Friday, November 27, 2009

Epiphany of purpose

Why manipulate, why
push and pull?
Or play games?

When -

Why the power struggle,
the point scoring
the vindictiveness
for vindication?

When you can -

Why compromise your integrity
or settle for bitchy
jumping on a slipped word
pressing your point
just to win?

When you can love -

Why even try to win
the struggle of me versus you
the do-what-I-want

When you can love
and achieve far more
than mere winning.

Monday, November 23, 2009

Mums and content relevance

Responding to Grant McCracken's Invading The World Of The Mommy Blogger:

I recall Clay Shirky mentioning that mums are the ultimate test of your website or service, because they have no time for things that are not ultimately useful.

The same reason underlies usability studies finding people unwilling to waver from their "click, hunt for relevance, click, hunt, back" behaviour. People (and particularly mums) are very goal-oriented, and until you provide something that helps them, they don't care about you (thanks Kathy Sierra).

So, you need to pull your target audience in with the application of your book/theory - what could a mum change about her day-to-day routine that would show the theory at work? Mums will read that, looking for a nugget that will help them. If they agree and it works, they will come back for the rest of it. But calling something "Chief Culture Officer" does not suggest a resource that will get better-mannered children and free up some of their time to see their friends.

In essence: identify your audience, find out what their problems are, work out what one theory application will have the most impact for them, and package that up for them. People will come back to the source of good advice.

Tuesday, November 3, 2009

Does the hill level out? [The difficulty of discipline]

Every day, every moment,
I face a choice.

Step up the hill
against the gradient,
Eastward, towards
activity, discipline,
doing the right thing.
This way lies achievement
I'm told.

Or relax,
Let fall the strive
and roll
gleefully, unfettered,
to the little ditch.
But here only frustration
I've seen.

keep I upward.
Push I, by force,
But I, striving
so I waver.

Does the hill level out?
Can I trust that the toil
will one day yield
at least a shallow slope
or greater muscle to overcome it?

Were I truly lazy,
I would have no internal motivation.
Nothing would push from inside
and in that ditch I would be
But in it I am not.
Not happy, that is,
but frequently there I find me,
entertained by an endless stream
of little amusements. But inside
my spirit darkens with self-hate.

So in a spirit of reluctant agreement
I make another pact
with the inner striver
and turn to face the hill.

And then some days I wonder
is this the wrong hill for me to climb?
Or would I better push
knowing that that rock hides
a treasure easily
worth the climb?

Perhaps the ditch merely reflects
my fear of the unknown.
Beyond the rock, over the ridge,
perhaps I fear I will not make it.

Save me, self-deception,
from facing it all.
Your ditch I know.

Monday, October 5, 2009

Fallacy of the wise

How deep the fallacy of the wise,
of mine?

How far goes the well,
how deep the hole
in the ignorance of humanity?

I stumble,
so walking always,
yet each step feels like
a leap, a feat,
an achievement.

Most wisdom is merely
recognition of our limits -
humanity's fallacy, its
myopic self interest.

So teach me, universe,
teach me, nature.
Limits be my guide,
exponentials my pointer
to the unending asymptotes of
my blind, dumb stumble.

Null my ego,
flatten my puffing chest,
drive it down and make me


Knowing even this,
I am proud
of my humility.
Fie, fool! Fallacy

Friday, September 18, 2009

Seeking Significance

Underlying always,
Beneath my drive,
I am seeking significance.

Somewhere to the right
in the caverns of my mind
I am drawn
to be something more.

Secretly I measure myself
to Einstein, to Feynman.
Men of great achievement
of significance.

I do not want to live
nor tirelessly work

It is said that young men seek
significance, and find only age
which brings a dulling
of the hunger and a
wider emotionality.

But this is my hunger. Will I be
a great man? Or will my dreams be forever
just dreams? And I
just a man,
just anybody.

Tuesday, August 25, 2009

Fat Photographer

> I'm at a seminar in Eindhoven. The photographer is you when you were fat.

I lurk
neath your conscious,
waiting to subvert
your unwitting context.

