My 500GB main disk died. No click of death. No warning from SMART. Nothing.
The disk had been running a little poorly for the last few months - an unfortunate fight between ACPI and APM meant it's XP partition was never the same again.
But this morning - nothing. Not even a parting goodbye.
The darn thing won't register. I doubt it's even spinning up.
Now, it wasn't completely unexpected.... I have a new 1TB drive with Vista Ultimate on it (and please don't start on why not Win7... I did not have $400 spare) that I was progressively moving to, at a speed the wife would accept.
So now we're on Vista, I have a dead HD I'm wondering to do with, and a lot of stuff to migrate very quickly.
But it got me thinking about the cloud, for the first time in a long time (especially given it's my job to have my head in the clouds).
I've lost no email. My precious media of the children is on a seperate HD (which is about to be backed up AGAIN!). But I would love to be able to have a safe store for what is important to me and my family.
With my TelstraClear Cable Internet, I can restore my email and any PC pretty easily - although getting Vista and the apps patched up again came to about 3GB in a day - but for real content? forget it. WHo can offer me 1TB of storage? and what Internet service can I use to upload that amount of info?
A home server might be the answer... but that also has a hard disk that will eventually die. On the story goes.
I've been struggling to think about what use a fast fibre network could be. This is one of the uses.
But then economic reality steps in... my wife says she'd pay $30/month to back our data. That's real world consumer expectation, and she doesn't care what's involved in making it happen.
SO who will be first with a 100Mbps Internet service and unmetered 1TB in the sky. $30 a month up for grabs.....
And streams _very_ well at HD on my cable connection at home...
Under the heads of agreement announced this afternoon, Telstra would provide access to Telstra facilities and progressively migrate Telstra traffic onto the National Broadband Network, subject to regulatory approval. The agreement for these terms will have an approximate value of $9 billion.
Separately, the Federal Government has agreed to progress “public policy reforms” with an attributed value of approximately $2 billion. These basically involve changes to Telstra’s current universal service obligations with the establishment of a new Commonwealth entity – USO Co – which will deliver unprofitable services. USO Co will receive a maximimum of $100m in annual taxpayer contributions with the rest to be funded by presumably increased industry contributions. It will take over Telstra’s USO obligations from 2011.
Telstra also said it has received a written agreement from the government that it will be able to participate in LTE spectrum auctions under the deal.
“This is a sound outcome for NBN Co because when finalised it can maximise the use of existing infrastructure and accelerate the roll out of its network,” NBN Co CEO Mike Quigley said in a press release.
NBN Co added that Telstra would likely become its largest customer. NBN Co will pay Telstra for migration of traffic on to the NBN and the decommissioning of its network.
The Heads of Agreement also provides for NBN Co’s use of Telstra’s “existing fit-for-use infrastructure, such as ducts, pits and conduit and a right to acquire Telstra backhaul services and space in Telstra exchanges. While there is a considerable amount of negotiation and contractual work to go, we believe this agreement is a significant step forward to creating a more competitive telecommunications industry,” Quigley said.
Telstra expects to be able to place the deal to a shareholder vote in the first half of next year. The deal is subject to both that vote and ACCC approval.
About this time 9 years ago, I was sitting in a management meeting with my Director of Marketing, and we were reviewing the latest iteration of a truly dismal situation. The telco bubble had burst, 3G was a bust, and a huge internal project to install GPRS and MASSIVE ISP infrastructure had just gone through it's 3rd rescoping. From the hubris of a fully connected world where everything ran off a PDA - a state the world only managed to get to in 2009 with the iPhone 3GS - came the realisation that the only proposition available was. basic colour WAP over GPRS, using the classic Ericsson T65 handset.
Out of sheer frustration, my director uttered the immortal words in his best James Bond "Gentlemen. I want a killer app in 2 weeks". a meeting that still amuses me to this day, although it was anything but funny at the time (the project bill by this point was bordering on ?55m). Since that time I have many more 'amusing' meetings, where out of sheer frustration at the world of complexity, decisions have been made to pursue a technology or a specific way of selling come hell or high water, hoping this eventual silver bullet will prove to be what unleashes torrents of success and money.
