mechtopia

Graphically selecting smart phones

Here's an interesting way of picking a smartphone. Create a cloud of data points, then just delete the ones you don't want.

Posted at 07:13PM Feb 09, 2012 by ohp in General  |  Comments[0]

Abellio has a bad first full day of service

image: www.ukrailtweetmap.co.uk

It hasn't been a good day for Abellio so far, with today marking the first full day of service since taking over the old NXEA service yesterday. With the snow, website problems, a lorry hitting a bridge at Shenfield and a broken down train at Colchester, It looks like a faltering start for the new Greater Anglia Service.

Posted at 01:58PM Feb 06, 2012 by ohp in General  |  Comments[0]

Austerity Christmas tip #232: DIY Chocolate advent calendar

Chocolate advent calendars are often pricey, and the chocolates inside disappointing. Why not make your own advent calendar by cutting windows in the top of a chocolate box.

Posted at 05:11PM Dec 20, 2011 by ohp in Projects  |  Comments[0]

Hivegraph: collaborative infographics

www.hivegraph.com

In the upcoming weeks I will be blogging new charts created using hivegraph. Hivegraph is a new site that I've created. The site has a single data set that can be used to generate a wide variety of graphs. Current functionality is based around comparing items, and users can collectively add data, and create new charts. As new graphs and charts are created, I will also extend the functionality, and let users do more.

Posted at 07:52PM Oct 17, 2011 by ohp in Projects  |  Comments[0]

I thought I'd add my bullshit riot explanation to the mix.

Image by Bluesquarething

The media is still in a post-riot speculation frenzy as to the underlying cause of the recent rioting in London. So far I've heard the following:

• Social deprivation
• Racial tensions
• pseudo - racial tensions
• Government cuts
• Gang culture
• Anarchism
• Alcoholism
• Authoritarian conspiracy
• Lack of policing

... And many more.

So what's my bullshit explanation? well, it's more complicated than any one explanation. People yearn simplicity in explanations, so that they can find a simple fix. Sadly, life can be more complicated than that.

My hypothesis is that the events that unfolded formed around an idea: an idea to have a riot and to start looting that formed simultaneously after the reporting of the Tottenham riots, when people saw that there was violence and looting that was nothing to do with the protest, and that it was possible to exploit a situation without regard to an underlying cause.

This triggered a fire-storm of communication heralded by ubiquitous handheld smartphone ownership that has come about very recently. Maybe some of the reasons listed above were used to goad people into coming out on the streets, but none of these were central to the idea. A form of peer pressure and hysteria (perhaps with a little alcohol influence here and there) blinded people to the consequences of their actions. This brought people out to do things that they would never ordinarily have done by themselves, or even in small groups of their peers.

So how do we stop this from happening again? Personally, I don't think that it will happen in such a widespread fashion again for a while: The subsequent arrests have shown that the idea of rioting and looting in a city with such high CCTV coverage was stupid. The social networking that enhanced the transmission of the idea was also one of the most powerful tools to identify those involved.

But in 10 years, when this event has faded from memory, it may happen again.

Posted at 02:06PM Aug 14, 2011 by ohp in General  |  Comments[0]

The app is dead. Long live the web!

Recently I was asked to speak at a local dorkbot, and I thought I'd take the opportunity to have a bit of a live audience inteteraction experiment. Nothing fancy, but just enough to see if enough people had smartphones to make such interaction useful.

An app seemed like a good idea, but the more I thought about it, the less appealing it was. If I wanted to write an app that covered all of the possible phones out there, I'd have to pass the following hurdles:

Six App stores

Apple, Google, Blackberry, Nokia, Palm and Windows each have their own app stores, so I'd have to set up a developer account with each one. Some of them have a charge, bringing the up front cost to around 140 pounds. 6 app stores mean 6 developer agreements, 6 lots of making sure that they conform the the agreements, 6 lots of worrying if the apps will be accepted or not, potentially flushing months of work down the drain.

Four Development languages, Six SDKs

Android, blackberry and ovi could be developed in java. I'm proficient in Java, so it might take a few days getting the hang of several APIs, but not a big problem. Palm OS uses an html/css based approach which I'd need to investigate. Windows mobile will require me to dust off some rather rusty .NET skills, and for Apple, I'd have to start again from scratch with Objective C.

New hardware

My desktop environment is Linux.. Maybe I could run the windows SDK in a VM, but the real killer is going to be Apple. I can't develop iOS applications on Linux, or in a VM, I have to buy a mac. A basic mac is going to cost me at least £500.

Why not just develop for the big two?

I suspect that's what a lot of people end up doing. I'll only have to learn one new language, but I still have to buy a mac. But I'm excluding members of my audience who have smart phones! That's a tough choice to make.

The Web

You could see where this was all leading.. The web. The web has worked well for us for 20 years now, why doesn't it work well on smartphones?

If you want to do a 3d game, html and javascript still has a long way to go, and it doesn't work well offline. There are latency effects that can make user interfaces a little sluggish, and monetisation isn't so easy with web pages.

It does, however have some strong points. One SDK can be used. Tweaking for multiple phones doesn't require a rewrite. There are no app store agreements to sign, your app won't be pulled without notice. You won't have to go out and buy a mac.

HTML5

This is one reason why things are going to get better for the web, including the web on smartphones. Vector graphics and 3d graphics will soon be possible, and are already possible on some smart phones. Many companies (such as amazon) are providing apps via HTML 5 already, in order to get around restrictive App store policies.

Ultimately you might be using the web already. I understand from some insiders in the app business that a lot of apps are little more than a front page, and that the real functionality is provided by a back end web page.

Conclusion

I think the smartphone app still has legs. If you want to do something CPU intensive, want to easily monetise your product, or want to do something low level with the hardware, then an app is probably the safest bet. However, the web is fighting back, and HTML 5 will play a large part in that fight back, combined with better and faster mobile broadband connections. More basic data entry scenarios and even basic graphical applications (I've seen a demo of a javascript mobile game that is quite playable) are more easily and quickly brought to the market using the power of the web. The audience interaction system was written in a day, and supports every smartphone I've tested it with.

Posted at 12:50PM Aug 10, 2011 by ohp in General  |  Comments[0]

Refractive index bubble display prototype

Here's something I've been playing with. The idea is that you have a block of acrylic (or other transparent material) and within the acrylic there is a chamber milled out. When the chamber is filled with an oil with a matching refractive index, the outline of the chamber becomes indistinct. You can "turn on" the display element by draining the oil from the chamber, at which point a bubble appears within the acrylic.

Element "off" (filled with oil)	Element "on" (oil removed)

It was influenced by an exhibit I remember from the science museum in London, which had a perspex block containing a number of different fluids with different refractive indices.

I demonstrated this basic prototype to a gathering of Ipswich science in the pub on Tuesday 19th July, 2011, and we discussed various ways of improving the design.

There are still lots of issues to overcome:

• It's messy: oil gets everywhere
• Pumping oil to multiple chambers requires multiple drivers, which makes it expensive and unreliable.
• It'll be hard to ensure that the drivers don't underfill or overfill the oil in the chambers.
• Long term, the properties of the oil and acrylic are likely to change, especially in the harsh lighting conditions that show up the effect.
• How do you account for thermal expansion?

Currently I have a strategy for increasing the contrast between the two states: You can easily see the machining marks inside the perspex in my prototype, and I'll be making the main chamber rounder (rounded chambers show the effect much more), and the feed tubes squarer in order to enhance this effect, and using heat to remove the tooling marks. I'll then look at glueing the two perspex halves together, so that the block becomes one sealed unit.

I'm not sure if this is patentable, I haven't found any other patents in this area, and much googling has thrown up nothing. I did consider trying to patent it, but I'm releasing this idea into the public domain, as I didn't expend much research money coming up with it, and I don't think it's a practical display technology that's likely to make much money, but may be a fun idea to exploit for people who perhaps want to build it into art installations and one off pieces.

I eventually want to build a digital clock that uses this as a display, with different segments being driven in this manner. You could also make simpler designs, perhaps mill a single chamber into lettering, creating a sign that you could turn on and off. OPEN/CLOSED maybe? This car park is full? There are also a number of options for drivers, from solenoids to servos.

Posted at 12:00AM Jul 25, 2011 by ohp in Projects  |  Comments[0]

Q: what is the storage capacity of 78 floppy disks? A: 27 litres.

I've been thinking about using discarded objects to make useful items. This is a box made from 78 floppy disks. It is made from floppy disks and hot glue only.

Posted at 05:00PM Jul 12, 2011 by ohp in Projects  |  Comments[0]

At the exact moment (A disaster amplifier)

This is something that I made for the curiosity collective's time show last year. Time show 2 is on the horizon, so I thought I would write it up.

One thing that has caught my attention over the years is the concept of stopped clocks during disasters. Often a museum will display a clock that stopped at the exact moment a ship sunk, or a bomb went off, or an earthquake struck.

These disasters are large, life changing events, and yet we devalue the concept of such disasters with our everyday, mundane disasters: perhaps we failed a test, we broke a glass, we ran out of milk.

I then wondered how I might artificially amplify these mundane disasters so that they had the same impact as a proper "stopped clock" disaster, and so I made a machine.

The disaster that I chose was running out of chocolate bars. This was selected because I could mechanically sense when the last chocolate bar was removed from a stack. When the chocolate bars run out, a mallet is released, which smashes a clock.

When we ran out of chocolate bars, we could then exhibit the clock with a message that told of the conditions that led to it having stopped. It was very much like a pile of sticks game: you didn't know if you were taking the last chocolate bar, and woudln't know if your bar was the one that destroyed the clock, creating a heightened level of suspense. In all, it went well. The trigger mechanism needed adjusting slightly, and we did run out of chocolate much more quickly than anticipated during the reception event, but ultimately that was the point: those mini disasters were forever marked with the stopping of a clock.

Posted at 03:59PM May 27, 2011 by ohp in General  |  Comments[0]

Audience voting and participation with smartphones

I've often wondered if there are enough people with smartphones in a specific audience to be a useful means to provide audience feedback. I was asked to speak at the recent dorkbot anglia and I took an opportunity to create an application that might test this hypothesis.

I thought I'd start with something simple: smartphone users would select a colour from a range of colours using their touch screens, and the corresponding square for their seat would light up on the screen. This would allow me to ask the audience questions that could be expressed in terms of colour.

The process starts with individual QR codes attached to each seat: these are 2D barcodes that can be interpreted by reading software on smartphones, and converted into web site addresses. Each QR code resolves to a different web address.

When you load the page at the address, a grid is served which contains coloured squares. Selecting a square will communicate that colour to the server, which is running on the internet

The laptop running the projector at the front of the room has a page that shows the currently selected colour for each seat

So what have I learnt from this trial?

1. It shows that you don't need apps for this kind of application. Indeed, apps would be difficult to implement, as you would need a different app for every kind of phone that you might anticipate, and exclude phones that had browsers, but no app engine.
2. This audience was pretty tech savvy: I'd say that around 70% of the audience had a suitable phone, but it shows that a usable proportion of people may have an appropriate phone in many situations (although you couldn't use it where you wanted instant feedback from every member of the audience)

Where do I move from here?

I'm planning an interactive charity pub quiz that will use similar technology to this (although it'll be a little more complex) and I'm also planning on experimenting a little more with the javascript capabilities of mobiles.

If you are planning an audience participation project, feel free to contact me at the email link in the sidebar.

Photographs courtesy of Ross Scrivener

Posted at 02:58PM May 04, 2011 by ohp in General  |  Comments[0]