Fat, I live forever.

Sunday, August 16, 2009

One more standard page widget: short URL to permalink

Most pieces of content on the net from flickr photos to blog articles have a row of widgets somewhere near the item, inviting the reader to Digg it, Reddit it, Slashdot it, email it to their friend or make it their "it's complicated" on Facebook.

So, here was my suggestion: If you have a favoured URL shortener, or you have your own, provide one more link at the bottom called ShortURL.


Well, it turned out this thing is already baked in the 'Tweet This' link, coming soon to a blog near you. So we see another item added to the growing list of these widgets. I quite like the way a few people modify the 'standard' content interface on their blog and the change becomes viral.

Generalise: personalised widget lists

This widget list is growing in length, and will suffer scaling problems: I am a user of slashdot and Twitter, but not Reddit. And I don't email my friends with links. So I'd prefer to have a custom list of items. Perhaps, then, we could add a tag or special container to the blog so that a browser addon could replace it with my preferred set, using the links provided. If it could pick up this list from a website that stores my preferences, so much the better.

Semantic web

This brings us closer to the semantic web, where the stuff on the page is interpreted rather than simply displayed. Incidentally, this is where browsers can deviate from being PDF readers - they become more than merely a delivery mechanism.

The content interface

I can see a normal interface being slowly generated for content on the web, with each item having its own permalink and registered short URL, and some associated way of commenting on it. I expect it will also be possible soon to get a summary of its ranking or quality measure on various sites like Digg or Facebook. All this metadata, despite being available with the content via the browser, is actually maintained in many silos, and aggregated repeatedly. This is a sparse object, a distributed dataset.

At last, integration of platforms and services is possible at the point of consumption, rather than needing repeated changes to the website by the producer.

Saturday, August 15, 2009

How could Labour win the next general election?

A disclaimer: this is an issue I don't have real depth of knowledge on. I'd love to see Nick Robinson, Ian Hislop, Johnathan Dimbleby, Martha Kearney and a non-BBC commentator as a panel facing a public audience to thrash it out, absent any party lines and attention-seeking from politicians.

What could Labour do to win it, then? The public reviles politicians, particularly Labour, and particularly the incumbent leader. Even impressionables like me have finally got tired of Brown and the various expensive schemes like ID cards that have proven foolish.

Reading Ian Hislop's page on wikipedia, he has apparently expressed support for Vince Cable as potential treasurer, which made me wonder. What if the Lib Dems joined up with Labour? They could give Cable the treasurer spot, put in a smoother operator like David Milliband to run the show and set a fixed date 6 months away to have an election. This should give the country time to get used to the idea, push through some high-profile Lib Dem policies (needed to demonstrate that it wasn't just a gimmick) and get some battles fought against the Conservatives in the Commons. They would do well to field Cable nearly as often as Milliband in the Commons if the PMQs rules allow it, because although Cameron might rate higher in public minds than Milliband, Cable must thoroughly trounce Osborne.

They would need, inevitably, to keep a few less well-liked faces around - Mandelson for his astute manoeuvring, perhaps a real surprise like Blair as foreign secretary for his experience and respect beyond the borders.

Policy-wise, having the Lib Dems on board would be a fantastic excuse for rewriting the policy book - they could drop ID cards, 42 days, clarify Iraq and Afghanistan, commit to cutting spending (which was always inevitable, but now acceptable), and so on.

It would be an utterly catastrophic result for the Tories, whose position is currently so defined by New Labour that they would struggle to find a footing, perhaps long enough for some memorable moments like Michael Howard's Paxman interview. Taking a solid devolutionary line would pacify the Old Labour North with the prospect of more local rule, and the lower taxes would be a step squarely into the Tory sweet spot.

The only Conservative defense would be that they are merely rearranging the deckchairs on the Titanic, or the derision of having to resort to joining up with the Lib Dems. But the flaw here is that these are political arguments. The public cares about policy arguments at the moment - what will you do to the NHS? What will you do for the economy? How will I know you are a good politician, not a typical, dodgy one?