A little bit like throwing $48bn at a fibre network (Australia) or $10bn (NZ - you didn't think the country would be rewired for less than that, did you?). This is the silver bullet to unlock all our problems. Especially if you actively destroy the existing technology and force everyone to start again.
Or maybe deregulating the Building Act, to make it faster cheaper and easier to build new houses. now there's a silver bullet to get the economy going. Too bad that people don't get the difference between treated and untreated timber, or what living in NZ's wet climate really is like (we have been here for a great many years. not that much has changed!)
The expression, 'No Silver Bullet', comes from a book written by software engineer Fred Brooks in the 80's. He wrote
".there is no single development, in either technology or management technique, which by itself promises even one order of magnitude [tenfold] improvement within a decade in productivity, in reliability, in simplicity." While Brooks insists that there is no one silver bullet, he believes that a series of innovations attacking essential complexity could lead to significant (perhaps greater than tenfold in a ten-year period) improvements.
Over the christmas break - while having a ball with my Telecom XT connection - I read a book called 'How the Mighty Fall' by Jim Collins. His main premise was that successful companies get arrogant, start to decline but fail to recognise until the decline is too late, then throw resource and money at the problem until it's fixed. which almost never happens, and they carry on as 'zombie' companies until closed down or sold off. Basically you'd have to be spectacularly bad to fail outright. it's more a case of gradual decline. Probably true for a large corporate trading on old successes.. I think smaller companies don't have this sort of luxury!
So how many companies s do you see out there lurching from strategy to strategy, product launch to product launch, always consuming more and doing more but never quite seeming to succeed?
Success is'nt about just striking the right mix - although that's important to. It's about endeavouring to understand your market, trying, adjusting, reflecting, going back and refining and keeping at it. It's about consistency mixed with execution. It's about staying the course and holding your nerve, and not backing away when it starts to look hard. And it's about trusting the people around and beneath you, while also supporting - and verifying - that things are on the right course. It's also about communicating widely and taking feedback from all places - not just those you work with day by day, but everyone. If the cleaner can tell you what you're about - you know a lot of effort has gone into living and breathing what you're doing.
It may not be widely known, but Apple launched the iPod in 2001, as a followup to the nascent MP3 market started by the Diamond RIO (I had one of those. good gadget). But they applied continuous effort to innovation, change, and were lucky enough to have some powerful management able to leverage content. In 2003 iTunes went Windows. In 2004 the 15GB iPod's came out, and then things really started getting interesting.
Imagine sitting in that meeting at Apple in 1999, with a Director proclaiming "Gentlemen. I want a killer app in 2 week", and proposing what iPod has become. Brass ones for sure.
Bring on the Broadband.
Over the the christmas break in 2009/10, I took the opportunity to buy a Telecom Prepay Mobile Data stick, running on the new XT network, and used it while away on holiday at the beach. Modern lifestyle means it can be very hard to be away from what you get used to: instant access to information, banking, email. and even sometimes just filling in quiet time when it's just too hot to go outside (and to avoid the other great holiday mode behaviour. eating!)
I also compared coverage with the Vodafone 3G network, which has the benefit (or curse) of being able to fallback to GSM/GPRS mode. In that review, I found the Vodafone network wanting.
This Easter break, we took the opportunity to have another beach break, this time in the Coromandel (near Hot Water Beach), and I have availed of modern technology during the high noon heat. The township I'm in is small but well formed, some of the houses look to be worth staggering amounts (especially those with walks down to the beachfront), and it has been genuinely gratifying to see people out camping and benefitting from the facilities.
The campsite we're at suffers power failure like clockwork between 7 and 9pm, when the draw is high. Water pressure falls to a trickle around 4pm. The general store reminds me of NZ in the late 70's (pile stock high, sell straight from the box), and some of the buildings reflect the building glory of that era (shades of brown everywhere).
But the township also has wide area WIFI Coverage (really) at $8/hr at not so bad speeds, spread across a area around 1 square kilometre, run by a group of shops on behalf of the community and visitors to the area. Vodafone's coverage and speed is good, but my gadgets fallback to 2G far too much for me.
Telecom's XT coverage is brilliant, with HSDPA readily available. Not bad for a coastal town with few residents.
