Google autocomplete’s 15-step infinite loop starts and ends with love

While walking up the high street this morning, I suddenly wondered whether there was an infinite loop triggered by following autocomplete results in Google. The methodology would be thus.

  1. Type <Google> into Google and take a note of the first autocomplete suggestion.
  2. Type that extension into Google and see what that term’s first autocomplete suggestion was.
  3. Repeat step 2. until the loop was closed. I.e. the same term occurred twice.

If you type <Google> into Google, its first autocomplete suggestion is for <Google Maps>. Search for <maps> and you’re prompted with <Maps quest>. And so on. Below is the full set of results.

And here we loop back to step six.

(For the record, this post was autocompletely useless.)

GCSE scoring: #fail

The school examination system is fundamentally broken: the way in which marks map to grades is wrong.

This year, the percentage of GCSE passes increased for the 23rd year running.  While the percentage of A*–C grades increased by two percentage points from last year—to 69.1%.  Meanwhile, the A-level pass rate increased for the 28th successive year.  (They last  decreased on the same date as compact discs were released in Germany!)  And 29.9% of those sitting the exam achieved an A* in Mathematics (Further).

In my view, the percentage of entrants in a subject achieving each grade should stay static year-on-year.  So, for example, 16% of Maths entrants should be awarded an A grade.  Period.

Only through evolution are we becoming more intelligent as a species.  And the rate at which evolution, er, evolves is sufficiently small for it not to affect exam results across a generation or so.  As such, a gradual increase in people’s success in exams merely serves to penalise the old, at the same time making it more and more difficult for prospective employers, universities and the like to differentiate between pupils.

If a paper is easier than last year’s then by maintaining proportions as proposed above, the mark necessary to achieve a certain grade is pushed higher.  I’d be happy for the percentages to change from one subject to the next—one may choose for a higher proportion of English exams to score well.  But changing them year-on-year serves no purpose whatsoever.

Seems simple to me.  So why so hard for the politicians?

Everybody move

If everyone inhabited a strip of the earth 15 degrees wide stretching from the Arctic to Antarctica (I’d suggest Europe/Africa to maximise land; N and S America are deceptively not on top of one another latitudinally), then we’d all be in the same time zone, and no other time zones would be of interest.

Here’s the maths to support the argument:

It’s quite a big project to make this happen, but I think I’ve presented enough of a business case for someone to start running with this. High-level plan anyone?

The River Twitter

My good friend Alan wrote:

Twitter, for me, is a river.  Every so often you step in and watch the flotsam drift by.  Sometimes there is stuff that you pick up and look at, rarely you keep it to examine later.  But what came before and came after is lost to the ocean of debris.

I love the analogy.  Your Twitter feed is like a river.  Whether or not you’re present to witness it, it keeps on flowing.  And for anyone following more than a hundred or so people, keeping up with every snippet every day is a pointless goal.

Instead, just as with a river, you pop along every so often—to admire the view, to watch the boats pass, to see how the sun is glistening on the water.  You might even take off your shoes and have a paddle.  But however often you go back, the river will always be flowing, there’ll always be something new to take in, and you’ll always be welcome to get your feet wet, skim some stones or dive right in.  That’s what I love about Twitter—and rivers.

Changing space/time for fire engines

The world is an inefficient place.  If fire stations are positioned such that all places can be reached within a specified length of time, then their circular catchment areas necessarily overlap, assuming that distance from the fire station is directly proportional to time taken to get there.

It would be much better if space/time was warped such that fire engines could reach every point around the perimeter of a square within a specified length of time.  That would make for a much more efficient fire service.

155.6 swimming pools of oil

I love data.  And I love it when people take abstract concepts and try to put them in terms that people can relate to.  The most recent example has been the oil spill.  It’s nigh-on impossible to visualise the quantity of oil that has leaked from the Deepwater Horizon well since 22 April.  So many have tried to give estimates that we can relate to.  All of the quotes below are from different sources, found courtesy of Alan.

[The] flow rate […] would fill more than 50 swimming pools per day.

The Coast Guard [said] earlier it was at least 1.6 million gallons [since the spill started]—equivalent to about 2½ Olympic-sized swimming pools.

BP’s estimated 5,000 barrels a day […] would be approximately 150,000 barrels (or 6,300,000 gallons). That’s barely enough to fill 286 swimming pools: sixteen feet, by thirty-two feet, by eight and a half feet deep.

Once the pipe has been cut, the oil will spew into the Gulf of Mexico unobstructed, enough to fill an average swimming pool every hour

In just one day, the oil leaked would be enough to fill up six Olympic-sized swimming pools .

The only consistency here is the unit of measurement.  (It seems that the SI unit for oil is the Olympic-size swimming pool.)  The quantities are either formed of very varied guesswork or extremely poor arithmetic.  Probably a combination of the two.

My own estimate, based on an average between a low and a high estimate of the volume, was that as of 3 June, the oil would be sufficient to fill 155 Olympic-size swimming pools, with enough left over to start filling the 156th pool.  But this meant nothing to me.  So I changed the frame of reference.

If the oil was solid and laid evenly over the entirety of Hyde Park, it would form a paste 40 metres deep, equivalent to nine London buses stacked vertically.  Sounds a lot, doesn’t it?

Yet if the oil was to fall as snow over the entirety of the UK, it would amount to 0.2 seconds of snowfall.  Nothing at all, right?

While units that we are more familiar with help us visualise things that are intangible to us, they also serve to create dramatic headlines, headlines that could equally be replaced with innocuous ones if the newspapers were driven by ambivalence.

And who’d want to swim in oil anyway?