But they wouldn't do it, would they? Either they'd rather lose this election and pick up after a Tory term, or they are too shortsighted to try something really radical.

Thursday, August 6, 2009

Wednesday, August 5, 2009

Could Markov chains be useful cryptanalysis?

The Enigma code-breaking relied on the discovery of common pairs of letters. Markov chains are a way of analysing common sequences of symbols. Could Markov chains be useful as a cryptanalysis tool?

Tuesday, August 4, 2009

Apple, tethering and ARPU

Girtby blogged on Apple allowing carriers to detect and block tethering, affronted that Apple would act so unfairly to its users, because the carriers are all 'Apple's bitches', right?

Unfortunately the carriers are not all Apple's bitches, as you put it. As per the normal business negotiation, each side brings their cards to the table. Apple had the iPhone, for which it wanted moolah. The carriers had moolah, for which they wanted the iPhone and as many restrictions they could get. For each restriction they will be willing to offer more money, so if Apple want to be nice to a small group of users that want to use tethering they have to be willing to give up some revenue for that.

In creating restrictions, carriers push a larger proportion of users into buying addons to their contracts or into using the expensive bit of the tariff (international calls by banning net phone apps, etc), so driving up their Average Revenue Per User. Remember, the better the iPhone, the more Apple can squeeze out of the carrier and the more the carrier has to raise the ARPU to get their profit margin back. So Apple leverage the hype they can build to get a better deal from the carrier. And the carrier know that if you're spending $40/m on a contract, you probably won't mind another $5 for tethering.

Yes, they are all there to take your money. That's Big Business at work. The terms of the deal to prevent tethering, to quietly throttle certain apps, were all written into the contracts at the negotiating table. This was a done deal a long time ago. As time goes on, and more apps and uses are prevented, the terms of that deal will be slowly revealed, clause by clause. Apple has not just turned on its loving fans, it has already sold them. Did you expect anything else?

Thursday, July 30, 2009


In everything there is an exchange of some problems for others, some freedom for others, some limitations for others.

So as you wait for the new thing - a record contract, a home of your own, a wonder-job - see afresh the freedoms you will be giving up, the joys you will lose to memory. By all means, grasp the new thing if you wish, but expect to grieve the old as you celebrate the new.

With new birth comes the death of the old, and the inevitability of more death. We invite it with every breath. But don't be depressed by it, be inspired! Your death will invite new life.

Life is minor and major, stop wishing for utopia and live it.

Wednesday, July 8, 2009

[Idea] Fie to scammers: A public URL database

I've never understood why URL shortening has not become a part of the browser... surely it's easy enough to shorten a URL by compressing it, so the browser could see and decompress it without a single http request.

Still, the world loves ease of use and hence we get the inevitable abuse of URL shortening to send people to spam/scam/phising sites.

So why not have a public URL database, where a people enter the shortened URL and the database either gives them the URL or tells them it is spam/scam etc? Crowd-source it - let people use browser extensions to report bad links.

Importantly, to avoid it just being yet another redirector, the service would need to provide public database dumps so that others could provide competing services, verify the data, etc. Vitally, ISPs could provide access or even use the list to warn the users about the content. Think of it a bit like the freedb or the imdb - public service databases.

It would allow the browser to look up every link on a page to find out where it ultimately points, and take the URL shortening services out of the loop entirely. It could even communicate via the minimal overheads of http error codes - 404 if the link is not known, 307 to provide the correct redirect (skipping advert pages, further pages of redirection), 303 to indicate a dodgy site and supply the URL of a relavant page (e.g. the Dodgy Scammy Warning page).

Come on people, one link gets read thousands of times, so if a small proportion are willing to mark dodgy links, that could save a lot of people a lot of headaches. And we could always use it to tag links, and vive la web semantique.

Thursday, July 2, 2009

Take me back

Take me back
to simple.

Take me
and return
to humble and quiet faith.

To quiet and humble
sweet and waiting
patience and hoping

Let my arrogance and force,
my pride and envy
drop from my shoulders
ease my burden.