This got me thinking about the pace of change in this country, and how telco is evolving as a whole. In my first role at Telecom (1996), the manager at the time told me how 'every year since 1987' had been the year of Mobile Data, the next big coming thing for the industry (1987 being the year Telecom rolled out the AMPS network - target forecast 50k subs by 1995). It really took the launch and settling of GPRS - 2003 - and widespread launch of 3G -2005 - along with the iPhone - 2007 - before anyone could really say that Mobile Data was here to stay.
But it is, and Telstra's model for Next G - go far, go fast, go deep - shows exactly what you can do if you build a high-quality mobile network. Too bad it's data pricing can't compare with a $35/month DSL connection. Mobility still enjoys a premium - for now - but the relentless downward pressure from government, consumers, competitors and so on, means this business is changing just as the huge pressure in fixed calling dropped prices to 'irrelevant' status (compared to mobile).
So Telecom suffered unforgiveable failures with XT, and is now running ads in the papers of photogenic staff who lived through that period (although Chris Q really looks just over it). So Mr Hamburger has stopped the Florida commute, and maybe a local will be appointed to take ownership of the line. The network is there, it's running. and it's still performing well.
That mobile network failures can take this much media attention shows exactly how important Telecom - STILL - is to this country. The difference: they are getting over the introduction and settling of new technology (just wait till the IP Voice services start rolling out), and just getting into the game for the next 10 years.
Where's everyone else?
Spot muggins under the Poster... notice his coffee on the desk next to him.... and how bald he thinks he looks....
Well, what a february it was, as a start to the new decade.
Telecom suffered the most ignominious of sustained failure, with their shiny new WCDMA 850Mhz network having fallen over several times, with the blame eventually falling on AlcaCent (the supplier) and a sacrificial head from Telecom (Frank Mount, whose contract ended in June 2010 anyway). Add to the woe problems with the sacrosanct 111 emergency call system, and the Telecom must feel like someone has a voodoo doll of Spot furnished away, full of broken fibres (in the stitching). Add the new comments from the former boss Ms G about how things were'nt like this when she was there, and you can't help but feel that someone probably doesn't like you!
But never fear, as Vodafone NZ and 2degrees have invoked the fates and called on the combined forces of Murphy and Sod, taking switching campaigns to market on the back of these failures, inviting customers to join a reliable network. so now it's just a waiting game for their network to fail (as indeed it already has, but not to the same scale).
I'm interested in what is happening inside Telecom, to it's sales staff, support staff, field staff and everyone involved in service delivery. Because it is from these folks that customer's present and future take their lead on purchasing decisions. and there is nothing that has more inertia or destructive force that delivery entity that refuses to supply. A loss of confidence is hideously difficult to turn around - collective memory being a self-healing, reinforcing wall of despair (at least where salesfolk are concerned). After all, who wants to sell something they perceive to be a lemon, and have to deal with the fallout?
It has been my jolly good luck to have taken career roles where a turnaround has been required, coming from high expectations to failure to having to rebuild. and it never quite is the same again. The hubris is gone, replaced by an air of 'whatever', and a market that actively questions what is being said, even if all the elements are fixed. My article on using Telecom Mobile over the Christmas period has been challenged, but I stick with my assertion that Telecom has the better network. especially now that the failures will force people inside the company to sit up and take notice.
Microsoft required Windows 7 to get over the failure of Windows Vista, even though the service packed product is quite usable (I have Vista on one of my machines and it works very well). But markets have long memories, and product quality is so good these days, that customer replacement cycles have extended and people are more selective.
Service providers (like Telco's) can be frustrating to work in at times - you know good stuff is coming but you can't really say anything, because inevitably Sod's Law kicks in and you miss the dates you said. Equally people don't want to hear 'coming soon, we are working on it really really really'. they want solid dates, and they expect them met. Fair enough - but technology is such a complex beast that dates given are almost always under threat, and compromises/workarounds have to be made to get to the dates - or if the above can't be done. you slip. As companies get larger and have more systems, the complexity becomes exponential. so things take longer.
Speaking to some folk I know (none of which are involved with XT), we surmised that somewhere in the deployment of XT a compromise was made. A workaround was introduced to meet a date - "of course we can reuse xxx component, it's a few years old now but it still works" - which is pure poison with technology. Vendors can only backup, within reason, their own kit (and even then they struggle against all permutations), but telco's rarely rip and replace an entire system.
