John Linton
You may have seen various press reports earlier this week claiming that an independent study had shown that Telstra provided the best ADSL2 service in Australia - hardly Earth shattering news had it been true and most people would expect that to be the case without spending money on such a study. However, at least according to this article:
http://www.itwire.com/content/view/18837/127/
the news that Telstra's ADSL2 service was the best, as stated in their previous article, was incorrect because the organisation purported to have run the study hadn't actually commenced testing yet.
Very odd when you think about it.
Nevertheless, the point remains, why run such a study at all? (and I'm assuming that it would actually be possible to effectively test the various ADSL2 services offered by the different carriers/service providers - which I personally would doubt).
Telstra has the largest network and therefore offers services in more than twice as many locations as anyone else - by definition Telstra's ADSL2 service in those locations is the best because there is no other service.
In all other locations (including the CBDs of the major cities) where other providers would have viable numbers of customers to make viable provisioning of back hauls possible, Telstra would still be the 'best' because it would have an enormous advantage in already established back hauls and hardware infrastructures and certainly wouldn't run into whatever the issues experienced by Adam and Internode late last week/early this week - a factor always likely to occur in new network infrastructures.
Having said that - what is it possible to test? I'm sure IDC, with its decades long record of 'testing and evaluation' will have come up with sensible criteria but reading the New Zealand report it was pretty much ho hum and didn't seem to address the actualities of network performance across the multiplicity of aberrations that are the actuality of the Australian residence to local exchange network and which cause so much angst in individual customer's use of ADSL2 services.
Exetel as a user of two carrier's ADSL2 networks in Australia for almost two years with 15,000 customers currently connected (not a huge number but a 'sample' likely to be far in excess of any 'sample' IDC end up using) can categorically state that the performance (under optimum copper residence to exchange conditions) doesn't vary between the two carriers. Both services are excellent and deliver consistently high speeds over two years with no latency/drop out/speed degradation by time of day issues.
HOWEVER:
both carriers, on different sectors of their national network at different times, have delivered less than optimal speeds for varying lengths of time. This is very annoying to the end customer and, in the case of a 'piggy in the middle' such as Exetel probably something much more than annoying. Having been in the communications business for more than five minutes it's not surprising (not any excuse) and as with all growing networks the problems reduce in frequency over time. Anyone who used the Optus wire line network in the early days of deployment may remember some of the issues that regularly occurred then but have been non-existent for more than ten years now.
Personally, as an end user of an Optus ADSL2 service, I've never had a problem in almost two years (Mosman/Sydney Lower North Shore). As a supplier I am more than well aware of the fact that Optus has underprovisioned different sectors of its network at different times and has caused a lot of unhappiness to various groups of customers each time there is any sort of delay in upgrading the back hauls from the different LACs around the country. Optus is unfortunate' compared to the smaller ADSL2 network deployers as it much more quickly reaches 'saturation' points on its different network sectors than the smaller deployers (with their much smaller numbers of customers do). Again, that's no excuse for Optus but it is an explanation that can be understood.
I recently looked at a 'smaller' ADSL2 network that had zero contention on its 1 gbps loop and that would be the case for some time (because of its relatively low number of users). However I would be unsurprised if that network deployer was in any hurry to incur the very significanrt capital (even by leasing) costs associated with putting in the replacement 10 gbps exchange controllers once the 1 gbps limit was raeched on the first exchange - I could, of course, be quite wrong.
So you would have to say that Telstra, having learned from the early days of its deployment of 8192/384 ADSL1 services (which from Exetel's experience had a significant provisioning problem on various back hauls in the early days - none since) would have not provisioned its ADSL2 network below the 10 gbps level or much higher based on their known huge user base and likely take up. Who could match that? No-one in Australia.
So when IDC gets round to 'measuring' the performance of Australian ADSL2 services it would be a total surprise if Telstra's services weren't ranked Number One (or Numero Uno).
However my experience, and the experience of the vast majority of the other 15,000 'Exetel' ADSL2 users is that our ADSL2 experience is trouble free over a long period of time and is delivered at a vey low cost.
I don't see the point of such a 'study' (other than it makes IDC a lot of money).