There are times when lables just aren't helpful.

Anybody who has ever realised that their hierarchical document folders now go fifteen deep, total about 200 folders and have duplicated several of them, will realise that sometimes it's best to be a bit vague about things.

This is why I like tagged organising systems - you can get at the same information several different ways which, as any halfway-competent website designer will tell you, is a Good Thing.

Could we, I wonder, stop fucking about with motorcycle categories? Adventure bikes are also Road-Trail, Dual-Sport, All-Road and Big Trailies. We now have Sports Tourers, Garnd Tourers, Tourers and, apparently, Hyper Tourers. Every time something new and interesting comes out it get its own category, probably in an attempt to convince people that it's innovative (which it invariably isn't - "Hey, if we stuck an even bigger engine in that, we'd invent a whole new category of bike!" Hello Suzuki B-King), and WTF is a "muscle-cruiser" anyway? Things have gotten so out of hand that when somebody builds a bike that's just a bike that works and does bike stuff, Americans saddle it with the name "Standard".

Go to some manufacturer's websites and you have to second-guess your way through too many menus to find what you're looking for. Moto Guzzi Australia do this well: One page, a little picture of everything, all properly hyperlinked and even with pop-up information giving price. Of course, if Honda tried that you'd still be waiting for all the images to load.

And now I find out that not only is the Yamaha TDM900 the most underrated bike in existence (every journalist to ever test it comes to that conclusion - and then wanders back to their R1), but it's so poorly understood that although technically it's a Dual-Sport (or Adventure, or...) and regularly gets Sports Tourer thrown at it (which is fair enough) at least one website has thrown it in with the Naked's which, clearly, it technically isn't.

I advocate an independent body coming up with a maximum of about ten categories for road bikes, and if manufacturers try and sneak their own fancy names in they get egged at the next trade show.

P.S.: Should customers be worried when Italian motorcycle companies sell own-name-branded battery chargers?

Lies, damned lies, and statistics: Road safety edition

I have a serious problem with most authorities-derived news reports relating to road safety and the road toll.

Well, actually, I have a couple.

Number one is that they only report on the road toll, which is to say the number of people who die. Why do I have a problem with this? Repairing or replacing vehicles can run into many thousands of dollars for uninsured surviving drivers, and drives up insurance premiums for the insured. The ambulance system is already stretched to warping point, and doesn't need the extra work. "Survivors" can end up needing care costing more than $100,000 per year for the rest of their lives (I am not making that up). Emergency rooms and emergency room staff are no better off than the ambulance system and its paramedics.

But never mind that, what's the other problem?

They never report the same numbers, and they never report comparisons.

Take this news story, which appeared in my news feeds this morning:

Drink driving blamed for 105 Queensland road deaths in 2007 (Courier Mail)

It's your basic "People died last year because they were doing stupid things x and y, and Police can't believe that they're still doing them" story, which only serves to reinforce the worrying gullibility, blinkerdness, short-sightedness and ineffectiveness applied to the road safety issue in modern Queensland.

It also contains this paragraph, right at the top in approved inverted-pyramid news writing style:

Queensland Transport research shows drink-driving was to blame for 105 of the 360 lives lost on state roads in 2007 - more than speeding (95), fatigue (65) and inattention (53).

Crunching the numbers, this leads to:

• Drink-driving: 29.2%
• Speeding: 26.4%
• Fatigue: 18.1%
• Inattention: 14.7%

Which, if you ignore petty little things like multiple causes and the difficulty of actually assessing speeding, and the fact that "speeding" includes "inappropriate speed" as assessed by things like skid marks at the scene, not actually "breaking the posted speed limit" which is all the "enforcement"*cough*punishment*couch* activities focus on, is fair enough. Except that the 2003 summary, where they reported non-fatal crashes as well, read like this (fatal crashes first, all crashes in total second):

• Alcohol and other drugs: 38%/11%
• Failure to obey traffic rules: 29%/41%
• Inattention: 26%/29%
• Speed: 16%/5%
• Fatigue: 13%/5%

I'll leave you a few seconds to stare, scratch your head and say "What the fuck?"

We can come up with these conclusions:

• Road rules (they are no longer bothering with this one why?): nearly half of all crashes, which means that running lights and signs, not giving way etc. is, not surprisingly, responsible for a lot of work for panel-beaters, insurance companies and hospital staff;
• If you're pissed, you're not actually more likely to crash, but you're much more likely to die if you do;
• Watch the road!
• Going too fast and being too tired are actually really safe.

Excuse me, what was that last one? Oh yes, speeding contributes to 5% of all crashes, but you're more likely to die if you do. Oh wait, that was in 2003. In 2007, after four years of road safety campaigning and the introduction of more speed cameras, we have no idea about the all-crashes category, but the number of fatal accidents involving speed has gone up by 165%, from 16% to 26.4%. And fatigue-related fatalities have gone up by 139%, from 13% to 18.1%.