So 'integration' is required - words that are enough to send a shudder down your spine if you've ever had to do it. Integration is a level of joy reserved only for the brave.
But when integration goes right then wrong, products don't work well. When products don't work well, the humans cannot explain, respond or repair fast enough. When response is slow, confidence is lost.
Confidence lost is not a good place to be, personally, professionally or collectively. The YouTube clip of the Downfall/XT mashup is below, which I found quite funny. But the original is very sobering (when reality has struck home, when you have to take responsibility. and the first response is so very human, and so very appalling).
What conversations are being had inside the great Service Providers everyday? at what point are the accusations made of 'lies, betrayal, hiding things etc'?
I wonder how confident the Telecom folk are feeling today?
2degrees Mobile and Telecom XT launched in 2009. Vodafone, Orcon and towards the end of the year, TelstraClear, launched offers using unbundled copper lines (telecom copper connected to other providers infrastructure). Telecom Wholesale created and launched a huge range of IP-only access products, allowing a service provider to offer Internet, Internet+IPVoice, or Private Network products, with a range of service speeds and characteristics.
The Commerce Commission waded in on commercial wholesale rates for voice calling for fixed and mobile networks, and of course the government pushed the Crown Fibre Company concept. The field service industry was shaken by Visionstream in Northland, and the march of Downer EDI as the technicians of choice. Chorus began the hard work of turning variable grade copper and successive investment decisions of the last 30 years into a semi-viable business model. and the last vestiges of the NZ Post Office were tackled (there are still chalk and blackboards in use out there, in the depots.).
TelstraClear flipped back to Vodafone for mobile, Black and White, Digital Island, Compass and Callplus all took Mobile (Wholesale) offers to market, most settling around the 25c/minute mark for mobile (good rate that.). FX Networks continued to cobble together a fibre network of own-grown and in-sourced, the utility companies tried to become telco's (Northpower, Vector.).
So lots of technology investment, lots of change, lots and lots and lots of money being spent. Or at least, planned to be spent.
NZ was about 4.2m people five years ago. I think we're about 4.3m now. The NZ telco industry is forecast to shrink from $5.6bn to $5.2bn by 2012. and the collective investment of the last few years has been about $4bn as I recall.
Commercial models based on the old world of TDM - POTS for Voice, with DSL grafted on top, moving up to Centrex or Trunk lines - is disappearing as the technologies become just another service, sitting on top of an IP-enabled access circuit. Mobile phones' continue to be the access device of choice (note I said MOBILE - not the mobile service providers).
The notion of paying $xxx for a phone line, $yyy for mobile, $zzz for Internet, and a raft of rates and charging models (1 minute billing, per minute billing, mobile mb being expensive but fixed being cheap). is all challenged, thrown in the air, and being thought about by many people. Some have no experience in economics, the reality of making targets, the challenge of recovering expensive installation costs. Some like to push numbers as a soundbite. Others overthink the market or allow inertia to act as a cushion to aggressive or poor sales practises.
Identity is the next big change to happen, as a way of standing out or just surviving this brutal market. Superior technology was an excellent way of creating identity, until the competition caught up. Telecom has caught up, and at a national level - so the technology line is reset (good time to buy shares btw - their performance can only improve from here). What do the others stand for? are their services even relevant now? Being $5 or $10 cheaper than the next one in line is hardly a compelling reason to change providers. And would you really trust your critical business apps to a company that comes in thousands of dollars cheaper? that gap in revenue is being achieved by some means, and not all of it is new technology - a lot has to do with tiny investment in backup systems and resilience.
The major movie franchises spent most of this decade being successfully recast - REBOOTED - along the lines of reality, grit, plausibility, and connection with the audience. Daniel Craig's James Bond is brilliant. Batman is enjoyable. Sherlock Holmes was enjoyable and grounded in the real when dressed in fantasy. So it goes.
2010 marks the start of the reinvention of NZ Telecommunications. The models of the old are challenged and under severe strain. The game has been lifted for all players. There is less money to be made but more people chasing it.
Companies will fold. Products will fail. Teams will be reorganised. Strategies will be challenged. Pressure will cause innovation. Careers will be made. Companies will succeed.
I look forward to being part of it!
