Category Archives: experimental

Everyware

I got a notice in my inbox today that Chumby Industries are finally (after what seems a loooong time) beginning to ship the first Chumbies to early adopters. I tried very hard last year with a series of increasingly sycophantic emails to Chumby to secure myself an beta model, and failed dismally, but at least I seem to be on the mailing list for the first people allowed to buy one…

chumbyThe Chumby, for those who haven’t come across it before, is a small Linux-based wifi device with screen – er, ok – it’s a computer – which sits on your wireless network. The content it displays is entirely hackable – you can point it at any number of Chumby widgets which sit within the Chumby network. These range from simple clocks to news to just plain weird stuff. The basic idea is that the plain ole’ desktop or web widget is now beginning to show itself in the real world.

I think this is exciting for two main reasons: 1. I reckon that widgets (in general, but right now, web and desktop based ones) are the biggest thing happening to content consumers right now, and 2. Ubiquitous computing / internet of things / the spime – the whole pervasive “internet in real world” thing is going to change the way we use web resources in a big way.

So what else is there? Well, mobiles, obviously – generic computing devices which we all carry with us, everywhere. Offshoots from here include SMS, MMS, QR tagging and mobile web browsing. When commentators like Tom Standage start saying things like

“…mobile phones are the most numerous digital devices on the planet, and truly deserve to be called “personal computers…”,

and major telcos start concentrating on the mobile web (T-Mobile’s “Web n Walk” and Vodafone “The internet is now mobile”), you know that these kind of approaches are leaving the steep bit of the hype curve and entering the mainstream.

MarcelAnother device in the Chumby space is the Nabaztag. Those lovely types at the Science Museum got me one of these as my leaving present and I’ve been hacking it every since. Marcel (picture on left, looking only mildly like a drug dealer..) connects to the web through my wifi network much like the Chumby and receives emails, weather reports, podcasts. Most usefully, I’ve also discovered that he has much more authority for telling my 3 year old to go to bed than me… ;-)

Meanwhile in the museum space, Ross Parry and associates recently presented on the concept of the Live!Label – small screen-based labels for exhibits which can be updated anytime via a wifi network.

The general premise of all of these devices is the same: the real world is where we live and move but the internet can start to be layered on top of that world, Matrix style, rather than separated from it. As wire-free (wifi, 3G, GPRS, EDGE..whatever!) access gets faster and more ubiquitous, this layered internet will begin more and more to play a part in our real, not just virtual, lives.

Au revoir, Science Museum…

The 14th September 2007 marked the end of an era, for me anyway. I’ve been at NMSI, the National Museum of Science and Industry, for just over 7 years, and that was my last day.

I move on, as anyone does from a job they’ve lived and loved for that length of time, with a huge range of emotions. I’m terribly sad to no longer be attached to an institution with such vast public kudos. I’ll miss the people hugely – I’ve never worked with such an interesting, creative and open-minded bunch before. I’ll probably never work on such a huge range and variety of projects. But it’s time to go, and I’m delighted and excited about the future I’ve got coming up in Bath.

As part of this post, I thought I’d jot down three or four of the key developments in the history of NMSI online since August 2000 – mainly this is indulgence, but I thought it might also cast some light on how and why things changed over the years. Personally, I’ve learnt hugely important things about the web, people, the complex set of politics which exist in any institution of this size and scope, not to mention museums online and the vast range of technologies available to us.

This is, by the way, an entirely non-exhaustive history. One day I’ll get round to charting everything out, but today is not that day ;-)

I first started at the Science Museum back in August 2000. I’d left Waterstone’s online at that point in any job when you start twitching: It was hugely hard work, but I wasn’t learning anything new. At the time, I was massively excited by the opportunity of working at the best museum in London (sorry, but it’s true..), but also arrogant enough – the dotcom boom providing 2 or 3 job offers each week – to negotiate quite hard with the museum prior to an offer. I told them I wasn’t going to come along unless they increased the pittance of operational budget which was then allocated to web, and also find some additional people to help make it happen. They agreed.

The first few months were terrifying, but exhilarating. There was a lot of blagging on my part: at the time I knew nothing whatsoever about how to put together an agenda or chair a meeting. I’d never managed a budget (having just negotiated a bigger one, this was particularly daunting…). I had a server to look after (and knew nothing about server-side technology). I couldn’t code. I had a vision, but no people who could help me do it. The museum had just reached an impasse with an agency who will remain nameless who had built them an interestingly exotic(!) “CMS”. The site had just been re-designed but loads of snagging issues remained from the old site (wait for images to load for full effect!).

I muddled through. I bought books on ASP. I junked the CMS system the agency had built and installed my own homebuilt version (not particularly popular, that move, given what had been spent…). I patched up the server as it memory-leaked and limped its way from day to day. I did frightening things like find and replace the file extensions on the entire site to convert it to .asp…(and yes, I backed up first…)

Shortly after that, I persuaded Daniel Evans to come to the museum from Waterstone’s Online. He proved an incredible asset. Just after that, the dotcom boom crashed into the world of online bookselling and the 60-strong staff at W/O was “consolidated” into 3 or 4. We felt good having escaped.