But, people not paying attention is less dangerous than it used to be, and I'm sorry, but I just don't buy that at all. You'd have to try very hard indeed to convince me that all collisions are not a result of either a.) misjudging the situation or b.) not being aware of all the variables, and unless your car spontaneously breaks or a sheep falls on you, not being aware means you weren't paying attention. There may have been an innocent party in the equation (possibly), but someone was guilty of either one of those crimes.

But I digress.

Gee, haven't they been successful?

Short of drastically changing how they assess causative factors - and I'm not saying they haven't - the last four/five years have seen a fairly big re-arrangement in the reasons that people crash, and the only road safety advertising I can remember lately has been speeding, fatigue, inattention and a handful of petty rules like seatbelts.

Which says to me that out of three major issues, one has been successful and two have been, shall we say, counter-productive.

More people dying in speeding-related accidents, more people dying in fatigue-related accidents (question: How do they work this one out? I can only assume that it's only possible to collect statistics if the driver survived. Which means that the true figure would be much higher than this, or it's complete garbage because they tick that box if you ran into a tree without tyre marks on the road, and they can't see the spider that dropped out from behind the sunvisor).

Can't they just stop spending money on advertising and direct it somewhere useful, like education, and see if that makes a difference? And how about not putting speed cameras in places where people suddenly realise they're there and oh-shit-brake-and-take-eyes-off-road-to-glance-at-speedo which means that an entire line of traffic is changing speed (dangerous) while distracted from the road (dangerous) simultaneously. This is improvings safety how?

P.S.: When I say "petty rules" I mean things likely to have a major impact upon your change of crashing, not upon your chance of being projected through the windscreen and having your spine compressed beyond it's capacity to support life when you do.

A nascent guide to talking to a vaccination denier.

(last updated 18 September)

Okay, it seems as though my next exercise in science education (while I should be educating myself in Defamation law as it relates to the media), is going to have to be the spurious "link" between vaccines and autism.

I got into a "discussion" at work with a support worker who made just about every error possible, from thinking that caring for a group of people gave him an insight into how they got that way, to thinking that a chronological coincidence between vaccination and onset of autism is evidence. From taking doctorsaredangerous.com seriously to thinking that maintaining an open mind means something.

I felt my brain melting from the committed stupidity.

Unfortunately, although many, many bloggers with, you know, qualifications, have been talking about the issue, there does not seem to be any real definite collation of evidence, and of refutations to arguments, in the theme of How to talk to a climate change skeptic. Although the Autism Wiki looks promising. So, for my own use, I'm going to have to do it. Not spell out the science myself, I don't claim to have that thorough an understanding (at least not yet...) but just a list of resources.

It's going to have to grow quite a bit over the weeks in order to be useful, and short of time to do some deep digging, it'll grow as things appear or are suggested to me.

But, here it is, in roughly alphabetical order:

(P.S.: If the formatting of this is a little off, it's Blogger's fault for being different every time I come back to edit it).

Overviews:

• Medical Journal of Australia
  

Aluminum/Aluminium in vaccines:

• National Network for imunization information (US-centric)
  
• Medical Journal of Australia
• Respectful Insolence
  
• WebMD

Autism

• National Institute of Mental Health
  

Mitochondrial mutations:

• A Photon In The Darkness
• Actual mitochondrial disease, from the Mitochondrial Medicine Society.

The MMR (combined Measles, Mumps, Rubella) connection

• Children's Hospital of Philadelphia
• Respectful Insolence (includes list of other references)
  

The autism "epidemic":

• Respectful Insolence
• Slate
  

Thimerosal (mercury based preservative):

• Children's Hospital of Philadelphia

Vaccines aren't necessary/don't do any good

• Denialism Blog: Who are the denialists? (Part IV)
• Neurologica Blog: First Measles Now Mumps
  

IT Support, oh dear

Look, it's a real, real worry when someone with no formal training and only six months of actual professional experience, working part time (this would be me) appears to know more about providing IT support to an organisation than the actual IT Officer of said organisation does.

The latest wonderful fiasco:

While labouring away inside a 15-page document that had taken me a month of two-day weeks to build, Word 2007 decided to crash without warning. Not even a dialogue box saying "Ooops! This piece of shit is crashing now, there's nothing you can do about it but I thought you'd like to now!" Following which, my file had disappeared off the server. And I mean vanished. There wasn't even one of those usually very annoying temporary files that Word could use to rescue the file. In act, when I reopened Word it didn't even realise that something had gone wrong and that it needed to rescue the file.

So I lodge an IT request, and I get a file back off backups. From three weeks previously. With only 8 pages in it. A polite email later reveals that there were actually no backups between 24 July and 11 August, which happened to be the afternoon of the day in which my 15 pages of research disappeared. Hang on, I thought, aren't backups made every afternoon? Apparently, yes. Actually, apparently not. Was nobody checking this? Was nobody even slightly concerned that a server observed to have been a bit flaky in the past wasn't being backed up reliably?

On the other hand, thank the lord for web browser histories. Seeing no need to blank my history at work, I was able to dive back in and unearth the fantastic resources I had found and which I only had the links to within the original document. So there's something I need to change right there, then, along with making my own backups in future...

This can't be good (Google)...

Since its inception, Google has had a reputation for bare, minimalist but supremely useful user interfaces (born, by the most delicious of irony, from the self-confessed fact that neither Sergei nor Larry knew HTML when they coded up the engine). And, although the word-count of the home page has grown over the years, and although things like Picasa Web and Blogger have been positively florid in comparison, this has remained - it's not generally, although on occasion true, too difficult to work out what your options are and how to get there.

But go and have a look at Google Analytics, at www.google.com/analytics.

It kind of looks like a Google product, but on the other hand it also kind of looks like a hard sell to industry, and the nice and above all familiar box saying "sign in" is replaced by a button which reads "Access Analytics". Excuse me? Access? Isn't this corporate-speak for "I'm a wanker who uses long words when a diminutive one would do"?

It took me a few seconds to work out how to continue, and it left me with an oddly dirty feeling to be using it at all.

Me don't like, and me faintly worried by what they've done.

To be offensively offensive.

Consider, if you will, the two meanings of the word "offensive". Work out the slightly but significantly different pronunciations associated with each meaning.

Isn't it fascinating that two such seemingly disparate meanings can grow out of one word? To offend. To go on the offensive.

One meaning inherently aesthetic, one meaning inherently aggressive.

Does that mean that aggression is, inherently, aesthetically bad?

And does that mean that the current trend, as exemplified by TV ratings schedules, to give the merest hint of sexuality a sterner eye than casual violence, is a modern thing and that the evolution of language has been in the other direction which is, I would suggest, a more hopeful sign?

Discuss.

Desktop search comes to Symbian

I, after a few moments of doubt, decided to give the beta (and therefore free) trial of T9 Nav a go. T9 Nav is made by the same company - Nuance Communications - as T9 predictive text itself, and starts there. It's essentially a desktop search tool, for those of you who are familiar with that notion, and what it does is give you fast access to (theoretically - caveats apply) all content on the phone, from the home screen. It handles the fast part by building an index of all applications, bookmarks, documents, media files etc. This sounds scarily as though it'll bog the phone down and make it run at a crawl, but only takes a few seconds while you watch.

The cool part is how it handles the access part. Simply start typing, using normal T9 rules, and voila! A list of all matching content will be displayed upon the screen. If you are dialling a number, simply ignore the results, finish the number and hit dial. It learns, too - just like a proper implementation of T9, the results you select most often become favoured and are put at the top of the list.

This works so well, it's amazing. I admit to skepticism to begin with, but colour me a convert. The first advantage will be obvious to anybody who has used an S60 phone - it bypasses a messy and faintly arbitrary menu structure and goes straight to your result. But it also taps into an existing (assuming you have it) skillset, using automated and therefore rapid text entry instead of a clunky menu-up-left etc. selection.

For example: To get to MobiReader in my heavily customised menu structure, I could enter Menu-down-select-down-select-select, and that can only be done smoothly through repetition. Or, I could type mob(662)-down-select. Now which one seems easier to you? Finding the unit converter, which I have used a couple of times lately but have no idea where it lives in the menus, would take of a fair few seconds and a bit of head-scratching (is it in the Office folder??), or I could just do conv(2668)-down-select.

It's brilliant.

Okay, it's not perfect. Yes, it is still in beta and it does have issues. Several commentators have mentioned that it doesn't seen to index the music player, but I have noticed that neither does it index Sync, nor index text files, such as this one, and I still have to go through a file browser (yes, I'm compiling this on the phone).

At this point, I have to admit that it's not unique, and that the company responsible for T9 doesn't even seem to have been the first to think of this implementation. If you want a program that isn't still in beta, try the ludicrously named SkyeQuiKey from SkyeStream. These people have obviously never struck upon the idea of product names that are easy to google for. SkyeQuiKey is not free - and I'm expecting that when T9 Nav comes out of beta it'll no longer be free either - and although there is a trial period, it's the amazingly useless 1 day.

I'm sticking with the free T9 Nav until it tries to charge me for it. At that point, I will be hard pressed indeed to not shell out for it.

Link to T9 Nav website
Link to SkyeQuiKey website