BTW: This is the first article I wrote using Windows Live Writer and posting via the Metaweblog API Mauricio made available quote some time ago... worked very well...
This is a light article looking at Mobile Internet options for the holidaying, professional. You might think it sad that one would want to remain connected rather than swap to holiday mode – but take a minute and consider the world of today, in NZ at least.
The last decade has seen an embrace of ‘The Internet’ for providing services, information, tools and entertainment. There is so much that is now online, that only when you don’t have access do you realise how different the world is. Add some real spice and introduce two children under 10, and the weather we’ve had in the xmas/new year of 2009/10.
When was the last time you banked a cheque or used telephone banking? Or called a movie theatre to find out listings? Try going ‘internet-free’, then go on holiday… and try to manage your life. Technology may be considered a scourge of modern life. I see it as a way of making life better, easier and that much more manageable.
Babies learn to get attention by crying at a certain pitch, designed to induce response in tired parents. Pre-tweens learn two more dreaded words designed to tweak that nerve - “I’m Bored”.
I called a local movie theatre to get movie listings for the kids, which had a recorded announcement of all the movies on. The audio level was terrible – I struggled to hear anything – and there was no answer when trying to get a human. Movies were read in random order and with too much detail (rating, movie length and observations on movie) –too slow and too irrelevant for me. A quick trip to their website – sorted in 5 minutes flat (and $40 later for the tickets!). The online world, especially once it is familiar, is tough to leave.
Continuous Improvement: Product Launch’s & Compromise I’ve worked for several industries, mostly in telco and mobile. I have launched many products, and been through the subsequent after launch hangovers (when reality sets in). So I thought I would apply what I know to one of the biggest launches in 2009: the Telecom XT Mobile network.
A product launch is a series of compromises between commercial dates, technical realities, your own expectations and those of the sponsor(s). At some point, you need to go to market with whatever open bugs, missing tools, missing material are, for one simple reason: launch creates focus. It also puts the spotlight on, which means every move, success, failure, complaint or problem has huge visibility and is pored over.
No product is perfect at launch, no company is ever ready to deal with every type of usecase that gets presented – nor should they have to be. Being ready to solve all of yesterday’s problems is insanity – but you need to be ready to deal with ‘in-life’ as it crops up, and a company needs to make resources available while they work on ‘the next big thing’.
This state is sometimes referred to as ‘crossing the chasm’, between the hubris of launch and the dullness of reality. Not dealing with the chasm, or pretending it hasn’t really happened, is just crazy. A product launch is just the start – rapid response to in-life is more critical to success.
Telecom NZ has never run with GSM in NZ, although it has come close to swapping from CDMA. It took several years and a management change to sign off building a new network - WCDMA 850Mhz (same as Telstra for Australia). It’s just one technology - no legacy GSM network here – and everything is brand new, with little legacy baggage to contend with. The only real challenge is migrating the CDMA customer base to WCDMA, without having to give away 1.5m new devices.
This approach means you lose a mature, stable and reliable platform. Tools and technologies developed for the old have to be rebuilt, product enhancements and pricing plans need to be re-justified to the latest round of management, business rules need to be re-qualified… the list is endless. The recent failure of the XT network is a classic example – everything is new, which means the technology is still being stabilised.
The same conversation is happening for the PSTN network – moving to an IP Voice service forces the same discussions to be had. New technology often doesn’t have the same richness as the old - or the costs - and comes with its own foibles. Changing core technology is easy to justify in a presentation, and a lot of work to live through in practise – having been through it several times at every company I’ve worked for.
Knowing this I avoided buying an XT service for many months, waiting for the network to stabilise, for the sales channels to be trained and educated, for new propositions to be developed and refined, and for XT ‘mark 2’ to come out. Having a need for a Mobile Internet service, I took to seeing what the market offered and went from there.
The purchase Around the XT launch in May 2009, Vodafone launched a ‘$1 for 10mb a day’ offer, for any type of Mobile Internet use, with subsequent use costing $1/mb. The 10mb is reset daily, and is quite a good offer for very light use with a Smartphone, but not so hot for a datacard.
Vodafone do have options for prebuying 100mb or 512mb, but these require the user to be on ‘Supa Prepay’ and to purchase an add-on that recurs every month. Not so good if you only need it for a few weeks (I would have preferred to purchase a once-off block as I need). I have a Supa Prepay sim, as well as a Vodafone HSDPA datacard.