Rolling out the CPS, or Content Publishing System – a simple VBScript application which let users around the museum edit their own content – was the first major milestone for us. It marked the point at which we seriously began handing ownership of the content to the organisation. At last, people started to appreciate why they should own and change their stuff. At the same time the system largely side-stepped the “resource bottleneck” which so often exists in web teams, but also left publishing control with the web team. We tried hard not to edit too much, and it also gave us a chance to prevent Comic Sans showing its horrible face on our site…

At around this time, I started working with Ann Borda (now at JISC) to develop a concept for what would later become Ingenious. It started life as “Science & Culture”. It’s interesting to note, given the vast remit (the first cross-NMSI project) and the huge timescale (over 3 years), that the very first sketches we presented were pretty much what we finally delivered in June 2004. This was the first lesson for me, and the first real resistance I developed to that all-pervasive museum treacle: projects are better done over short bursts, with small groups of stakeholders who are capable of moving fast and deciding quickly. It took huge energy (which to everyone’s credit, they retained over 3 long years) and quantities of strong coffee to make the site happen. Although it has very obvious shortfallings (second lesson: less really is more…), I’m still proud of what we achieved.

Making the Modern World Online followed soon afterwards. For the most part, this project ran outside the web team, but we had input on accessibility and design, and helped steer it (mostly) in a strategic direction which roughly followed what we wanted to achieve for the museum online. During this time, we worked hard on developing web policies and strategies to support us in everything we did. The key lesson we learnt here (it looks obvious now, but it was a revelation at the time..) was to align – 100% – all our strategic thinking with the goals of the organisation, literally drawing lines between what the wider business wanted to achieve and what web could do to support those goals. We coupled this with incredibly close work with the Visitor Research team. Third lesson: end users are the best friends you can possibly have, and will provide you with endless ammunitation to throw at the internal politics..

Next up was continued development of Sciencemuseumstore and then the launch of the Dana Centre. The original website for this was put together out of a small budget and little time, and this time we pushed forwards with more efficiency – although a small project we did it quickly with just a sidewards glance at the politics. The site was later re-built and relaunched by Frankie Roberto, the museums second Web Developer, and by gum it looks and works a whole lot better than it did the first time around…

Meanwhile, Daniel and Joe Cutting had started working on a vision for Antenna, our rapidly changing science news section. Initially, I have to be honest, I wasn’t 100% sure what exactly they were banging on about: XML and XSLT were new and mysterious things to me, and I couldn’t really see how they would help (cue sheepish grin..). Luckily they just got on with it rather than trying to explain it to me, and ended up writing a content system which revolutionised the way that the Antenna team edit and publish content. In brief, the system allows creation of a single XML-based content source which is then re-purposed to both web and gallery kiosks. Dan did some tests at some point and found that entire gallery stories could be built and published in around 12 minutes, a huge resource saving to the 2-3 days it was taking prior to that. The system is still in place today, and was really the pre-cursor to our understanding of how a good CMS system should capitalise on XML to deliver content to multiple channels. A similar approach was used by the agency who developed the fabulous Energy website, and continues to be the end-goal for Content Management at NMSI today: one “pot” of content delivering to gallery, web, mobile and anywhere else we choose…

Meanwhile, we were working hard on the Science Museum website, consolidating content, tidying up bad code, trying to CSS the whole thing. Behind the scenes, I was rallying for budget to re-develop it. Note “re-develop” rather than “re-design”: we knew we wanted to do something radical with the entire thing rather than just re-skin it: this is what we’d done in 2000 and apart from making it look better, it had still remained badly broken under the hood.

Eventually we got budget. The entire re-development project probably took about 3 years – again, far too long – but we remained incredibly enthusiastic with the vision we put together and the agencies we took on to do the work. The energy remained pretty high, which is always the most important thing. The new site went live on 26th March 2007. Beautiful, isn’t it?

At the same time (and looking back I can really see that we took on far too much in one go..) I was also working on putting together a vision for Content Management at NMSI. After a long procurement process we bought Sitecore, a fabulously powerful, standards-compliant .NET system. The ultimate, organisation-wide vision of building in Enterprise Content Management to everything content-related is still in its infancy at NMSI, but Web CM is the first, very visible starting point on that journey.

Of course we also continued to build in user generated content and new technologies wherever we could. Our web strategy took the organisational direction and applied UGC, drawing parallels between what our stakeholders wanted and what the web can usefully deliver. This ranged from SMS messaging during risqué Dana debates, encouraging visitors to bring in toys, a range of RSS feeds, allowing users to Ask Glenn – to mention but a few…

Next?

I have no doubts at all that web will continue to grow in importance and stature at NMSI. The vision, the environment, the beginnings I’ve had the privilege to be involved in – all point to an incredibly interesting future. I’ll be watching (with only occasional twinges of regret..) and undoubtedly blogging about it too.

