Weblog

The video for Sleepless by Tropics, directed by my brother,...

Thu, 15 Mar 2012 11:21:27 +0000

The video for Sleepless by Tropics, directed by my brother, Ewen, assisted by Nick Macdonald, Rik Green, Danny Atkinson and myself, starring Daisy Keeping. Made for no money, shot in Epping Forest (on a freezing Saturday morning) and in my spare room (on an extremely long Sunday).

Using Mailgun on Heroku to create user mailboxes for *vidbox

Thu, 08 Mar 2012 14:07:22 +0000

To distract me from my hangover the other weekend I made a small thing to scratch an itch I have with internet video… It’s called *vidbox and you can read about it in my original announcement.

Here’s the crux of the problem:

When I’m out and about I check Twitter, Facebook, email etc on my iPhone.
Many interesting-looking videos come my way.
I want to watch them when I get home.
I need some way to get these videos posted to my TV at home.

Almost every app has a share via email option so that felt like the simplest and most flexible way around the problem.

So my requirement was: every user of *vidbox should be able to send emails containing links to have them added to their queue.

Cue a quick look at the Heroku addons and a flick through the docs for CloudMailIn, Mailgun and SendGrid. It was close between CloudMailIn and Mailgun but I decided to try the latter.

Mailgun has a load of fairly complex filtering features; I just wanted all emails to be forwarded to a webhook in my app for me to deal with and it wasn’t entirely obvious how to do this so I thought I’d write it up.

To get Mailgun to forward all email to my app I…

installed the addon on Heroku.
went to Routes.
created a new route.
entered a “filter expression” of “catch_all()” and an “action” of “forward(http://myapp.heroku.com/controller/process_email)”
saved it.

Good to go. Now every email will be passed on to my app.

And that’s basically it. The best thing about Mailgun (and the bit that surprised me the most) is that you never actually have to deal with email. At all. They break down the email into its constituent parts and post them to you in the data. All I needed from the incoming params was recipient and body-plain but a full list is available on the Mailgun docs.

Every *vidbox user is given an email address of token@vidbox.mailgun.org, where token is a 16 character hash, unique for each user. For each incoming message I can then find the sending user from params['recipient'] and the link from params['body-plain'] (with a simple regular expression).

So when I find a video that looks interesting while I’m out and about I just forward it to my *vidbox address and voila - it’s in my queue. Sweet!

*vidbox - your sofa-friendly video todo list

Mon, 27 Feb 2012 00:59:06 +0000

This was a Sunday hack project to scratch an itch.

Netflix (in the UK at least) don’t have any way to bookmark stuff to watch later, while iPlayer, 4OD, YouTube and Vimeo all have their own. Cue much switching of tabs. Services like VHX and Spool are great but seem to be aimed at tablet watching, with tiny text that’s hard to read on a TV screen from across the room. On top of that, I want to be able to add to my list from my Macbook at the office and Tweetbot on the bus.

Long story short, I made a thing. It takes in links from a bookmarklet or email, puts them on a 10ft-readable page that doesn’t require re-login and tracks the ones you’ve clicked on.

*vidbox - your sofa-friendly video todo list

Hammer Principle, Lean Thinking and motivation

Mon, 13 Feb 2012 11:47:11 +0000

While on holiday over Christmas I finally got around to reading The Lean Startup by Eric Ries. I’ve been meaning to pick it up for ages but thought I’d probably know most of it already, having seen him speak and talked to a lot of folk about Lean at various socials. I was completely wrong… It’s one of the best and most inspirational books I’ve read in a while. It also happens to be one of the most misrepresented. I hear a lot of talk about minimum viable product; few mention the scientific method, which was the revelation of the book. This focus on running experiments, in the form of small hypothesis -> test -> learn loops, got me really exited. Talking to David about it, we decided to try out a few Lean principles on Hammer Principle, our neglected baby.