Give me
gentle joy of being.
No more
fearful waiting.

Soft white curtains billow
in the sanctuary of my mind. Sought
and seeking
to escape frought
and fleeting.

Take me back

Thursday, June 25, 2009

Some suggestions for

I'm a user and subscriber (I pay), because I think it's a great service. The 2 great things about it are that I can just log in and listen for hours, without the hassle of files, and that I can discover new music that I would never have encountered otherwise. Having used it for quite a while, I have a few suggestions:

  1. Bring back pause There used to be a pause, but it was taken away. Now when someone interrupts me at work, or the phone rings, I have to either leave the radio on, or give up listening to the track at all. Some tracks can only be listened to every so often, and that can get quite annoying.

  2. Merge the players One mini player (that used to be the only player) sits on artists' pages and plays you single tracks, and the new megaplayer plays radio, shows you slideshows, etc. I see no advantage in keeping the old mini one.

  3. Make 'loved' and 'library' tags I'd like to listen to tracks in my library, or loved tracks list which are also upbeat, or pop, or rock, because I have eclectic tastes. So let me have special tags to use in the multi-tag radio.

  4. Fix playlists Perhaps I'm too stupid to use them, but I haven't worked out how to use playlists easily. They must have 45 playable tracks by 15 different artists, the playlists pane on the megaplayer doesn't do anything at all for me, I can't see how to make a playlist or add tracks as I go.

  5. Help, I got distracted I listen at work, and frequently people interrupt me and I leave my desk for a time. Unfortunately, if I leave playing its recommendations, I get some mediocre stuff that I don't like cluttering up my library, and confusing last's recommendation engine. It's now at a point that I don't often listen to my recommendations or my library, because so much has been scrobbled that I don't want to listen to. Had I had the ability to remove the last hour of scrobbles way back when, I would have a reasonable library at the moment. How can I get rid of the noise?

  6. Learn from musicovery The key to musicovery is the ability to click on a combination of light/dark and slow/fast to get a playlist that fits your mood. It doesn't know me like last does, but sometimes it's better because I just want fast music to work to, or calmer music to chill to. If I just type in some tags I'll get loads of crap tunes that I have to keep skipping past, distracting me from my work. The music is there to evade distraction!

  7. Let me compare my tastes with others Perhaps stepping into the Facebookesque world of 3rd-party apps, a way to compare my tastes with my boss's or my colleague's would be a huge laugh. I could be embarrassed by my pop tunes, they could be embarrassed by their 60s flower power, it would be great.

  8. Simplify the radio We've all used Google, we can all type several tags into a box, so why do I have to choose Multi-Tag radio or Multi-Artist radio? Why can't I mix them up?

  9. Provide playlists from chartsThe charts are great, but I can't just click 'play chart' and have it start at the top. It's a list, I want to play it, why not a playlist?

  10. Let me queue Not my irrational British tendency; I want to queue tracks up to be put in my radio playlist. I don't want to stop what I'm listening to and play this other track, I just want to listen to it soon without interrupting things. I want to pick out some tracks and queue them up so that I can go away and work without having to click every minute or two. At present, Last is quite demanding of my attention.

  11. Work on inferring If I have queued up a few songs to listen to, or created a playlist with a few songs in, or typed in mutually incompatible tags (rock, trip-hop and strings, say), then Last should be able to guess at some more tracks - think Google Sets. Maybe there aren't any more tracks tagged with all those tags, so start calculating which tracks are most like them, or have 2 of 3, etc. Make it like search.

Broadly, my problems stem from having no easy way to manipulate the data Last keeps. I can't edit easily, I can't recombine lists and lists to make other lists, I can't search the database and shove the results into a playlist, I can't compute statistics or similarities or group similar tracks in lists. I can't even see the next 20 tracks and prune out what I don't fancy listening to.

Come on, it's pretty good, but it could so easily be great!

Wednesday, June 24, 2009

Ninthers, Tukey and statistical approximation

A response to a post on The Endeavour about Ninthers and Tukey:

There are 2 useful things in the ninther concept: approximations to bulk statistics and the calculation of a median without any sorting.