Before Xmas, Telecom launched a Prepay Mobile broadband offer. For $99 (or free to existing Telecom Xtra Broadband customers) you get the MF626 USB Modem (with USB extension cable), Prepay SIM Pack with $20 usage and 512mb bundled data. Subsequent data is either $29.95 for another 512mb, or partial usage that goes something like – “climbing usage of 10c/MB to 300mb, then 212mb uncharged usage thereafter”. Telecom gets to realise the revenue faster this way, rather than spread the $29.95 over 512mb (Money stored in a prepay account can’t be classed as revenue until it’s used for a service; while in the ‘idle’ state, it’s a liability on the companies balance sheet).
The purchase experience was one of the best I’ve experienced; I went into the Leading Edge store in Palmerston North on the 28th of December, and the entire process took 7 minutes. The girl at the counter was well-briefed and answered all of the tricky stuff I threw at here around compatibility, coverage and usage. She volunteered I shouldn’t activate the SIM Card until the 1st of January (else my 512mb free usage would be lost on the 1st of the month – this is a quirk of the Telecom billing system, but a nasty one nonetheless). Kudos to the Telecom retail folk – the training and execution was superb.
The environment I am staying near the beach, and depending on where I am in the house, I get between 0-2 bars for Vodafone, and 0-1 bars for Telecom - when near the window. Despite locking the Vodafone Mobile Connect software to 3G only, the Vodafone card still spends a great deal of time swapping between the UMTS and GPRS networks – which means lengthy delays as the card cycles between technologies. This is disappointing and should not really be happening in 2009. Both Vodafone and O2 developed their mobile datacard software simultaneously in the UK around 2004 (I know this because I led the O2 side for their 3G launch), and both use the same software vendor. Five years of product enhancement and development should mean the software is pretty robust.
To be fair, getting the Telecom software to install on my XP machine was an exercise. One only needs to insert the modem in a free USB slot, and your PC will auto-launch the installer – in theory. My machine recognised the Modem as a USB stick only and would not give me access to the embedded software – some obscure setting somewhere prevented autorun from working. I have Windows 7 on the same machine and was able to correctly install using that OS, as well as retrieve the software and copy to the local harddisk, to manually install with Windows XP.
I wasn’t interested in calling the helpdesk for support – I’m not doing a critical review of Telecom – and am lucky that I am able to self-support. No doubt driver install wobbled somewhere in the process – drivers being the biggest bugbear of Windows – and am pleased that the driver experience is better with Windows 7.
In Use What a difference new technology makes! Overall service speeds on Vodafone Mobile Internet, where I am, quite frankly suck. The card I have is not faulty, and I would not expect many other data users where I am so cell loading should be very light. My comments are subjective – I am experiencing the service as any user should – and to me, the Vodafone was just poor. When locked onto 3G for long enough, actual speeds for Internet, was disappointing.
Telecom (which has lower signal strength according to the Connection Manager) rocketed by comparison, with average performance of a vintage 512k broadband connection – which was more than enough to let me achieve what I wanted (banking and a little web use).
What I had forgotten was how greedy Windows is with an Internet connection. The moment a tunnel was available, all manner of updaters sprang into life (Windows Update, Windows Defender Update, AntiVirus Update, Apple Update, Spybot update and so on). I switched off as much as I could find, and even then there was usage going on (some malware perhaps?). On Vodafone, 10mb gets chomped through pretty quick, although they are good enough to notify you via text when you are 2mb away from your data cap. The Vodafone portal has excellent self-management tools – another example of services moving to the web – while I struggled with the Telecom website, and don’t wish to waste my precious bucket o’ data trying to figure it out.
Using Google Chrome, the surfing was fast and I achieved what I wanted – managing money, movie tickets for bored children, and keeping up with the rest of the world, as I have always done.
Conclusion I wrote this article to explore the concept of being truly mobile anywhere, with a useful piece of kit (a laptop). I also wanted to comment on product launch’s and product evolution, using a real-world example… and I think the Telecom experience is better than Vodafone. The self-service website needs a lot of work, but for what matters – actual customer experience – Telecom Mobile broadband is great.