Despite getting a constant stream of traffic, Hammer Principle has been essentially untouched for the last 18 months or so while David and I have concentrated on Aframe. We started a complete rewrite to mark its year anniversary but it got derailed and abandoned.

This time around, following the Lean philosophy, we formulated several hypothesis about users’ behaviour on Hammer Principle, concocted little experiments to test them, wrote small features to gather data and added Mixpanel tracking to try to track the results.

To give a quick example, Hammer Principle gets constant traffic to programming languages and databases but very little to martial arts. We posited that this was because people didn’t know about our other nails (what we call categories). Experiment: make it obvious that there are other nails to visit by adding a global navigation bar. Measure: people who use the navigation bar who go on to participate in another nail. The initial results have been encouraging (we’re getting about 10% of visitors following this funnel) and we now have a baseline to improve upon. This was our first experiment and I’m not convinced that we necessarily got our process right but there was an unexpected consequence…

Leaving the whole Lean circus aside, having targets and metrics has made working fun again. We develop something small that doesn’t take long and watch the counters tick. Conclusions are drawn and we then decide what to do next. Instead of pouring an arbitrary amount of effort into a black hole, hoping headline metrics like traffic and retweets will let us know whether we’ve done the right thing, we get a real sense of progress towards a goal, which is satisfying and motivating.

It doesn’t matter whether you’re doing a startup or just noodling on a personal project, I highly recommend reading up on Lean and playing with an experiment or two.

Which fish?

Mon, 04 Jul 2011 10:00:06 +0100

My friend and partner in crime, DRMaciver, pointed me to David McCandless & Derek Guo’s shocking visualisation of Atlantic fish stocks in the Guardian. I’d been vaguely concerned about our fish-eating habits. I happen to love fish but the murmurs about the state of fishing has been making me increasingly uneasy, fuelled by documentaries like Shark Water and Blue Planet. But I find it very hard to make a decision about what’s okay to eat. I simply can’t keep the information in my head.

An IRC buddy of mine, Zarkonnen, had a similar problem and turned the Guardian’s data list into HTML and JSON. I pulled in his JSON and turned it into a mobile-friendly, home-screenable list. Nothing fancy but I’ve been using it for the last few days and have found it genuinely useful already.

So, here it is: Which Fish?

TallyHo.it - keep a count of anything you want for whatever you want

Sun, 29 May 2011 14:13:00 +0100

I’ve been getting very into personal instrumentation recently, tracking energy expenditure via a Fitbit, weight via some Withings scales, mood via Mappiness, movies, gigs and books via my very own @oo5 project. I started noting all this stuff out of curiosity; I didn’t have any concrete plans for the data. There were vague thoughts of some kind of visualisation to help me find correlations between my habits and my general wellbeing. I ran across an anecdote from the Evernote CEO, related by Tim Ferriss, which got me thinking though. This chap had been trying to lose weight for years. He’d tried every fad diet under the sun but while the weight shifted, it always came back. The single most important thing in his road to a leaner, meaner self was simply to weigh himself every day and let his subconscious do the rest.

Since reading that I’ve noticed something about my behaviour as it relates to what I do, where and when. My Fitbit makes me think twice about getting the bus all the way to work; my Withings scales influence my lunch destination; @oo5 (with the help of my Kindle) effects whether I browse the internet, catch up on Instapaper or read a proper book. The simple act of measuring things has changed my behaviour.

When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the stage of science.

— Lord Kelvin in 1884

Curious, I wanted to try measuring some things I wanted to change. I tried using the very beautiful Daytum but it was just a touch too slow to inject into my daily routine. After a bit of a think I decided to make a simple iPhone-installable counter, that could be launched fast, used quickly, and forgotten about. So on the way back from SXSW I hacked together TallyHo. You can create counters for anything you like, add them to your home screen, and away you go. I’m tracking a whole bunch of things, including cups of coffee, which I should drink less of, glasses of water, which I should drink more of, and plastic bags, which I want to use less of.

