Are we starting to see Broadband Fatigue in New Zealand?

I wrote about Telecom finally returning full-rate DSL to customers after all kinds of silly wrangling to avoid doing so. You'd think that after all the bitching and moaning about the poorly performing broadband here, the news that Telecom will take the handbrake off DSL by October 25 would be reasonably big news.

Not so: there's not one word about it in the Herald, Stuff/Dompost or NBR.

It's the first step towards removing all the artificial constraints that make NZ DSL so sub-standard, and it will benefit businesses for instance wanting to use Voice over IP and remote access - interleave can now be switched off to reduce latency and the higher upstream speeds will make the service far more usable than before.

Was it just too "geeky" and difficult to cover this? Maybe, but it's more likely that media are bored with the whole broadband debate - just like most other people are. Since DSL came out seven years ago, it's been marred by endless negatives: bill shocks, poor reliability, high pricing, sluggish performance and of course, the farcically convoluted Yes, Minister style regulation that Telecommunications Commissioner Douglas Webb concocted.

The public is expecting more of the same, and the perception of DSL is that it's a complete dog that's only worth buying if there's nothing else available. Telecom maneouvred DSL into this position instead of developing the service (we should all have had ADSL2+ last year already).

Full-rate DSL sounds great but thanks to Telecom's chronic underinvestment, its backhaul network isn't up to delivering the high speeds people quite rightly expect from their newly unshackled broadband connections. Telecom hasn't upgraded its Alcatel ASAMs which have only a single 155Mbps ATM circuit, and which are already showing signs of straining under the load of rate-limited customers. More full-rate customers and well, it's just not going to be the fast broadband you think it will, and there you go, another negative DSL story.

Posted by Juha Saarinen at 12:57 PM | Permalink | Comments (5)

Today's Herald carries a story about the Commerce Commission rejecting Telecom's argument that power has to be reduced on short copper loops in order to provide DSL service on longer ones.

The Commission's decision follows a report from consultant Paul Brooks, which in essence rubbished all of Telecom's proposed spectrum management claims.

In fact, the report says the term "spectrum management" is a misnomer, and that we should be talking about "interference management" instead. Any reputable carrier should put in place an interference management plan as it rolls out DSL, the report says.

However, despite starting the DSL deployment in 1999 already, Telecom has yet to do so. That's right, Telecom's been running the DSL network by taking some pretty severe shortcuts for the past seven years. Even so, it's worked more or less fine now, the report says, so why the hurry to deploy a managment plan now?

Was Telecom trying to prevent an accelerated DSL deployment on the back of further wholesale providers? It's not entirely clear what the incumbent's reasons were for launching the delay strategy, assisted by obfuscated technological arguments that its network manager Alcatel augmented in a report to Telecom, and which was released to media as well. That report backfired on Telecom though, as it was widely seen as an indictment by the network manager on how poorly the incumbent maintained its copper.

Clearly what Theresa Gattung said when the new regulation was announced is... complete rubbish. The above shows that there's no change at Telecom. It's still trying to obfuscate and delay at any given turn, and if you don't believe me, look at Telecom's submission to the Select Committee deliberating on the proposed new regulation currently. Telecom's trying to wangle out of any clause in the new regulation that would matter, and proposes instead a regime that would be different to today's only on paper - in effect, Telecom would carry on regulating itself. I hope the Select Committee is able to see through Telecom's "legalistic argument" (which incidentally Theresa promised not to hide behind).

The ComCom's spectrum management report and decision is hard to find on its Byzantine website, but here's a direct link to it. The report by Brooks is a good read, as it demolishes Telecom's arguments one by one. I'm curious as to why Telecom is allowed to present obviously fallacious arguments like that to a government agency actually.

Posted by Juha Saarinen at 10:44 AM | Permalink | Comments (4)

Symantec reckons it has found two new proof of concept worms that "attack AMD processors" rather than Windows as usual.

It's not completely correct to say that the malware in question attacks processors, but this quote from Vincent Weafer, senior director of Symantec's Security Response Group, illustrates the danger:

"If I can get to the processor level, potentially I can really start tying myself into the core hardware. I can potentially evade some of the kernel protection and user protection. There is an attraction to virus writers to get to the lowest level possible," Weafer told vnunet.com.

