Thursday, October 31, 2013

Next time, SMH, get the surfboard to write the column



Really, it's not often that anyone – other than News Limited columnists talking climate change – can manage to cram so much utter drivel into the space of one column as this.

“I want to think we're too smart to pour 40, 50, 60 billion dollars, whatever the final cost, into a fibre-optic network system that sounds impressive now but may look like the greatest white elephant since the attempt to turn the stinking (literally!) Salton Sea near Palm Springs into a rich thang's playground; that a dozen years down the track the NBN will be the equivalent of "visionaries" 150 years ago building a network of hundreds of thousands of stables across the country to cope with the growth in horse transport.”

I think this individual, the founder of a surf magazine yclept Derek Rielly, is a subscriber to the “fibre will soon be obsolete” school of what passes for “thinking” (even when we used entangled photons to overcome the speed of light, fibre will be a better way to move them around than through the air. Noise, and all that.).

“According to the projections of the NBN Co we'll be getting a gigabyte per second once it comes into our lives.”

Well, there's a scale thing right there. The NBN is measured in bits (and megabits, and gigabits), not “a gigabyte per second”. And it's not a projection, the kit is already gigabit-capable wherever the network is, but you know, this is a column. It's like climate change: facts are optional when you have a dumb opinion to tout.

“Wouldn't it make more sense to just invest 20 billion dollars into developing the world's finest compression software?”

Well, no. Actually, no with a double serve of clue-stick, multiplied by “what on Earth?”

“Something like America's Manhattan Project in World War II that created a way to annihilate the world in just six years.”

Yeah, thanks for that analogy. But let's continue …

“All of us are familiar with today's compression software such as jpegs and PDFs which work little miracles every second of every day.”

At which point, head meets desk, because this sage advice, presumably passed by an editor between lunch and Alka-Seltzer, comes from someone who knows not the difference between “software” and “file format”.

Oh, and (say) a PDF is already compressed, which is why (say) PKZip ignores it.

Save me.

“Can you imagine the value in creating compression software so good it shoves the paradigm of computing way into left field? We'll be emailing movies, high-res photo albums, complete TV commercials, magazines, all over our existing networks and through our existing software.”

OK, that's enough quotation.

Compression is mathematics, and it's very well understood. It's a problem the world has been working on for nigh on sixty years, in the digital world.

Of course, even the ancients practiced a form of compression: setting a fire on a signal tower to signify invasion, for example, is a very compressed message in a specific context.

But as for modern communications: compression starts when Claude Shannon explained the limits of any communication medium: channel minus noise equals bandwidth. For any medium, you can precisely predict how much information it can carry.

So compression tries to reduce the amount of information that has to be carried – it says “what can I get into 'bandwidth'?” And it uses a lot of maths, and lots of billions have gone into it over three decades, and the world has created very good compression.

And all of the decades of compression work have been a matter of increments. Every now and again, someone claims a bigger breakthrough than anyone ever before.

Like the Adams Platform.

Next time, get the surfboard to write the column.

No comments: