. Spend 5 minutes with gamers from all over Australia & you’ll see this is patently false.”

                Um … StarCraft II player here, dude. Platinum League …

                • Posted 28/11/2012 at 11:04 am | Permalink |

                  A sample set of 1 is far from conclusive.

                  Meanwhile, I play WoT with another 200+ Aussies who can barely play the game at times, let alone play & use voice comms. The main factor is their copper, not their provider.

                  Over the years of gaming & supporting networks (ADSL, ISDN, Fibre), the majority of fellow gamers & users tend to have problems with their copper. Whether it’s gauge, faults, or line length.

                  • clownface
                    Posted 24/06/2013 at 6:44 pm | Permalink |

                    As for my two cents all I can say is you can hardly play internet chess at times – but then I live in Perth and we all know what an absolute scam fest it is over here!

      • Geoff
        Posted 27/11/2012 at 11:42 am | Permalink |

        “In God we trust; all others must bring data.”

      • Posted 27/11/2012 at 2:02 pm | Permalink |

        Of course if you ask Telstra if its network is ‘rooted’ theyre going to say no.

        If you can actually GET adsl, then you have issues like dropouts, pair gain, RIM access, RIM and Exchange congestion… list continues …

        Then theres the evergrowing issue of Telstra NOT extending the copper network by adding more exchanges. So its just more RIM access or micro-exchanges, both of which are totally useless in provisioning ADSL if its too far from the actual exchange. You can only extend the RIM path so far, even by fibre before it becomes a liability.

        So what does Telstra do? Doesnt extend the network or upgrade it beyond what they have to; then sells NextG Wireless services instead which have a massive margin.

    2. looktall
      Posted 27/11/2012 at 11:35 am | Permalink |

      Telstra itself still appears to believe that its network could be upgraded to support FTTN

      well of course they do.
      they’re covering their arse in case the coalition win the next election and push for FTTN.

      they’re not about to come out and say “well actually our network can’t support FTTN, so you might as well stick with FTTP”. that would be shooting themselves in the foot.

      much better for them to say “hell yeah we can do that. bring that mofo on”.

      • Posted 27/11/2012 at 2:03 pm | Permalink |

        ^ this guy’s onto it.

        +1.

      • Gwyntaglaw
        Posted 28/11/2012 at 8:27 am | Permalink |

        If Telstra does find itself in the position of negotiating to sell its copper tail after the election, I have no doubt some clever analyst could find evidence of any line remediation priced into the deal.

        Either they will have to do a spot of unexpected spring cleaning of their network prior to sale, or you will see the assets priced down on an assessment of their condition.

    3. GongGav
      Posted 27/11/2012 at 11:37 am | Permalink |

      I dont think its oranges v oranges Renai. Maybe oranges v mandarins, but definitely different comparisons.

      The little daleks have a squillion connections in them, and the insider claim is that 30% of those connections are faulty. Thats different to what Telstra is saying.

      They’re saying that 98.5% of customers have no problems, and they’er probably right. But given the squillions of connections in the pillar (side note: how many ARE there?), the redundancy levels means both statements can be right.

      There can easily be 30% failure rate, but enough spare connections that the end user wont notice.

      Is that important or not? Is the final result that 98.7% of people wont notice more important than 30% of the network might be unusable?

      • NBNAccuracy
        Posted 27/11/2012 at 11:45 am | Permalink |

        Exactly what i was about to post. Fault rates and how much of the copper is degraded are two different things. If you read the comments for the ABC article you will also see a lot of people who say they have worked for Telstra saying that if anything that what was said was pretty conservative and is probably worse than mentioned.

      • Dan
        Posted 27/11/2012 at 12:14 pm | Permalink |

        I think Gav is spot on here. The pillars might have 99% active and sound connections to customer premises, but there may be as much as 30% of the available copper pairs “rooted” as per the claims by Nick Ross’s source.

        This would mean that both parties are making valid claims.

        Telstra are also talking about voice services, not pure data services, which means that their numbers are highly inflated as a pair of exposed wires with measurable resistance across them will still carry voice, but not a viable data signal.

      • CMOTDibbler
        Posted 27/11/2012 at 12:33 pm | Permalink |

        “There can easily be 30% failure rate, but enough spare connections that the end user wont notice.”

        How would these faults be detected if not by the end user reporting them?

        • SimonB
          Posted 27/11/2012 at 12:52 pm | Permalink |

          I haven’t gone back and read the article, but from memory it said that the contractors who make the initial connections for a new premises test the ports at the pillar, and if the port they are supposed to use is faulty (and allegedly that fault rate is around 30%) then they simply choose another one. The end result is the user doesn’t know there was an issue with the port and from their end things appear seamless. From Telstra’s end, though, they have an incorrectly mapped connection and a potentially unreported hub fault.

          • CMOTDibbler
            Posted 27/11/2012 at 1:21 pm | Permalink |

            Thanks. So these are faulty pairs detected when connecting a new service rather than faulty lines on existing services. That might explain the discrepancy in the numbers.

            • nonny-moose
              Posted 27/11/2012 at 2:32 pm | Permalink |

              One party is talking about active lines/services, and the other is talking about all possible/available copper pairs in the network. I don’t see it as discrepancy as much as difference in focus. We know there are more pairs than services, so I thought the telstra statement a tad disingenuous.

              If said pairs are not or are poorly tagged than that complicates the work of laying in a fttn network; in terms of installation labour as that’s another (time consuming) task on their list. address and service matching is a big part of the nbn, it comes built into it. The same would apply for any fttn build.

              That’s what really bugs me about fttn. Its not that it can be done, it sure can. But it is adding costs, time(=money) and complexity, for a position that will be revisited anyway, while the in progress policy is far simpler and removes many anachronisms and bad features of the old network in one swell foop. Nearly all the issues with fttn revolve around retention of the copper at last mile, and that’s traditionally where a large portion of telstras yearly upkeep goes, whether thats good or bad I don’t know but logically retaining that portion retains its maintenance costs as well. So I think the question about ALL local Telstra lines/pairs and not just the active ones has some merit.

      • djos
        Posted 27/11/2012 at 2:06 pm | Permalink |

        In my experience the ABC article is correct; the DataCenter I used to manage ran out of usable PSTN pairs after only a year and half of operation* – there were easily 30% of the total pairs that were faulty and unusable. We requested Telstra fix them and the answer we got was essentially “no, tell customers to have their services provisioned via fibre” (and fibre was something we had plenty of from all the major Telco’s & ISP’s).

        *amazingly some customers still use dialup modems and ISDN links

      • Anthony Wasiukiewicz
        Posted 27/11/2012 at 2:40 pm | Permalink |

        +1
        Was my first thought.

    4. Tinman_au
      Posted 27/11/2012 at 11:41 am | Permalink |

      I suspect the fault rate that Telstra is using, is for it’s USO which is why the ACCC commented on it (ACCC has oversight of the USO doesn’t it?). A voice POTS is pretty resilient, you only need a few Kb for voice.

      The fault rate mentioned by the “unnamed source” would be for data connections (as that was explicitly mentioned), which need a much higher quality of line, and is less fault tollerant.

      “Fault” and “Line Quality” can be two very subjective things if they are not defined, and unfortunately the sources haven’t actually defined them here.

      • Tinman_au
        Posted 27/11/2012 at 12:03 pm | Permalink |

        Opps, meant ACMA, not ACCC….was still thinking about that copyright article :o)

    5. SMEMatt
      Posted 27/11/2012 at 11:42 am | Permalink |

      Deer Telstra.

      Lets say I have a 100 network points in my building and have had 1 fault in the last year. Now only 70 of those points are used and I know another 30 US and can’t be used. What % of my copper infrastructure has had a fault in the last year. You say 1.4% because 1 of the 70 active services has a fault I say 31% because 31 of the physical connection are either faulty or have had a fault. One number tells you the state of your active services the other tells you the condition of the underlying infrastructure.

      • Dan
        Posted 27/11/2012 at 12:16 pm | Permalink |

        This +1

        PS, for other readers, US = unservicable

    6. jasmcd
      Posted 27/11/2012 at 11:42 am | Permalink |

      As pointed out, there is a big difference in the quality of line needed to service telephony vs quality required for VDSL.

      Pick up the phone, hear a bit of crackling? It all good, you can still here the person on the other end, sure, but now try any online video services or downloads. For my own line, if I don’t use a download manager, I am lucky to get any file over 50MB, such can be the rate of the drop outs I experience.

      VDSL for me? Not without substantial upgrades to the part of the copper network that services my household.

      • Michael
        Posted 27/11/2012 at 12:48 pm | Permalink |

        This is called a “fault” and you should, you know, get your ISP to demand it be fixed.

        If your connection is literally dropping out because of the line quality that is still considered a fault.

        • Tinman_au
          Posted 27/11/2012 at 1:42 pm | Permalink |

          I think jasmcd’s story is actually pretty common, I know I have a lot of friends that have similar issues with ADSL. This is a big contributor to why Australia is ranked 60th when it comes to promised line speed verses actual line speed, and 34th when it comes to line quality.

          (Source: http://www.netindex.com/promise/allcountries/ and http://www.netindex.com/quality/allcountries/)

        • Bpat
          Posted 27/11/2012 at 4:58 pm | Permalink |

          My parents have a click on their phone line. It comes from a neighbors electric fence. Telstra cannot fix the fault without pulling up a couple of kilometers of copper or finding out where the electric fence leaks. They have chosen to do nothing.

    7. Brett
      Posted 27/11/2012 at 11:48 am | Permalink |

      VDSL maybe, VDSL2 (the upgrade path)? maybe not. VDSL2(2+?) needs 2 active working pairs. 1 pair in my 2 pair line is gone, Copper is about 30 years old. Alot of new houses only have single pair lines to save money for some large big corporation.

    8. Brad Cann
      Posted 27/11/2012 at 11:51 am | Permalink |

      yeah sorry, comparing the ongoing fault rate, with what the original article said which was dead or faulty pairs, is not apples for apples.

    9. Abel Adamski
      Posted 27/11/2012 at 11:51 am | Permalink |

      I also agree with the other posters.
      Telstra is being disingenuous, there is fault rate and state of network , a tenuous connection at the best of times.
      It was worth reading the comments to Nicks article.
      It is not just one anonymous source, but also many others who have confirmed Nicks assertion, they are in many cases actively or have been actively involved either in the CZN network or attempting to use it

    10. Tinman_au
      Posted 27/11/2012 at 11:56 am | Permalink |

      And I don’t really see a figure of 1 in 10 services having a fault as that good really, extended out nationally, that would mean 2.2 million services a year will have a fault.

      The FiOS fault of 1 in 111 is what I’d call “good”, that would be a much more acceptable 198,198 a year.

      This assumes a population of 22 mill and everyone being connected (unrealistic, but you get the idea).

    11. Stephen
      Posted 27/11/2012 at 11:58 am | Permalink |

      To paraphrase Mandy Rice-Davies;

      ‘Well they would say that…’

    12. Tinman_au
      Posted 27/11/2012 at 12:06 pm | Permalink |

      Actually, another “takeaway” for me from this article is that the copper alone in an FttN network may cost “hundreds of millions” of dollars a year to maintain, on top of the rest of the system. I wonder how many “hundreds” that are talking about?

      • GongGav
        Posted 27/11/2012 at 1:37 pm | Permalink |

        I believe the figure mentioned was something like $700m or $800m per annum.

        Could be wrong, willing to be corrected.

    13. Brad Cann
      Posted 27/11/2012 at 12:07 pm | Permalink |

      and of course telstra feels it could upgrade its network to support FTTN, it stands to make bucketloads out of an incremental upgrade, AND it also gets to keep its assets and have the public pay to upgrade them AND get charged back to them at the same time.

      and a few years after that it will get yet another upgrade at yet an enormous amount of money to just put in the fibre the “last few hundred metres”

    14. Gamertech
      Posted 27/11/2012 at 12:36 pm | Permalink |

      “pointing out that in October this year, only about 1.3 percent of its telephone lines nationally suffered any kind of fault, and that it invests hundreds of millions of dollars each year keeping it that way.”

      When I set up ADSL2+ at my premises i had to get Telstra to swap between 8 different main pairs back to the exchange before we found one that wasn’t so degraded that it could actually hold a connection. (only 2.5km cable length to the exchange)

      After all that TW wants ME to pay them $2,500 cash to repair the fault on the current line to get rid of the static on the phone service and the daily dropouts on the DSL. TIO, ACCC, ACMA and the local political member all tried to help but TW just told em all where to go…

      Definitely sounds like the copper network is up to scratch…and lovely maintenance of it too…

    15. NBN Choice
      Posted 27/11/2012 at 1:41 pm | Permalink |

      I think people should believe the company that operates the network, rather than some random insider.

      Look, the Telstra network is not just one design.

      Much of the Telstra network is already designed as FTTN. Look at Mt Cotton in Brisbane – once part of the Redland Bay ESA, it was a subexchange. It is now its own ESA with the exchange feeding nodes across the entire suburb. Most people are within 300m of a node. Tomorrow, Telstra could put in VDSL2 cards and give everyone 100Mbit download speeds – and many would be OK with a single copper pair.