In other words, anti-viruses may not be able to detect such a virus, as it is able to hide in hardware. Such a virus would also evade Microsoft's kernel security enhancements on Vista, and virus writers could conceivably target any operating system this way.

Presently, the proof-of-concept worms that Symantec have found, w32.bounds and w64.bounds (for 32 and 64 bit processors respectively) are rated as low risk and harmless. They infect Windows executables only currently.

This development is quite ironic, as AMD has made rather a song and dance about the virus protection in its processors.

With both AMD and Intel looking at hardware-based security solutions, it's no surprise that malware writers are heading in that direction too.

Security is indeed a moving target...

Posted by Juha Saarinen at 2:00 PM | Permalink | Comments (0)

Any content that can be digitised will be pirated it seems. PC World is no exception.

An anonymous PC World reader tipped me off about the August 2006 issue of the US edition being available on Bittorrent:

The reader says it's the complete magazine, scanned in and turned into a 25MB PDF file. It's 72 pages, content-only with the ads removed or blanked out.

A quick look at the same site shows that it's not the only issue of PC World available:

I haven't checked them to see what they're like, or which country editions they are, but... I'm amazed that someone took the considerable time to scan in all the pages in a fat magazine, snip out the ads and then convert the lot into a PDF. This clearly isn't a question of saving money because the time and effort expended would be far more than what the magazine is worth - and the person who scanned it probably bought it anyway.

So, why bother? I don't get it. Why not just buy it and get a higher quality publication that's easier to read - and in doing so help feed us lowly paid geek journos who come up with the goods in the first place?

Posted by Juha Saarinen at 11:28 AM | Permalink | Comments (3)

Someone had to pay for the Biggest Mistake Ever at AOL, and it is CTO Maureen Govern and two other employees.

In case you missed it, AOL released the search logs of some 600,000 customers to the web because... well, it seemed like a good idea at the time.

Russell Brown over at Public Address covered the gaffe and with the help of Netguide sidekick Nigel Horrocks discovered some rather disturbing stuff:

User 1627022 reeled off searches for "alt.binaries.pictures.erotica.child.male", "boylove belarus" and the like - and then "university jobs new zealand" and "social work new zealand", which took him to the recruitment pages of four New Zealand universities and, creepily, www.cyf.govt.nz. Someone might want to have a look into that. (A reader has pointed out that the string of access to very, very dodgy images may have been for legitimate research purposes, and that did occur to me, but I'm not sure that even a researcher needs to spend that much time looking at actual child porn. At the least, some questions would need to be asked for purposes of clarification.)

What's the bet though that sacking the people responsible for such a monumentally stupid act won't do much to restore people's confidence in AOL?

Update Computerworld NZ story here.

Posted by Juha Saarinen at 3:43 PM | Permalink | Comments (0)

I see in the National Business Review that property developer Peter Yealands plans to install optical fibre in a new project near Havelock by the Marlborough Sounds.

(Note to NBR: "1000 MB" is probably better written as "1Gbps" or even "1 Gigabit per second" instead - and why not say "network" instead of "high speed information and communication system"?)

That's all good stuff - kudos to the developer, which obviously understands what people in 2006 want.

However, putting good networking infrastructure into new housing projects is one of those blindingly obvious things that should not generate column centimetres in the media anymore. The optical fibre network should go in there together with other utilities like water, electricity, and gas. The houses themselves should be wired up with gigabit Ethernet inside.

What about the added cost though? Well, fibre is actually cheap. Copper on the other hand has shot up in price - just check out how much it costs to rewire the electricity mains for a house now.

The long and short of all this is that with a fat optical pipe to the outside world, you don't need to put in a copper networks for last-millennium phone and DSL service. I don't know if Peter Yealands is brave enough to Just Say No To Copper, but they should.

In fact, property developers should be encouraged to avoid copper completely. How about if councils around the country put into their development plans that no new copper plant can be rolled out, only fibre-optic - or a similar, high-speed capable networking technology?