Generalising from 3 groups of 3 to n groups of m, we could still calculate a median from a series of chunks of the dataset, but we would need to sort.

This would suggest problems when working on Very Large Datasets, but consider the case of the most annoying dataset - randomised data. If we take a chunk of data from this dataset, we can approximate the statistics of the bulk with the statistics of the chunk, or a series of chunks.

The answer may be, then, to randomly pick out a series of 9-value chunks, and calculate a series of ninthers. That way the number of comparisons per total values can be less than 1.

O(1) to O(N), depending on how accurate you require your statistics to be.

Monday, June 22, 2009

Changing the pattern

"Insanity: doing the same thing over and over again and expecting different results."

- Albert Einstein

I first heard the above quote as one about neurosis instead of insanity. The effect is the same: do you repeat a pattern of behaviour and thought, and expect a different result to last time?

I know I do. Every project, I'm expecting that this time I will magically be willing to follow through, to see it to the end. The initial buzz always makes me believe it. Or I will make a resolution and expect to keep it.

I have noticed that the real times that make the difference are not those promising, idealistic times at the beginning of the project when everything is blue skies and hopefulness. The real crux points are when I am bored, tired and resistant to the work, repelled by it, rebelling against the cords that bind.

Those are the times when things need to change. I listen to myself there and believe that I don't need to work on it now, that I can leave it to next time. Fool.

Thursday, June 4, 2009

One day

One day perhaps when I am 80, I will be able to dispense the wisdom of those who have gone before sufficiently untainted by my limited understanding to improve people's lives.


My hands will be clean enough to pass someone the bread they need.

I know that by then, I will be able to see clearly the foolishness I am currently unaware of - what idiocy do I labour under today?

Beginning wisdom

If a man begins with certainties, he shall end in doubts;

But if he will be content to begin with doubts,

He shall end in certainties.

[Francis Bacon 1561-1626]

I began with certainty, and find myself in the middle, with doubt.
Seeing the world afresh.

Thursday, May 14, 2009

[Poet] And go gets it

And go gets it
'Cept I don't.
I never do.

I begin
or pre-begin
or think
and muse.

But I do not do.

At least, I didn't do.

I always tried at first.
Always pushed and cajoled
hoped and cried
that I would finish.

But all dissolved to boredom.
To eeking and tweaking,
and hunting and fixing mistakes.
To dull dull dull,
the monolith grew.

Then I gained sophist
and saw the rolling story played
at school, at uni, at work,
of the repeated descent;
from lofty thinking mountain
to miserable dank undergrowth,
vast swallowing of the mountain hope.

But I am not repressed. I,
my creativity,
rebounds each time.
A quake, volcano rises,
dank foothills thrown to sky.


My wife suggests self-loathing, the
source from which the lava flows.
A reaction to familiarity,
and familiarity, self.

Do I greyly, merely
chase the new,
mesmerised by greener grass?
A fakery,
chasing foreign winds?

An inner, bitter remnant
suggests I quit.
Give up! And slide to mediocrity.
Sod the world, and hate it

I regard it as a clam regards the grit,
I will not yield.
If indeed, that is the choice.

Dare I dream?
Even if the dreams are worthless
in others' eyes?

Saturday, January 24, 2009

Synthetic synaesthesia, Part II: font-induced colours

Reading an old post that I didn't get around to publishing before, a new droplet has formed where the last one dried.

The gist of the first idea was to actively colour the letters on a page, to employ the brain's colour-recognition to improve spelling and reading ability. The flaw I found in reading further was that dyslexia and synaesthesia (seeing certain letters as coloured) are related.

In particular, this quote stuck out:

"My daughter, aged 11 is dyslexic and has grapheme-colour synesthesia. The colors negatively affect her ability to read and to spell, since some of the letters have the same color. E and U are both green, for example. She also tends to group the colors and hence inserts letters into words because she thinks their colors “go together”." - Mary G.