      Alcatel have designed a new ‘node’ which is so small it attaches to power polls. And it takes power from the connected CPE via the phone cable. Add in VDSL2 Vectoring and you can offer 100/40 to customers 400m from a node. Add bonding and double the bandwidth.

      Rapid deployment of FTTN is slowly becoming a reality, because NBNco are taking so long to connect and service customers. The Telstra top-hat upgrade is essentially an FTTN rollout without VDSL2 enabled. The Government could change the legislation tomorrow that prevents new 25Mbit+ networks being deployed, and Telstra could enable VDSL2 for a few million.

      • NPSF3000
        Posted 27/11/2012 at 2:03 pm | Permalink |

        “Telstra could enable VDSL2 for a few million.”

        Which, if it was as simple as you’ve outlined, would be a fantastic improvement for those few, for a few years.

        You still need the NBN to provide the ubiquity [needed now] and higher speeds [1+gbps needed soonish].

      • Brendan
        Posted 27/11/2012 at 2:17 pm | Permalink |

        I think you’re overstating the numbers a bit.

        Top Hat is targeting ~1850 odd locations, for a few hundred thousand connections. That’s a non-trivial upgrade, but it’s hardly millions of people.

        Also, to reach those sorts of speeds, you’re looking at VDSL2 (which would potentially require capacity upgrades, and more than one copper pair to consumers).

        To reach several million, the government investment would have to be equally non-trivial.

        And it still doesn’t address the costs of “FTTN > FTTH” upgrades. Who pays for that? It’s the end goal, and we’re already building that.

        • NBN Choice
          Posted 27/11/2012 at 3:05 pm | Permalink |

          The Telstra nodes use line cards. Swap out an ADSL2 combo card with a VDSL2 combo card. Worst case, swap the unit. All the cabling remains the same.

          VDSL2 is not new – and I have had Australian experience with it since 2008. Since then there has been a whole world of improvement – speed, signal quality, improved reach, reduced power, etc.
          http://www2.alcatel-lucent.com/blogs/techzine/2012/vdsl2-vectoring-delivers-on-its-promise/

          Also, the tophat upgrades went into areas where Telstra had not been able to upgrade the DA due to the space limitations – there are many other DAs out there that supported ADSL before tophat. There are millions of people who currently are serviced by remote exchange equipment, as they keep letting us know :)

          Add Tophat services, and customers within 500m from each exchange, and you can service most with single copper. Add in older RIM/CMUX customers who might be further out, plus exchange customers within 1km, and you can use 2 pairs (if the customer requires the speed).

          New estates get FTTP, and outer areas get PTP Wireless.

          Any customer within FTTN footprint can register interest to have upgrade to FTTP, as required.

          The Gov can then fund the gap, as it usually does, and rollout the rest of the metro areas.

          I agree, FTTP to replace the copper is good, but it really is going to take too long. Your asking some people to stay with ADSL2 for up to 10 years – what will the rest of the world be using over copper then?

          • Tinman_au
            Posted 27/11/2012 at 3:36 pm | Permalink |

            I think Telstra has had enough chances, time and money to get it right, and yet Australia still has some of the most terrible rankings in world when it comes to broadband, it’s past time to try something else to be honest…

            • NBNAlex
              Posted 27/11/2012 at 3:48 pm | Permalink |

              Although I think Telstra has improved markedly under Thodey and it seems they want to offer better deals (hence they buy Adam to involve themselves in the budget market)… I agree, they had their chance, all they had to do was take it… but no.

              Anyway that was the past… to the future –

              Come late 2013 if/when we have a new government, we can look forward to Telstra again getting back to it’s old ways (why wouldn’t they – as would any profit driven/listed company) when complete control of OUR is again returned to them, in the name of political ideology :/

              Unbelievable.

          • Posted 27/11/2012 at 3:40 pm | Permalink |

            @NBN Choice

            While I agree with your analysis of a possible FTTN rollout being feasible that way…..I think what you’ve lost sight of is the fact that the NBN IS HERE. NOW. In 3 years time some 1.5 Million people will be connected. That’s a HELL of a lot faster than waiting for the election, waiting till Feb of the new year to start legislation changes, mid year for passing, end of year to make Telstra do it and 2-3 years to do it.

            FTTN may have been a decent replacement in 2007 or even 2009 when NBNCo. was created. But not now. It is too late now. It would take more time to change direction now than would be saved in doing FTTN. Do you agree?

          • Brendan
            Posted 27/11/2012 at 4:34 pm | Permalink |

            Sorry, but you’ll not see an overnight change for FTTN.

            Both from a political standpoint and from a physical rollout. The timescale would be measured in years based on having to reverse NBN engagements and introduce new policy in Parliament, etc. Realistically it could be 1-2 years before anything actually happened.

            Far from “instant on”.

            It’s not going to be any faster, and as I’ve just said, doesn’t address what happens in a decade’s time, or more.

            • SMEMatt
              Posted 27/11/2012 at 5:24 pm | Permalink |

              I also doesn’t address the fundamental reason of why we are having this debate in the first place.

          • NBNAccuracy
            Posted 27/11/2012 at 5:15 pm | Permalink |

            NBN Choice, since vectoring is so successful, show me a link to a consumer grade VDSL2 modem that supports vectoring. Without modem support vectoring doesn’t work. As far as I am aware only multi thousand dollar commercial VDSL2 modems have just started to support it. A bit slow for something that was trial and found successful in 2008.

          • NPSF3000
            Posted 27/11/2012 at 7:45 pm | Permalink |

            “I agree, FTTP to replace the copper is good, but it really is going to take too long. Your asking some people to stay with ADSL2 for up to 10 years…”

            So are you – moving a few % to FTTN/VDSL doesn’t change the broad picture. However it could possibly slow down the Fiber rollout.

            • Shannon
              Posted 28/11/2012 at 3:42 pm | Permalink |

              Considering how long I waited on ADSL1 for ADSL2 while seemingly everyone else I know had it… and even now, all the lines in the street are so crap that I don’t get much better than ADSL1… I’ll wait for the FTTP

      • Ben
        Posted 28/11/2012 at 11:11 am | Permalink |

        I think you shouldn’t believe everything you hear. Why in the world would Telstra EVER admit their network was anything but good? The copper network is a huge asset for Telstra, they don’t want it to be devalued because of reports of poor quality copper. These sorts of reports could give Malcolm Turnbull an edge in any future deal made regarding FTTN.

    16. Tailgator
      Posted 27/11/2012 at 1:42 pm | Permalink |

      Corporate self interest further confuses the debate.

    17. Posted 27/11/2012 at 1:44 pm | Permalink |

      I pay for a 20megabit connection I am lucky to get 2 mb download on a good day I am 3 Kim’s from the exchange and the noise on the copper lines prevent me gettin a good connection speed. I am told that this is normal for my area and because the NBN is mowing soon they are not going to fix the copper connections.

    18. Brendan
      Posted 27/11/2012 at 1:54 pm | Permalink |

      Telstra would love to have someone else pay to have them (probably) upgrade their copper network.

      Enter stage left, the Member for Wentworth :)

      This would be the same Telstra whom hoovered up most of the handouts from the Coonan era to build a network that still has some regulatory issues, right? I don’t seem to recall a whole lot of complaining.

      Telstra will of course use active services as their measurement. Because, as far as their USO obligation is concerned, that’s all that matters. The percentage of “active” services that don’t currently have a voice fault, will be far below the actual figure of “inactive and damaged”.

      Telstra is simply playing both sides against the middle. They either profit from the NBN, or if it gets canned, they’ll profit from Turnbull turning the clock back 10 years.

      Can we use any of the above as some proof FTTN is actually viable after all? I’m sure Mr Turnbull is furiously typing something similar into his blog as you read this. PROOF, TELSTRA SAYS FTTN IS THE WAY TO GO.

      In reality, Telstra will say whatever the heck is most likely to ensure it will get money. Doesn’t mean a damning indictment of the NBN (or it’s push to modernise the network) or that FTTN is our saving grace.

    19. Douglas
      Posted 27/11/2012 at 1:58 pm | Permalink |

      So if the copper network is so old and decrepit, why the heck did we pay Telstra shareholders $11Billion to shut it down?
      If your answer is that it was to shift users onto the NBN, surely if the copper is that bad and fault ridden (and the NBN is everything it is made out to be) people would have just moved to the NBN anyway? Wouldn’t they?
      Have we given Telstra $11B for something that would have happened regardless? Are we mugs?
      It is my understanding access to the ducts is an extra rental on top of this amount also, so it wasn’t for that was it?

      • daniel
        Posted 27/11/2012 at 2:00 pm | Permalink |

        We are paying Telstra for the customer base + leasing the ducts/pits.

      • Brendan
        Posted 27/11/2012 at 2:05 pm | Permalink |

        It’s not the copper that was paid for. A refresher:

        http://delimiter.com.au/2012/03/08/telstra-seals-11-billion-nbn-deal/

      • Abel Adamski
        Posted 27/11/2012 at 2:06 pm | Permalink |

        We are NOT paying for the copper, that remains Telstra’s property, which is the issue the FTTN proposal will have to face.
        The $11Bill is for lease of Exchange facilities, backhaul, pits and ducts as well as customer migration – the customer migration is not the major component

      • Tinman_au
        Posted 27/11/2012 at 3:30 pm | Permalink |

        If an FttN NBN does get a go, they’ll need to pay more (probably a lot more) for the copper lines, the deal the NBNCo did was just rental for the ducts the copper runs through.

        • jasmcd
          Posted 27/11/2012 at 10:56 pm | Permalink |

          Wouldn’t that be grand, Telstra paid to implement a national FTTN system, paid for existing copper with work conducted on behalf of the NBN Co, whilst using all the money they pocketed to develop further Wireless Dominance. Possibly also offering their own FTTP network in cherry picked sites.

          On the wireless front, I am surprised they haven’t further dropped the prices of their plans to be closer to offerings from Optus. Telstra plans still have sting for those looking for decent data on a mobile phone plan.

    20. daniel
      Posted 27/11/2012 at 1:59 pm | Permalink |

      It took me over a month to get my connection fixed, and then a few months later, my connection stable at 4.4Mbps.

      If this doesn’t tell you the lame excuses that Telstra is putting out, perhaps we need to get the Goverment, early next year, before the Rampup of the rollout, a nation wide audit on Telstra systems.

      And Make sure what we are buying is good enough for the NBN and or Coalition Broadband Policy.

    21. Bill Tarrant
      Posted 27/11/2012 at 2:24 pm | Permalink |

      Anybody who has dealt with customers’ broadband faults over Telstra’s copper will have a very low opinion of that copper in general, although they will agree that in places it is very good indeed.

      They will suspect that Telstra is talking about voice faults, which is another issue. And they will wonder whether Telstra’s reluctance to log broadband faults, and their greater reluctance to escalate them, might also somehow lie behind these Pollyanna statistics.

    22. MikeK
      Posted 27/11/2012 at 2:42 pm | Permalink |

      Well I can only go by my experience with Telstra, at the previous unit I lived in I had Telstra come around to my place 3 times in 5 years to replace wall sockets, the copper had turned green and stopped working and still had unacceptable line noise every time it rained. At my present place of residence I cant even get ADSL but still the line has noise every time it rains. In my experience its rooted.

    23. MikeK
      Posted 27/11/2012 at 2:51 pm | Permalink |

      Your being bias in your photographs Renai, how about putting up a photo next to the one you have showing one of the Telstra terminals in poor condition.

    24. Posted 27/11/2012 at 3:35 pm | Permalink |

      I’d have to agree with most of the posters here Renai.

      I’ve no reason to doubt Nick Ross’ source. It’s just as legitimate as most of Malcolm Turnbull’s sources on whether FTTN is going to be ‘cheaper, faster and more affordable’ or his Telatra sources which tell him 24Mbps in 4 years is doable.

      These are 2 entirely different ‘faults’. One is current connections a, of which the fault rate is 1.4%. I don’t doubt that. One is NEW connections, which could very well be 30%. And NEW connections are the important thing with FTTN.

      I think comparing what Telstra are saying on this to Nick’s article is disingenuous on Telstra’s part at best.

      • Ully
        Posted 28/11/2012 at 1:18 pm | Permalink |

        I agree. Nick bets his un-named sources against Malcolm and his un-named sources, is a poker game from hell. Just shows how false Mal’s ‘evidence’ is. Now, I know someone who was told by someone, that this internets thinghy is not going to last. True, honest, believe me, as I am a politician. Ha ha.

    25. NBNAlex
      Posted 27/11/2012 at 3:59 pm | Permalink |

      Since we are talking evidence based, lets do a comparison we do ergo know.

      Regardless of whether the copper is rooted or not, new fibre will be vastly superior to either good or rooted copper.

      End of argument

      :)

      • MikeK
        Posted 28/11/2012 at 6:31 pm | Permalink |

        +1 Alex and the only government thats going to delivery the superior product is Labor.