The main focus is on what you’ve done today. There’s a big number in the middle of the screen telling you how you’re doing. Underneath is the medium term view - how you’re doing this week, and the long term view - how you’re doing overall.

I released TallyHo quietly a while back and uptake has been surprising. Somewhere over 1200 items have been counted to date, with tallies including:

Cigarettes
Coffees
Beers
Teas
Indigestion
Bars of chocolate
Bags of crisps

At the moment the data gets synced to my server and there’s an option to claim your tally by signing into Twitter, so you will be able to retrieve your data should anything happen to your phone. I might add a sync to Daytum too, to benefit from their lovely visualisations. In the meantime, drop me a line if there’s anything you’d like to see TallyHo do. I’d also love to hear from you if it’s helped you change something (anything!) about yourself.

My predictions for 2011

Fri, 07 Jan 2011 12:56:36 +0000

Web Designer Depot have published my predictions for 2011 alongside a bunch of my favourite designers. Here’s what I had to say, with bonus hyper-linky goodness.

If 2010 was the year that mobile came of age then 2011 will see it move into its own apartment next door to the desktop and start throwing wild parties. Having started as the younger, slightly neglected sibling it’s now on the verge of shouldering the desktop out of the way.

The industry (as articulated by Fred Wilson of Union Square Ventures) is coming around to the idea that mobile could often be the primary platform for new apps. As web developers and designers we are in a great position to take advantage of this shift through infrastructure enhancements like HTML5 and Phone Gap, and frameworks like jQuery Mobile, Zepto and Backbone, along with a horde of others.

All of these are based on technologies we already know and can start working with today. Rather than clumsily reskinning our websites, 2011 will be the year we embrace the philosophy of mobile to produce context aware, task tailored, fundamentally handheld apps.

This time last year - Flickr reminiscence

Wed, 05 Jan 2011 17:58:00 +0000

This time last year I had just arrived in Belize. I was standing on the beach in Placencia, on the world’s narrowest high street, with my brother and friend Brett. Our ferry had been cancelled and we’d spent the whole day working our way from Honduras up the coast by truck, boat, bus and lancha.

With the New Year just gone I thought it might be nice to reminisce so I build a little hack project to help me mine the hundreds of memories that my Flickr stream contains. It’s super basic at the moment but I thought I’d put it live anyway. So. Introducing:

This time last year - Flickr reminiscence

It’s really intended for use on my iPhone but I’ll add a desktop format at some point.

Rise & Fall: The Historical Trajectory of Power & Politics

Tue, 22 Jun 2010 12:55:09 +0100

While I was in Central America over New Year I found myself trying to place the Maya, the Olmec and the Aztecs into the timeline of world history. I don’t think I’ve ever seen them in context before - everything I’ve read deals with a single empire, civilisation or region. So I decided to draw a graph to find out. The result is Rise & Fall: The Historical Trajectory of Power & Politics.

Setting about Wikipedia with random abandon, looking up dates and following links I have creating an editorial list of the things I think are noteworthy. I’ve played it fast and loose with dates and inclusions. Broadly speaking: early entries are cultures; late ones are empires. There’s a fine line between them, when there’s any line at all. Cultures tended to be subsumed so their “end” is debatable; empires were conquered but have a tendency to decline so their end is also in question. What I’m trying to say is that this chart is just my interpretation. If you want to know more, have a read. If you decide I’ve made a mistake, drop me an email.

Out of curiosity I also decided to try to give some indication of the cultural legacy of the world’s powers. As a rough yardstick I’m using the number of Google search results for the quoted search “name culture”. This is skewed towards those whose names have continued - Chinese and Egyptian, for example - and against those who have died out - like the Mauryan and Bagan - but it’s an interesting measure.

A couple of features I like:

The straight line marking the end of the Phoenician, Kushite, Greek and Persian empires as the Romans rise.

The big gap between the Romans and the Spanish - the Dark Ages - while Asia and the Middle East flourish.

The Maurya, apparently a golden age for the Indian subcontinent, fading from Western memory and the chart.

Enjoy!

Rise & Fall: The Historical Trajectory of Power & Politics.

Elect or not? compared to proper polls

Thu, 29 Apr 2010 16:00:14 +0100

Over the last week or so I’ve started pushing a little thought tool/influence experiment I’ve set up called Elect or not? I’m trying to gently nudge people into thinking about why they vote for who they vote for while hopefully gathering some data on the halo effect while I’m at it. I’ll write more about the science behind it when I get some more time.

Out of curiosity I just crunched the numbers for the results I’ve had so far and thought I’d share… Bear in mind that this is people saying who’d they vote for based on looks alone, without knowing who the candidate is, who they represent or what their policies might be…

32% Conservative
25% Labour
24% Liberal Democrat

Compare these results to some of the proper polls…

YouGov (2010-04-28):

34% Conservative
27% Labour
31% Liberal Democrat

ComRes (2010-04-27):

36% Conservative
29% Labour
26% Liberal Democrat

ICM (2010-04-25):

33% Conservatives
28% Labour
30% Liberal Democrats

Tweetminister (experimental):

35% Conservative
30% Labour
26% Liberal Democrat

It’s a little bit scary but my numbers don’t seem too far removed from these far more sensible and official polls… See the data for yourself on the Elect or not? Parliament page.

Some Sass tools on Github

Wed, 07 Apr 2010 21:53:20 +0100

On a whim, and after an off-hand suggestion from Stef, I decided to pop some of my most-used Sass mixins - the ones that I include in every new project - up on Github. There’s nothing too ground-breaking here: a straight port of Eric Meyer’s reset.css, and a few CSS3 expanders.

The CSS3 expanders are probably the most interesting. I was getting irritated with having to type and retype certain of the new rules over and over again with the now ubiquitous -moz and -webkit prefixes. So I bundled them into mixins. Mixins are Sass’ way of adding very simple scripting to your stylesheets. Let me give you an example.

This Sass snippet:

div width: 500px height: 300px +box-shadow(#000, 0px, 3px, 5px)

Gets rendered as:

div { width: 500px; height: 300px; box-shadow: #000 0px 3px 5px -webkit-box-shadow: #000 0px 3px 5px }

Dead simple. I’ve wrapped up border-radius, box-shadow, gradient and box-sizing. Interested? Grab the files from Github.

YouTube feed to Boxee feed converter

Sat, 27 Feb 2010 19:00:00 +0000

To get my 4oD Boxee app working I first had to convert Channel 4’s YouTube feeds into Boxee’s format. I’ve gone with the simplest thing I possibly could - just including the bare essentials - but it serves my purpose so maybe it’ll be of some use to someone else.

In a nutshell: you pass it a YouTube username, it gives you back a Boxee RSS feed. Done.

Here are some examples:

4oD on Boxee

Thu, 25 Feb 2010 12:00:00 +0000

These days almost all my TV viewing has moved online. After much frustration I settled on Boxee as my home entertainment center of choice, running on a Mac Mini sitting on my shelf. It started quite rough but it’s been constantly improving and the new version is great. My favourite thing about Boxee is the iPlayer though. It puts almost all of my favourite TV content on Boxee. Almost all.

What’s been missing is 4oD, Channel 4’s IPTV service. Channel 4’s documentaries, in particular, are fantastic. Unfortunately when 4oD launched it was PC only. Last year they finally opened it up to Mac users but it was still limited to the browser, which is awkward to use from my sofa without a mouse. Awkward to the point of being unusable.

A few months ago, while waiting to be served at a pub in Soho, I overheard a man in a suit telling his colleague that Channel 4 were moving all their content over to YouTube. That sounded very exciting. Leaving aside the implications for the entertainment industry this meant that 4oD would have feeds.