The Idea
Mary G's daughter seems to see spelling as a series of colours, and can thus happily swap E and U. However, there are likely to be other aspects of shape that affect the colour. For example, what colour is a spikier E, or a U with a wiggle in it? Intuitively, the changed shape would affect the letter's perceived colour, which might help Mary G's daughter. If an entire font was created based on a fairly standard font, in which certain letters had particular changes made, that girl and others like her might find reading easier, and might make recognising incorrect spellings easier too.

Inevitably, as this is based on intuition, I'm probably wrong. If the girl perceives all Es and all Us as the same colour, regardless of font or handwritten, this won't help.

One alternative might be to attach simple musical notes to the different letters, making sure they harmonise, but that problem groups like E and U do not sound similar. As she types, the flow of notes will be recognisably right or wrong for a given spelling.

Admittedly, the girl could just rely on Word to correct her spelling all her life. She will also need to learn to spell when writing, and paper can't do customised sound effects.

But she could do a spelling bee with a little humming.

An email to Kyle Neath on IP and Capital

Reading That's My Property on Kyle's Warpspire blog, I saw some parallels with recent ponderings on capital versus cash flow.


Firstly, props for the blog, quite inspiring to see someone out there and getting on with things. Maybe I just like it because you're on my level. (I read your blog because you remind me of me - what a sentiment.)

Just been reading your thoughts on IP, and thought I'd draw some parallels with capital investment. I've been reading and thinking about capital, interest rates and salary, and it seems that the best thing is to try to turn a chunk of capital into an income, in a better way than just spending it in small chunks (mirroring your point on VC-hunting startups these days).

Suppose I have an idea. Not just any idea, but one of those 'this could change *everything* [for people like me trying to do what I'm trying to do right now]' ideas. The ones people get really precious about and fail to make use of for fear of exploitation. To me, that's like a chunk of money. Now, you can either use it as capital, and try to wring future returns out of it by renting it to other people (licensing), for which you need lawyers, or you can build your own company on it. Either way, it was only ever one idea, one chunk of money. It is limited in how much it can change the world, or get you early retirement.

Capital is limited. Much better is a flow of income. Saving £200 a month is better than investing £20,000, after four or five years. In a similar way, I treat ideas like a flow, not capital. I have some great ideas I'm wild about, but if I haven't done anything with them myself after a while, I will give them away. More ideas will come, formed and shaped by the experiences I have.

Ideas are always over-valued. Google was founded on one idea, sure, but they built on that initial investment with many, many others, and a lot of hard work.

I'm hoping to build a portfolio of the ideas I have, not to boast or to sell, but to convince someone, someday, that these things do not stop falling out of my head, and that it's worth paying me for a time so that I focus on *their* problems. Until then, I'm working on refining the filters, stimuli and breadth of search space available to my imagination so that what does fall out of it always improves in quality and usefulness.

I am, however, always wrong to some degree. Where am I wrong here?


Friday, January 16, 2009

Duck-typing vs static with interfaces

Reading Jason Baker's thoughts on the issue, here is my response:

Personally, I've been pondering duck typing vs interfaces for a while. It seems that the choice is between a catch-all (duck typing) and a square hole (interfaces). Interfaces are great for static languages, but are more restrictive at design time than going for duck typing. At the same time, duck typing means that an increasing number of tests are needed to make sure the system copes with different usage.

But what if we restrict ourselves to web services? A datastream (MIME, xml, etc) is passed in, and the service provider has an SLA contract for expecting it to 'work'. But since we're dealing with externally supplied input, we do all our input-checking at the start, to make sure it's all sanitised. After that, we have effectively already done the checking we need (if we've done it well).

Now, do you take the static+interface approach, because you already know what the input is like, or do you take the duck-typing approach, because we've already sanitised the input, and have our TDD tests aligned with our input sanitisation?

My answer would be to go with duck-typing here. If we've defined our SLA well, we have excluded non-compliant input data, which will be turned down at the point of reception into the system. Thus we can dispense with the overheads of interfaces, since we can (Atwood style) throw hardware at the interpreted language speed issue.