    26. Karl
      Posted 27/11/2012 at 4:20 pm | Permalink |

      “However, to pull out one un-detailed, unsubstantiated fault rate on one area of Telstra’s network and extrapolate that into a claim that Telstra’s copper network is so “rooted” that it would be “completely impractical” to deploy fibre to the node in Australia is just, in my opinion, horse twaddle.”

      Um, Renai, it was the insider who said those things not Nick Ross. Additionally, the state of the copper is not the reason he was talking about that would make FttN completely impractical, it’s the state of the record keeping.

      • Posted 27/11/2012 at 4:28 pm | Permalink |

        Statement by Nick Ross in article:

        “It puts the implementation of a Fibre to the Node infrastructure (as favoured by the Coalition as an alternative to the NBN’s current Fibre to the Premises-based roll out system) further into doubt in regards to its fitness-for-purpose.”

        • Karl
          Posted 27/11/2012 at 4:46 pm | Permalink |

          Yes further into doubt, but that’s from a starting point of “we don’t actually know which bit of copper is connected to which house”. That was the meat of the ABC article, Telstra have cleverly turned it into a discussion of failure rates by seemingly wilfully misunderstanding that.

    27. StevoTheDevo
      Posted 27/11/2012 at 4:45 pm | Permalink |

      It’s also important to note that for ADSL or voice only 1 line per premises is needed and, as opposed to 15 years ago, not every premises has an active line (Mobile only households are common). Therefore it’s easier than ever for a line tech to find a functional pair, and therefore easier than ever for Telstra to report high compliance numbers.

      FTTN however needs 2 lines per premises, and its foreseeable that mobile phone/data only houses currently unable to get ADSL, might want to get a hard line back into their house.
      All of a sudden, the demand for active lines more than doubles!
      This is where the failure rate of all the lines in a pillar, not just the active lines, becomes highly important.

      I suspect the 30% failure rate is accurate and IS indeed a major thorn in the side of any FTTN proposal.

    28. Posted 27/11/2012 at 4:50 pm | Permalink |

      @StevoTheDevo

      Just FYI, FTTN does not REQUIRE 2 pairs. Bonded pairs for VDSL2 speed improvements and range extension, require 2 pairs. But VDSL will, itself, work on 1 pair. But it will only give maximum 80Mbps at about 300m. That’s where Turnbull’s 80Mbps comes from.

    29. mr_tim
      Posted 27/11/2012 at 5:01 pm | Permalink |

      yes, and telephone exchanges don’t have fires ;)

      • Shannon
        Posted 28/11/2012 at 4:23 pm | Permalink |

        Mine is attached to the local cop shop, and we NEED the copshop, only reason I haven’t torched it lol

    30. Posted 27/11/2012 at 5:18 pm | Permalink |

      Hi Renai,

      I too got a press release from Telstra and also one from Turnbull’s office saying the same thing. We’re talking about two totally different things.

      This is not anything to do with the number of customer complaints coming in. This is to do with the copper in the turrets – 30 per cent of them are either failed or connected to the wrong line.

      As such performing a one-to-one translation would mean there would be around a 30 per cent failure rate.

      The customers attached to them right now are likely to be fine and aren’t necessarily complaining about anything.

      My guy did the audit. What he says concurs with what others are saying in the industry. I’ll update my article. Please update yours.

      N

      • Posted 27/11/2012 at 7:07 pm | Permalink |

        Nick,

        I don’t know what you’re talking about, but I didn’t get a “press release” from anyone. I got some information from Turnbull’s office and I asked Telstra for some further information.

        As for your comment about stats, I’ll be happy to do a new article if you can produce some evidence to back up what you’re saying … your article made the somewhat controversial claim that Telstra abandoned its FTTN strategy because of those fault rates, and the similarly controversial claim that fault rates are running at 30 percent.

        I was around when the FTTN shebang went south, and I can tell you from my perspective it didn’t have anything to do with fault rates — it was more the fact that the Government’s expert panel rejected every FTTN proposal at that stage, in favour of a FTTH model. The 30 percent claim may be correct, but without evidence it’s hard to know, and without context such a figure is essentially meaningless anyway (for example: Does BT have the same fault rate on its copper network, which is being upgraded to FTTN?). Did you actually see the audit which your source authored?

        Cheers,

        Renai

        • Brett
          Posted 27/11/2012 at 7:59 pm | Permalink |

          others could be, how well do they maintain their systems? I remember reading last month about how NBNco is having to remap the cable maps as the ones provided by telstra are inaccurate in some areas and engineers have to manually find what pair goes where.

          Also, theres different evironmental factors as well, ie 40 degree days in summer. When it rains, is it drizzling or do they have sudden downpours? It would be hard to say, what works for BT would work for aus/telstra.

        • Tinman_au
          Posted 27/11/2012 at 9:56 pm | Permalink |

          Hi Renai

          If Telstra didn’t abandon it’s FttN and go FttP then how would you explain http://fibretosouthbrisbane.com.au/ ?

          I fully understand there were “other issues” involved (Anna “encouraging” Telstra, locking in customers to non-ACCC/ACMA regulated infrastructure, etc), but they could have used wireless and/or FttN and achieved the same ends. But they decided to use FttP…

        • Ben
          Posted 28/11/2012 at 12:05 pm | Permalink |

          Hi Renai,

          I’m pretty sure Nick Ross was referring to before Labor’s NBN proposal. Back in the days when Telstra were looking into doing it on their own.

          I know that Telstra wanted regulatory certainty before rolling it out and that is probably the public reason they used. However, who really knows what the underlying reason really was. Telstra certainly aren’t going to admit that their network isn’t up to scratch.

          Thanks to both Renai and Nick Ross for being one of the few reporters NOT spreading NBN FUD.

          • Posted 28/11/2012 at 12:24 pm | Permalink |

            Um, what happened there was that Telstra tried to roll out their FTTN network without having the requirememnt to sell wholesale access to it (this is the way their mobile network operates). They knew that the ACCC and the Government would never agree to that. But the fact that they tried created the discussion which led to the 2007 $4.7 billion FTTN election policy of Labor.

            Telstra didn’t can its initial FTTN policy because of technical reasons. They canned it because of legal reasons.

    31. Abel Adamski
      Posted 27/11/2012 at 5:28 pm | Permalink |

      I note Malcolm has not updated his site since the innovation speech. No mention of either of Nicks articles

    32. Posted 27/11/2012 at 7:46 pm | Permalink |

      Here’s my follow up to this article. Hopefully sorts a few things out.

      http://www.abc.net.au/technology/articles/2012/11/27/3642266.htm

      • Posted 27/11/2012 at 10:33 pm | Permalink |

        I appreciate your work, Nick, but I still don’t think your new article demonstrates that FTTN is impossible in Australia … it has no international context. I’m sure BT has faced many of the same issues that you raise — but they are deploying FTTN in the UK right now, as is AT&T in the US, and also in many other countries. In fact, in Europe it seems like FTTN is very much in vogue at the moment.