Mobile is not yet at the point where it’s a true fixed substitute – the technology quality and pricing take care of that. But it’s good enough for the occasional user – and I do mean occasional – that prepay mobile internet is ok to use and fairly predictable (no bill shock either – if you run out of money, everything just stops).
Good on you Telecom for an acceptable service. Roll on the full XT Mark 2 – and then Mark 3!
Note: for clarity, I work for TelstraClear.
One thing that has done - and still does - my head in, is the discrepancy between speeds and the different technology types. It's easy enough to say 'it don't matter', except when it does, and I find it matters an awful lot, quite often.
What do I mean?
I mean those awful speeds. 24mbps Internet! 7.2mbps HSDPA! 1gbps Fibre, fwoar!
And yet, there is real nuance that affects what you get as a user.
For Internet, since that's the topic of NZ time and again...
The speed that matters I believe, is the IP speed you get for your applications. I use TelstraClear's cable service at home, and I can regularly download a file from a high-capacity server (eg microsoft.com) at 1.02MB/s - that megabytes per second, which measure the size of a file. To get that speed, I have the 10mbps - megabits per seconds, which measure the absolute speed of transfer at that second. As a rough rule of thumb, 10240mb /10 = 1024MB. Take 10 as being how many bits are required to represent a piece of information, with some overhead.
I know that I will be able to get those speeds, assuming all technology is lined up and functioning correctly (I am not getting into a debate over configuration and what TelstraClear does or doesn't do).
But now it get's interesting.
ADSL/ADSL2+ technologies use an older method of transfer called ATM (Asynchronous Transfer Mode)... this doesn't readily map with 100% efficiency into the IP world, and you get something called the ATM-Tax affecting speed. If your ADSL2+ connection can connect at 24mbps/1mbps to the local DSLAM... then the 'tax' will consume between 10-15% of that bandwidth in overhead and syncronisation... or about 2.5mbps... and on it goes. Except of course most people don't get those speeds, and they get something closer to 10mbps, which reflects that it's not a fair comparison between HFC and DSL. With DSL, especially on the Telecom network, sometimes you're up and many time you're not.
802.11a/b/g/n/q/s/x and every other letter also has wonderful overheads. At Radio layer the access point may be able to talk to the wireless network card at speeds from 11mbps to 300mbps... but once you take error correction, delay, latency, and generally poorly implemented software... you're lucky to get IP throughput closer to 5mbps on 802.11b, 20mbps on .11g, and who knows what on the .11n spec. Translation: what you PC gets is not what's on the box. But if you're trying to stream a High-definition video from a media PC to your TV on a wireless connection, you need about 9mbps bandwidth for the IP layer - if you're only getting twice that on average, wow!
Now for something else: Ethernet, which is a way of networking computers together, but also has overhead and loss. A 100mb Ethernet port will not give you 100mb IP - far from it. There will be loss to inefficiency, error correction's, retries and so on. If you purchase a service that is 100mb IP... then the Ethernet access had better be larger than 100mb so you can get the full bandwidth.
Head hurting yet????
Does 1GB fibre connection mean I'm getting 1GB IP bandwidth to my computer? of course not... and most PC's would struggle to process such a torrent of information. 1gb more likely will give about 500mbps, which translates to about 50MB... still huge, but quite different to what's on the box!
I won't go into HSDPA or that lot - that's a whole new level of panadol time...
Now, I'm not a techie, and i'm sure many services could be made to run more efficiently. But I think if you buy a service with a speed promise, then it should be at IP level - which is what your applications will use.
Off to get the panadol....
antoniosk's profileAntonios Karantze
I'm a born and bred Wellingtonian, and have chosen IT and Telecommunications as my industry, as a Commercial Manager.
Credits include but are not limited to:
- The O2 Xda smartphone range, and O2's range of 3G Mobile Internet services
- Numerous TelstraClear Mobile and IP voice products
In my journey through the industry I have worked at
- Bellsouth NZ
- Telecom NZ
- ICO Global
- T-Mobile International (formerly One2One Communications)
- O2 Plc
- Vodafone NZ
I'm a fan of technology, and what it can do for people and business... and I enjoy bringing new things to market and seeing them grow. Enjoy the blog, take the time to think about what I write - it's not technology heavy, and is my reflection on life and the people around me.