Last week I happened across YouTube’s 4oD feeds and thought I’d try hacking a quick Boxee app together - mostly for my own benefit. Working with Boxee’s UI XML isn’t the easiest thing in the world but after a lot of reading, poking, twiddling and cursing I finally got something functioning. It’s very much alpha. Maybe even pre-alpha. But if you want to try it out here are the instructions:

Launch Boxee.
Go to “apps”.
Down the left, right at the bottom under “extras”, is “repositories”.
In there is an option to “add repository”.
Hit that and enter “boxee.donotremove.co.uk”.
That should get you “Mike’s Boxee repo” and in there is “4OD”.
Hit that and go “add to my apps”.
Start it.

The UI is limited, to say the least. 4oD channels are down the left; shows are down the right. Pressing right loads the show from the channel you have selected; left goes back to the channels. My biggest niggle at the moment is that loading channel content is a bit laggy. Let me know how it goes though! I’ll make improvements as I find the time and as I learn more about making Boxee apps. Or maybe Channel 4 have their own app in the works… We can hope!

Soft launch for You Might Not Know

Tue, 09 Feb 2010 13:00:00 +0000

You Might Not Know is a place to share those little tips you’ve picked along this road called Life. These tips can be able absolutely anything. If you think they’re useful then someone else will too. Or that’s what we think, anyway.

David and myself have been busy working on the prototype for a while now and after much procrastinating we’ve finally decided it’s safe to tell people about it.

I’m rather proud of how simple it is at the moment - mostly thanks to some cunning ideas from David. Adding a tip takes moments and I’m finding it rather compulsive. We’d like to be to Knol what Tumblr is to Wordpress.

We have a hundred things on the todo list but at this point we’d just love to see a few people playing with You Might Not Know. Let me know what you think!

Me at TEDx

Sun, 31 Jan 2010 16:00:00 +0000

Towards the end of last year I had the privilege to be invited to speak at the rather marvellous TEDx event put on by CodeWorks up in Newcastle. Each of the talks was filmed to the exacting TED standard and the result is a very slick-looking video of me rambling on about some of my favourite things.

Dr. Frankenstory gets his own Lab

Thu, 29 Oct 2009 13:45:00 +0000

From the beginning of Frankenstory we wanted a way to showcase stories. Our first attempt was very predictable: a list of stories that we could add our favourites to. But something about it just didn’t feel right… Frankenstory is supposed to be fun, first and foremost; it’s not a competition. Or at least not in the usual sense.

Enter Nick, who had the rather marvellous idea of writing a blog from the point of view of Dr. Frankenstory. He busily invented a history for the good doctor…

The story of my success begins many years ago in the sleepy Alpine village of Yödelayoo where I was working as a humble librarian. One fateful day as I was dusting a big pile of Mills and Boons high on a rickety bookcase, the shelves gave way and I was buried in a veritable avalanche of romantic fiction and soppy chick-lit. When I regained consciousness, I awoke to find my face was buried in an open book –Genetic Experimentation for Dummies. What’s more, the concussion I suffered as I fell jolted a dormant part of my brain, leading me to instantly absorb all of the incredible information in this hefty scientific tome.

…set out a manifesto…

Hello dear friends, welcome to the Frankenlab, the place where I’ll be preserving some of your most interesting submissions for posterity in a glorious ‘Frankenstory Hall of Fame’. Think of it as a kind of taxidermy display for your stories where I lovingly stuff and mount some of your freakiest creations (before gluing googly eyes to them and using them to scare small children).

…and busied himself with exposing this new life of literary experimentation to the internet at large, using Frankenstories as illustration.