        I still believe Labor has the much better policy here, but I think you’re basically claiming that FTTN is not possible in Australia; but I also think you haven’t presented enough evidence to make that case yet.

        • Tinman_au
          Posted 28/11/2012 at 3:23 am | Permalink |

          I don’t believe FttN is as practical here as it is in those other places though Renai, can you think of any of those examples you site where they are rolling out FttN where it isn’t an existing telco using infrastructure (copper) they already own?

          The situation would be different in Australia. The Liberal plan would require either rental or purchase of the copper from Telstra, and if just using the ducts cost $11 billion, I think they’d want a fair bit more for what is the actual core network of the system. As you say, it wouldn’t be impossible, but I’m not sure Malcolms plan would actually work out any cheaper, or be any more practical than the current NBN, once you factor in how much Telstra would want for the “nuts and bolts” of the POTS system…

        • Ben
          Posted 28/11/2012 at 12:29 pm | Permalink |

          Hi Renai,

          I’d just like to point out that Nick Ross never said it was “impossible”. He was just implying that it is impractical to roll out FTTN here.

          If Nick’s insider is to be believed, and I think his insider is more trustworthy than Telstra, then FTTN is impractical for Australia. At least, in comparison to the Labor’s NBN. Telstra are never going to admit fault with their network because the second they do, it is worth considerably less.

          Do you really think that Telstra are going to go above and beyond their USO obligations? They’re a private company with share holders to answer to, they’re going to do the bare minimum. The USO obligations only require that Telstra have a working voice service available.

          As Tinman_au was saying, you can’t really compare a potential FTTN network in Australia with BT’s FTTN. It’s an apples and oranges comparison. I wish people would stop doing it.

          • Posted 28/11/2012 at 12:35 pm | Permalink |

            “I think his insider is more trustworthy than Telstra”

            On what basis do you believe this? An anonymous source with no verification or documentation, compared with one of Australia’s largest corporations, which publicly proposed to deploy FTTN only a handful of years ago?

            Evidence lacking ;)

            • Ben
              Posted 28/11/2012 at 12:56 pm | Permalink |

              Hi Renai,

              Over a company that is always going to put its own interests first? That has a requirement to increase profits for their shareholders. There is no way to provide evidence to support our claims because only Telstra and its contractors (unable to comment on it) truly know the true state of their network. Telstra are biased.

              What about this: http://www.smh.com.au/articles/2003/11/14/1068674351979.html?from=storyrhs

              Do the math: November 2003 + 15 years = November 2018. Copper network needs to be replaced within 5 years, from the horses mouth. One can reasonably assume that the copper network would have a lot more faults as it gets closer to its use-by date.

            • Lachlan
              Posted 28/11/2012 at 2:27 pm | Permalink |

              Personally, I think this is a classic case of comparing oranges and lemons.

              Telstra’s 99.92% availability refers specifically to the Customer Service Guarantee, which is VOICE ONLY.
              So by this measure, the copper network is an orange.

              However, for FTTN Telstra has given NO EVIDENCE for its network quality. Given that Bigpond is a big ADSL2+ provider, Telstra could do a simple data match showing say, ADSL2+ synch rate at 90%, median and 10%, and compare that to theoretical ADSL2+ line speed based on each services line length, all wrapped up in a nice graph. Telstra should be able to get the line length information straight out of their copper line GIS, if they are

              If the copper is usable, then for each line length the sync rate should be close to the maximum. Otherwise, as the median drops, it is evidence that the copper is actually rooted, and a lemon, as the insider has claimed.

              You should be wary of overselling the Telstra PR flacks spin, and actually require relavent evidence.

              • djos
                Posted 28/11/2012 at 3:36 pm | Permalink |

                When I was in the Telco/ISP industry the favorite trick (of last resort) to get lines fixed (very, very common), was to port the customer onto TW, get the customer to log a fault with Telstra and hey presto the fault would get fixed – if you didnt do this the line issues would be ignored by Telstra!

    33. Kevin Cobley
      Posted 27/11/2012 at 10:14 pm | Permalink |

      The copper’s totally rooted in Katoomba

    34. MarkD
      Posted 27/11/2012 at 10:52 pm | Permalink |

      If 1.3% of their active network had faults in October. Can we extrapolate that to mean 12 * 1.3 = 15.6% of their network has a fault per annum?

      Not 30% to be sure, but dang, that’s a huge proportion of faults per annum.

    35. sydboy007
      Posted 27/11/2012 at 11:38 pm | Permalink |

      I was speaking to an ex Telstra tech and he was telling me how when he was a linesman he’d get calls asking if he had jumpered up certain connections. He thought it strange why they kept asking him till he found out that bascially the system was set up so that if the jumpering was not marked as completed it would automatically be set to a fault status.

      This way Telstra could show they had a 100% success rate for jumpering services.

      Also Telstra is not really acknowledging the fact that a service does not have to be faulty for the connection details at the pillar to be wrong.

      Finally, there’s a lot of people I know with crap ADSL quality. Are they logging faults on it? Sometimes, but certainly the number of issues out there unreported are way higher than what Telstra is trying to pretend it is.

    36. Abel Adamski
      Posted 28/11/2012 at 12:55 am | Permalink |

      OK the Coalition forms government and after all the processes and turmoil eventually starts their Broadban Solution based largely on FTTN.
      What if many streets, especially smaller ones sucha as Courts are one or two pairs short being able to provide a service, what then?, run in new copper, individual fibres for the shortchanged citizens, or a local fixed wireless a la rural covering all the one or two pair short streets in the area

      Every cable pair will have to be identified and tested for fit for purpose.
      It will depend on who will own the copper , if Telstra they will have to make it fit for purpose as Telstra is having to do with pits and ducts.
      Or if the Coalition network owns the copper, have they paid also for the unserviceable pairs.
      If there is insufficient viable pairs, what then,? FTTH or run another copper cable or provide a wireless service a la Rural in which case own towers or lease space off Telstra’s towers – how many of those will be required?

      I am looking forward to the comedy show of implementing anything like the Turnbull plan

    37. WIlly
      Posted 28/11/2012 at 7:46 am | Permalink |

      Yes well my connection was fixed a couple of times just like my neighbour now he operates like me off optus cable on a voip service but na telstra said everything was fine as we had a voice service and the internet was working when the tech was there. Honestly I just got sick of the constant calls so gave up and had the service disconnected, I am counting the months till they roll the fibre down the street thankgod its soon to be weeks. Oh and I will be voting for labor again just because I think everyone should have a right to what I will have.