A scheme to upgrade the copper network in older areas to fibre should also be devised by planners. If we can commit lots of money to "undergrounding" power lines for largely aesthetic reasons, why can't we start work on modernising the country's communications infrastructure?

Going back to the Marlborough Sounds, before anyone gets too excited about the prospect of 1Gbps connectivity and rushes off to plonk down a deposite on a Yealands house, remember the local network has to connect to the rest of the world. I'd be surprised if ThePacific.Net which apparently is the ISP partner has 1Gbps national connectivity, or international bandwidth anywhere near that level. Still, it's an encouraging development, literally.

Posted by Juha Saarinen at 11:30 AM | Permalink | Comments (0)

What is Microsoft SMaRT? When you suspect malware on your machine you do one thing. You run a traditional anti-virus/anti-spyware scan. After scanning and cleaning if the machine is still displaying symptoms, what is your next step? Currently there is no solution/tool/answer for this problem. You are left 'trying' to identify processes, modules, aseps, infected files that may or may not be running on the machine. The problem is even more frustrating if the symptoms are determined to be caused by hardware failure and not malicious software.

That's the blurb for Microsoft's new Suspected Malware Auditing and Report Tool or SMaRT. This is what it it looks like: The idea seems to be to establish what software is kosher and what isn't, through a set of heuristics - unknown at this stage. Does this sound familiar? I think it does. It reminds me of so-called generic software, which looks for known malware behaviour. Microsoft's version seems to work by identifying software known to be good, and then flagging the unknown stuff. This seems to me like a risky idea. The 90-94% accuracy is impressive, but what if the false positive is a file belonging to a competitor? Microsoft could end up in very hot legal water over that. Or, what if the file is unknown yet crucial to the system operation? Some users may delete stuff flagged as unknown. Can't quite see how SMaRT will help against "rootkits" or Windows kernel extension malware either, as these are below the file system level and thus invisible to the tool. All in all, I'm not sure what Microsoft is trying to achieve with SMaRT - maybe it's a good idea to build up a database of "safe software", but surely there are better ways of doing than this? While it was still availiable as a clickable link,I tried to install the executable, SMarRTInstall.exe, but it seems to be a a weird mish-mash of HTML served up as a GIF image. That was about two weeks ago, and it still hasn't been made available. Hmm.

Posted by Juha Saarinen at 11:29 AM | Permalink | Comments (0)

Idly clicking through Auntie Hairoiled's website trying to find something that I can't remember what it was now, I read this:

PC World will today become the UK's first national retailer to stock gadgets that turn iPods into mini FM radio stations, albeit with a limited broadcasting range.

Ahh... what now? PC World moving into retail... in the UK? Well, no, not quite: over in Blighty, there is no PC World. Instead, it's PC Advisor for the Brits.

Idly clicking through Auntie Hairoiled's website trying to find something that I can't remember what it was now, I read this:

PC World will today become the UK's first national retailer to stock gadgets that turn iPods into mini FM radio stations, albeit with a limited broadcasting range.

Ahh... what now? PC World moving into retail... in the UK? Well, no, not quite: over in Blighty, there is no PC World. Instead, it's PC Advisor for the Brits.

That story in the Herald though (it's actually from The Independent) is quite interesting, about the iTrip FM transmitter accessory for iPods. At first, I thought, very nice, you can spam others by transmitting whatever's on your iPod. Wireless podcasting!

A closer read says otherwise though. I think the idea is that you get yourself an iTrip and then... play back the content of the iPod on your car radio or the tuner at home. How prosaic.

Why not build a two-way transceiver device for iPods, so you could share and share alike? I guess the music studios (and film ones too, now that we have video iPods) would go ballistic at the thought, but I reckon it'd sell.

Just imagine flash iPodMobs, where lots of dead cool gen-somethingortheother peeps hang and mash-up each other's iPod content.

Hmm, on second thoughts, scratch that idea. It'd look completely ridiculous.

Posted by Juha Saarinen at 6:03 PM | Permalink | Comments (0)