The result is Dr. Frankenstory’s Frankenlab. And I completely love it. It kills a whole flock of birds:

putting the often incredibly random Frankenstories in an entertaining, light-hearted, fun context,
minimising any feeling of elitism by putting the decision in the hands of the Doctor,
eliminating any pressure that may have been induced by a leader board,
yet still giving us some way to promote and reward our favourite stories.

The latest post went up last night: Djinn and Tonic.

Stories and Experience at TEDx Newcastle

Fri, 02 Oct 2009 18:00:00 +0100

On Wednesday evening I had the pleasure of speaking at TEDx Newcastle, kindly organised by the good folk at Codeworks. My talk was about the memory of experiences.

Unless we make underwear or the like, our products probably spend most of their existence in the memory of our customers. In Stories and Experience I run through what we can do to help ensure that the experiences we design become memorable stories after the fact.

Slides (on Slideshare)

Books

QIF Converter

Fri, 25 Sep 2009 18:00:00 +0100

I’ve been using Wesabe for a few months now and I love it. I’m finding accumulating financial data quite addictive. Unfortunately, of the banks I use only my current account has any kind of export. I wrote a little script to parse my downloaded statement HTML into QIF files and over the weekend I decided to polish it a little and slap it on Heroku.

The result is my QIF Converter. It’s dead simple: you copy and paste from your online transaction into the textarea, specify how it’s formatted and hit convert. That’s pretty much it. If you want to import the transactions into your own financial planning app then you can just download the QIF file. If you’re using Wesabe and you have the downloader plugin installed you can just hit upload to Wesabe and you’re done. I’ve been using it to import the ridiculous PDF statements my bank insists on pensioning my old transactions into too. Bing!

Learning Ruby

Sat, 27 Jun 2009 17:00:00 +0100

A friend of mine wants to learn to program. I recommended either Ruby or Python and petitioned the Lazyweb for good places to start. In case they’re of use to anyone else, here are the recommendations I got back:

Poignant Guide

Way out in front is the Poignant Guide. It’s by why. It starts with a cartoon strip. Awesome.

Tryruby

An in-browser console. Allows you to dip your toes in the water without having to get your systems all set up. Bonus for the terminal-shy.

Hackety Hack

Calling itself the “coder’s starter kit”, Hackety Hack seems to be a self-contained, multi-platform training course, including a console, tutorials, cheat sheets and a place to store your own little apps. It’s currently being re-worked but worth keeping an eye on.

Brand Equity

Tue, 26 May 2009 20:11:00 +0100

Microsoft Aims Big Guns at Google, Asks Consumers to Rethink Search

Many will argue that no amount of advertising Microsoft throws at the product will make a difference – the quality of search results is the only thing that matters. And that may have once been true; after all, Google built its brand on the back of a great user experience, results that were markedly better and zero ad support.

Google has conducted internal tests in which the company put its logo and treatment on another engine’s search results. Users still prefer the results with the Google logo, even if they’re not Google results.

Then there are some stats from a research study called The Effect of Brand Awareness on the Evaluation of Search Engine Results. From this study, the authors concluded that Yahoo actually have the best brand perception, with their search results being rated a massive 15% better than Google’s.

Weblog

The video for Sleepless by Tropics, directed by my brother,...

Using Mailgun on Heroku to create user mailboxes for *vidbox

*vidbox - your sofa-friendly video todo list

Hammer Principle, Lean Thinking and motivation

Which fish?

TallyHo.it - keep a count of anything you want for whatever you want

My predictions for 2011

This time last year - Flickr reminiscence

Rise & Fall: The Historical Trajectory of Power & Politics

Elect or not? compared to proper polls

Some Sass tools on Github

YouTube feed to Boxee feed converter

4oD on Boxee

Soft launch for You Might Not Know

Me at TEDx

Dr. Frankenstory gets his own Lab

Stories and Experience at TEDx Newcastle

Slides (on Slideshare)

Books

Links

QIF Converter

Learning Ruby

Poignant Guide

Tryruby

Hackety Hack

Brand Equity