Excel is like life

Life is complex.  Problems are complex.  And it’s rare that an intricate problem can be solved by a simple solution.  Or indeed that a single solution is the only one.  Excel is a good analogy here.

It has rows and columns and it was invented to organise data.

Yet in its 2007 version, it comes with eight standard ribbons.  (Further ribbons present themselves in specific situations.)  The Home ribbon alone has 42 separate items within it.  Twenty of these have dropdowns from which further options can be selected.  A very conservative estimate would be that an average of five sub-options are available for each of these.  If the other seven ribbons are similar, then that’s 976 options, and that doesn’t account for the plethora of formulae that can be written in each of the cells, and the canvas of colours available.  (As a slight aside, that would allow 16 trillion possible actions across the cells of a single worksheet.)

Admittedly, Excel was not designed to solve a single problem.  But the arsenal of tools available merely highlights the huge variety of ways in which problems can be addressed.  If you give ten people the same problem and ask them to solve it in Excel, each one will address it in a different way, sometimes subtly different, sometimes wildly different.  If you don’t tell them which tool to use, then your range of solutions will widen further.

No less than and no greater than

You often hear of the likes of Osama Bin Laden (actually, just Bin Laden I think) quoting certain events or holding a copy of a newspaper during video footage to prove unequivocally that the footage was taken after a certain event and therefore after the time of that event.  But I often wonder whether anything could be used, likely something that we haven’t yet discovered, to accurately time-stamp a video.  That is, confirm the exact timing of its filming, as opposed to giving a “no earlier than” date.

Because that would be kind of cool.

The afterlife is just too hard

The afterlife should be easier.

I don’t have a will.  Don’t get me wrong—I know I should have one.  But the barrier to entry is too significant.  I have to involve a solicitor, get counter-signatories, spend money and generally make quite a bit of effort to get one.  It’s something I’ll probably do when I’m 40 when death becomes a more real concept.  Let’s hope I make it that long.

And besides, the default position is sufficiently in keeping with what I would want that the benefit is not great enough to warrant the associated effort.

But wills should be easier.  I would like a good level of dynamism in defining my benefactors and the respective rewards that they would reap in the event of my death, timely or otherwise.  Particularly the smaller amounts and artefacts.  I’d like to be able to define how any little knick-knacks that I happen upon over time are to be distributed among my friends and family on my demise.  And every time I get a new knick-knack, I shouldn’t have to bear legal fees.  And I would like to be able to pledge relatively small amounts to my friends to enjoy a meal on me, maybe on the evening after my funeral.

I’d like to be able to manage everything up to a certain value through a web interface.  And I’d happily cross-reference said interface on a proper will.  (“Any items specifically mentioned on account name danosirra on act as a supplement to the details herein.”)  And I’d happily agree that any changes made to said inventory after a time-stamp at which I was medically declared doolally should be deemed invalid.

Can this be done?  Does it already exist?


Paul Clarke introduced me today to the pastime of Tubewhacking.  Similar to Googlewhacking, it involves finding an English word none of whose letters appears in the name of a Tube station, and for that station to be unique in that quality.  The most famous example I know is St. John’s Wood, none of the letters of the word mackerel appearing in its name, a claim that no other station can make.

I wondered whether any stations were themselves Tubewhacks of other stations.  So I got to work.

Fortunately, the number of columns in Excel has increased recently—I needed 7,616 columns to complete my logic, along with a tidy 11.8Mb.  And below is a summary of the results.

There are 59 stations that have Tubewhacks, although their Tubewhacks come from only 22 unique stations.  Bank is the most common, being the Tubewhack of a whopping (not Wapping) nine stations. Each of Vauxhall and Woodford accounts for seven Tubewhacks.

Below is the full list—station on the left, Tubewhack on the right.

Barbican: Temple
Becontree: Vauxhall
Bermondsey: Vauxhall
Bond Street: Vauxhall
Boston Manor: Chigwell
Brent Cross: Vauxhall
Burnt Oak: Chigwell
Camden Town: Ruislip
Canary Wharf: Temple
Canning Town: Shepherd’s Bush
Charing Cross: Temple
Chorleywood: Bank
Colliers Wood: Bank
Dagenham Heathway: Ruislip
East Putney: Woodford
Eastcote: Kilburn
Edgware: Pimlico
Elephant & Castle: Woodford
Elm Park: St. John’s Wood
Farringdon: Temple
Fulham Broadway: Epping
Gants Hill: Woodford
Goldhawk Road: Upney
Green Park: Dollis Hill
Hainault: Woodford
Hampstead: Kilburn
Highgate: Woodford
Holloway Road: Epping
Ickenham: Woodford
Kew Gardens: Pimlico
Knightsbridge: Oval
Leyton: Chiswick Park
Limehouse: Bank
Liverpool Street: Bank
Mill Hill East: Woodford
Mornington Crescent: Vauxhall
Newbury Park: Dollis Hill
Perivale: St. John’s Wood
Piccadilly Circus: Kenton
Plaistow: Debden
Putney Bridge: Oval
Queensway: Pimlico
Richmond: St. Pauls
Rotherhithe: Bank
Royal Victoria: Debden
Shadwell: Brixton
Shoreditch: Bank
Southfields: Bank
St. James’s Park: Hillingdon
Tooting Bec: Vauxhall
Tower Hill: Bank
Upminster Bridge: Oval
Upton Park: Chigwell
Warren Street: Pimlico
West Brompton: Vauxhall
West Finchley: Moor Park
West Ham: Kilburn
West Hampstead: Kilburn
West Ruislip: Bank

Liverpool Street and Bank form the only pairing in the above list that are one stop away from one another.

← Previous PageNext Page →