    38. stoffs
      Posted 28/11/2012 at 10:30 am | Permalink |

      lets go back to 2003……

      http://www.theage.com.au/articles/2003/11/14/1068674351979.html

      yet magically.. it’s going to be good from 2012 – > 2020+ ?

    39. Tinman_au
      Posted 28/11/2012 at 11:23 am | Permalink |

      “Telstra will replace its century-old copper wire phone network with new technology within the next 15 years, saying the ageing lines are now at “five minutes to midnight”.

      Well there ya go, the copper was “five minutes to midnight” 10 years ago, and we only have another 5 years to make sure it’s replaced (according to Telstra).

      • Ben
        Posted 28/11/2012 at 12:35 pm | Permalink |

        Since Tinman_au didn’t provide a link with his quotation: http://www.smh.com.au/articles/2003/11/14/1068674351979.html?from=storyrhs

        Renai and Malcolm Turnbull need to see this. Who am I kidding, Turnbull will just simply dismiss it or ignore it.

        • Posted 28/11/2012 at 12:51 pm | Permalink |

          With respect to that article, we can discount the BigPond email stuff, as it has zero to do with the copper network.

          For the rest, that ‘five minutes to midnight’ claim appears to have been disproven by the fact that the network is still functioning with millions more ADSL connections than Telstra had at that stage and that Telstra itself proposed upgrading it to FTTN several years later, in late 2005. The article itself doesn’t mention FTTN; not sure what Warren was proposing at that stage.

          Yes, Telstra’s network has problems. Lots of them. But there has been no significant evidence presented so far that these problems would make it “completely impractical” to upgrade the network to FTTN.

          • Ben
            Posted 28/11/2012 at 1:20 pm | Permalink |

            Hi Renai,

            “five minutes to midnight” is a subjective term. He was implying the network was on its way out.

            From the article: “Telstra will replace its century-old copper wire phone network with new technology within the next 15 years, saying the ageing lines are now at “five minutes to midnight”. ADSL or FTTN, the copper lines remain the same.

            The implication there is that the network needs to be replaced very soon, if not within 15 years from when that article was written. The fact that millions more are using the network is not evidence to support your claim. We are talking about the quality of the last mile, not the interconnecting networks which are fibre anyways.

            It is impractical to roll-out FTTN only to have to move to FTTH very shortly after your network is complete. There are many other reasons, some you are aware of and agree with that also make it impractical to halt the FTTH NBN and move to FTTN NBN. When I say impractical, I don’t mean impossible.

            FTTN can definitely be rolled out but no one, not even Telstra, really knows what issues are going to crop up with the ageing copper network. How much are they going to delay things and how much extra is it going to cost because of it? Suddenly FTTN isn’t looking like a cheaper or faster option.

          • Tinman_au
            Posted 28/11/2012 at 1:53 pm | Permalink |

            I think the main point of the SMH/Age article in this context, is it’s further evidence that the copper network is (at least in many places) a century old and has “issues” that will require it to be replaced, either with a different technology like fibre (to the N or P), or with new copper.

            I don’t think it would be a stretch of the imagination that for Malcolm’s NBN to be workable, a lot of the current copper infrastructure would need to be replaced/upgraded (not that that would make a lot of sense considering how the future looks to be fibre and wireless at this stage).

            I’m basing that statement off both Nick’s anonymous insiders anecdotal evidence (though if you look around, he isn’t the only one that’s ever commented on the state of the POTS), as well as the company that actually owns the infrastructures own public statements (at a Senate inquiry no less).

            I’m more than willing to accept Malcolms NBN to be a workable plan (and then argue the merits of FttP vs FttN), but there seem to be so many issues with it currently that don’t seem to be well thought through (how will MT’s funding work? How much will he pay to access the copper considering access to the ducts alone cost $11 Billion, how much of the copper is actually viable for a further 20-30 years and what will it cost to fix it up so it will be? etc)), that issues like this need to be addressed before you could say “Yeah, that’s doable” vs a system that’s actually in the process of being “done” right now.

            And I don’t blame you for wanting the facts of the matter, I find Malcolms “FttN will be quicker, faster, cheaper” with nothing else except “I’ll tell you after the election” (as you yourself have pointed out in http://delimiter.com.au/2012/09/26/turnbull-wont-disclose-rival-nbn-policy-details/) to be too light on facts and too heavy on aspirations to get me on board it being a credible alternative at this stage. If he wants people like me to see it as anything beyond a political FUD tactic, he needs to flesh it out a bit more (he needs to at least address/recognise that there could well be some serious issues with the copper he wants to use and how he thinks he can work around that).

            As the potential Communications Minister (and a “player” in this field), I actually do expect him to have answers to the questions people have about his plan and a “ball park” idea of what it’ll cost, the actual “dollars and cents” costing of it can come later.

        • Tinman_au
          Posted 28/11/2012 at 12:51 pm | Permalink |

          My bad, the link was in stoffs post ( http://www.theage.com.au/articles/2003/11/14/1068674351979.html though the SMH article is the same one).

    40. Rob
      Posted 28/11/2012 at 11:29 am | Permalink |

      Talking about the figures of maybe 30% rooted copper network. If a good deal of copper is now no longer redundant, and Telstra are not investing in copper upgrades now since the NBN is deemed to take over, imagine the costs involved if suddenly FTTN is announced.

      I also wonder what the quality of the copper cable is like in new estates? Is it good quality, is it as redundant, is it better protected? In comparison to copper pairs in an old estate? Clearly most of the problems are usually in very old suburbs, and in rural areas.

      It would be good to get a map of the age of the network in each area. I’m sure some places could do VDSL easily and others would struggle.

    41. Abel Adamski
      Posted 28/11/2012 at 12:43 pm | Permalink |

      http://www.abc.net.au/lateline/content/2012/s3642410.htm

      “MALCOLM TURNBULL: For the bulk of the brown field built-up residential areas of Australia, what we will do is take the fibre further into the field, but not to the customer’s premises.

      MATT PEACOCK: For those last metres to the home, the Coalition will negotiate to use Telstra’s existing copper wire – unless you’re rich enough to pay for more.

      MALCOLM TURNBULL: You can have a fibre service area of a couple of hundred households getting the fibre node connected to the copper, but there might be one person – perhaps it’s a business – that needs to have very, very high speed connectivity, and you can selectively run fibre optic cable to that premise. Of course they’d have to pay more for it.?”

      Basically the Old British Class Distinction which I see echoed in many comments .
      The peasants get second rate that will become unreliable over time, the Nobility and affluent can afford to pay to have FTTH

      This is an assumption that only the Nobility and affluent will operate business from the premises or are worthy of remote medical services or have children that can use FTTH in education/research or being innovative entrepreneurs on a shoestring budget.

      Yet the Conservatives squeal “Class Warfare” at the drop of a hat

    42. Abel Adamski
      Posted 28/11/2012 at 12:51 pm | Permalink |

      A reminder
      The NBN is a ubiquitous Business capable readily and cheaply upgradeabl Communications Platform for the Nation for many decades to come that will pay for itsel and return a profit for the taxpayer, at which point profits will go to reducing charges and extending and upgrading the Network.

      Broadband is no longer a commodity for those that can afford it, it is now essential National Infrastructure

      A further quote from the lateline article

      “The Opposition spokesman on Communications, Malcolm Turnbull, says that in government the coalition will continue the NBN in a reduced, privatised form rather than destroying it as previously promised.”

      Privatised so welcome to the world of rip off and minimal service and standards – profit and ROI are the ONLY priorities of the private sector

    43. stoffs
      Posted 29/11/2012 at 10:13 am | Permalink |

      no wonder they have such a low “error” rate….

      i’ve been on the phone regarding a line outage at one of our offices…. 15-20 mins on hold, then get put through to another department… 15 minutes on hold..

      only to be told… that i’m being connected back to the original people i talked to…. currently on hold

      I’m sure most people would have just hung up …

      so from telstra’s point of view… no fault there!

    44. stoffs
      Posted 29/11/2012 at 10:19 am | Permalink |

      oh yay…

      put through to the wrong department again !

    45. Phil
      Posted 29/11/2012 at 10:53 am | Permalink |

      Telstra are perfectly to hide behind the Truth In Numbers while being very careful to NOT tell The Whole Truth.

      FACT: in many regional/rural settings Telstra will QUITE HAPPILY give up (completely and forever) on your copper landline and install a 3G ‘base station’ (with associated high-gain antenna) so that you can have a fault-free phone service.

      Sure this increases day-to-day costs for them to supply your service, however:

      – no faults means no costs from constantly repairing faults
      – no faults means they can claim surprisingly low ‘fault percentages’
      – as an “added benefit” no copper line means your ONLY choice for broadband is via-the-basestation (ie Telstra captive customer) or Satellite which is almost worse than useless (latency, poor service quality, expensive, etc)

      Please continue to ignore any and all statistics released by the PR mouthpiece speaking for Telstra, it’s 95% marketing-speak and 5% self-serving “technically true while ignoring the whole reality” rubbish.




    Get our 'Best of the Week' newsletter on Fridays

    Just the most important stories, one email a week.

    Email address:


  • Enterprise IT stories

    • Super funds close to dumping $250m IT revamp facepalm2

      If you have even a skin deep awareness of the structure of Australia’s superannuation industry, you’ll be aware that much of the underlying infrastructure used by many of the nation’s major funds is provided by a centralised group, Superpartners. One of the group’s main projects in recent years has been to dramatically update and modernise its IT platform — its version of a core banking platform overhaul. Unfortunately, the $250 million project has not precisely been going well.

    • Qld’s Grant joins analyst firm IBRS peter-grant

      This week it emerged that Peter Grant, the two-time former Queensland Whole of Government CIO (pictured), has joined well-regarded analyst firm Intelligent Business Research Services (IBRS). We’ve long had a high regard for IBRS, and so it’s fantastic to see such an experienced executive join its ranks.

    • Westpac dumps desk phones for Samsung Android mobiles samsung-galaxy-ace-3

      The era of troublesome desk phones tied to physical locations is gradually coming to an end in many workplaces, with mobile phones becoming increasingly popular as organisations’ main method of voice telecommunications. But some groups are more advanced than others when it comes to adoption of the trend. One of those is Westpac.

    • Ministers’ cloud approval lasted just a year reverse

      Remember how twelve months ago, the Federal Government released a new cloud computing security and privacy directive which required departments and agencies to explicitly acquire the approval of the Attorney-General and the relevant portfolio minister before government data containing private information could be stored in offshore facilities? Remember how the policy was strongly criticised by Microsoft, Government CIOs and Delimiter? Well, it looks like the policy is about to be reversed.

    • WA Govt can’t fund school IT upgrades oops key

      In news from The Department of Disturbing Facts, iTNews revealed late last week that Western Australia’s Department of Education has run out of money halfway through the deployment of new fundamental IT infrastructure to the state’s schools.

    • Turnbull outlines Govt ICT vision turnbull-5

      Communications Minister Malcolm Turnbull has published an extensive article arguing that the Federal Government needed to do a better job of connecting with Australians via digital channels and that public sector IT projects needn’t cost the huge amounts that some have in the past.

    • NZ Govt pushes hard into cloud zealand

      New Zealand’s national Government announced a whole of government contract this morning for what it terms ‘Office Productivity as a Service’ services. This includes email and calendaring services, as well as file-sharing, mobility, instant messaging and collaboration services. The contract complements two existing contracts — Desktop as a Service and Enterprise Content Management as a Service.

    • CommBank reveals Harte’s replacement whiteing

      The Commonwealth Bank of Australia has promoted an internal executive who joined the bank in September after a lengthy career at petroleum giant VP and IT services group Accenture to replace its outgoing chief information officer Michael Harte, who announced in early May that he would leave the bank.

    • Jeff Smith quits Suncorp for IBM jeffsmith4

      Second-tier Australian bank and financial services group Suncorp today announced that its long-serving top technology executive Jeff Smith would leave to take up a senior role with IBM in the United States, in an announcement which marks the end of an era for the nation’s banking IT sector.

    • Small business missing the mobile, social, cloud revolution iphone-stock

      Most companies that live and breathe the online revolution are not tech startups, but smart smaller firms that use online tools to run their core business better: to cut costs, reach customers and suppliers, innovate and get more control. Many others, however, are falling behind, according to a new Grattan Institute discussion paper.

  • Blog, Enterprise IT - Jul 5, 2014 13:53 - 0 Comments

    Super funds close to dumping $250m IT revamp

    More In Enterprise IT


    Blog, Telecommunications - Jul 5, 2014 12:12 - 0 Comments

    What should the ACCC’s role be in guiding infrastructure spending?

    More In Telecommunications


    Analysis, Industry, Internet - Jun 23, 2014 10:33 - 0 Comments

    ‘Google Schmoogle’ – how Yellow Pages got it so wrong

    More In Industry


    Blog, Digital Rights - Jun 30, 2014 22:24 - 0 Comments

    Will Netflix launch in Australia, or not?

    More In Digital Rights