The next huge thing on the immediate radar is the launch of Launchpad, the flagship hands-on gallery at the museum which is due to re-open – much bigger and improved – later in the year. I’ll be posting very, very soon about the online element of this. I’ve had the privilege of helping develop the concept for this and have watched it grow into something absolutely outstanding. It’s one of the best things I’ve seen for some time. But you’ll have to wait a little while before you too get to see it…

Netvibes universe for NMSI

I got involved with Netvibes a long time ago, first as a user and then briefly when I helped them out with some dodgy English translations. That’s how I came to be invited to set up a Netvibes Universe before the beta was opened to the public.

If you haven’t used or come across Netvibes, you’ve been seriously missing out on a major productivity improver – it’s essentially an Ajaxified tabbed start page which lets you embed feeds, calendars, video and searches into one place. What’s more, the recent opening up of the universal widget API now means that developers with little more than XHTML skills can create widgets that do pretty much anything – you could have (and at some point I’m going to build) a search box for your collections, for example. There are vast quantities of widgets available.

NMSI Netvibes UniverseNetvibes have recently announced the concept of the Netvibes Universe – a place where institutions, groups or societies can set up a page with a specific focus, embedding feeds and so on that are specific to that particular field of interest. Once you’ve created a Netvibes account you can add Universes to your page. It’s all a bit difficult to explain but should make sense once you’re there…

The National Museum of Science and Industry Netvibes Universe can be seen at www.netvibes.com/nmsi – I’m still adding content and playing with what we can do with this space but I think it’s an interesting slant on using content which is about but not necessarily generated by a particular institution.

It’ll be interesting to hear what you think.

Stufflinker – I need your help

I’m looking for more beta testers for Stufflinker, my mobile and social bookmarking tool.

Stufflinker lets you save thoughts, urls and other, er, stuff, while you’re out and about using text messages, as well as the creation of ‘stufftags’ which link through to specific content on the web.

To take part in the beta you have to be in the uk, have a mobile, and be up for some fiddling.

Please sling a comment at this post if you’re interested and I’ll get in contact.

Cheers…!

Why am I learning this stuff?

As I mentioned on a previous post, I smuggled my PC on holiday and had a go at learning Ruby on Rails. I’m not going to spend much time talking about what I think. Needless to say, I had fun starting from knowledge = nil and gently climbing up the learning curve towards knowledge = 0.1. And yes, Ruby is obviously very cool.

After two weeks of having a go at this stuff on an evening, I thought I’d sit back and re-assess exactly what I was doing, and why.

The lowdown is this:

> I’m an average programmer (possibly slightly above average because of my extreme / anally retentive progamming neatness) in ASP/VbScript
> There was a day when I did this as part of my day job, but not any longer. Now, I just send emails a lot.
> I have been able to solve most web application problems the world has thrown at me.
> I know well that VBScript/ASP has a limited lifespan, won’t be supported by MS for much longer and is laughed about in “real” developer circles.
> I accepted a looong time ago, grudgingly, that I’ll never know everything about everything. Maybe.

So apart from having a nice time learning something (and there’s a lot to be said for that without needing another reason), what exactly am I doing, trying to learn this new – completely new – way of coding?

Well, I have aspirations to retire any day now, and want to wake on a morning and know that I’ve truly made it by having to swim ashore to do my weekly shop. I intend to make this happen either with an extraordinary spate of deeply cunning crime or by building a web business that is so completely cool that no-one else will be able to touch me. The web business seems more sensible (for now at least) and I’ve got enough ideas to keep knocking them out faster than I need to. And here’s the crux. Bring on the graph…

Technical learning graph

[I love a graph almost as much as I love a diagram]

Here’s what’s going on. Up the X is technical proficiency from zero at the bottom to frightening at the top. Along the Y axis is time. Now think about me and my knowledge of VbScript. It’s up there – granted after a couple of years worth of hard slog – but I can do stuff in it, so my technical proficiency is pretty high. My knowledge might actually still be increasing so the line should be off-horizontal but there is also a threshold implied by its flatness.

Next up in this case is Ruby, but it could be any serious web development framework/language. When I started leaning Ruby, I knew nothing, but you’d hope my proficiency would increase with time. So there’s a gradient to it, which is me learning.

Now, there are also two thresholds which I have called the Production threshold and the Prototyping threshold. The Prototyping threshold is the point at which I’d feel confident enough to knock out some alpha versions, do some user testing, play about with the look and feel. I’d probably still be using an Access database under the hood though, and the code would be terrible. In other words, if I got more than three visitors my application would probably emit a funny smell and die on its arse. The Production threshold is the place that serious developers want to be (actually, really serious ones want to be at the Geek threshold which isn’t shown here. You know the people I mean…)

What I’m coming to realise is that I’ll probably never now be the person who writes production-level code – instead, I’ll pay, bribe or beg someone else to do it. So for my particular purposes the Production threshold and anything beyond is an academically interesting thing but actually nothing more than that. Instead, I’m only ever going to knock out experiments and demonstrations for ideas which other people can then take and make production-level. I don’t actually care how fast, well compiled, transportable, modern or supported my prototyping language of choice is. In the slightest.

The time it takes to get me to Point A is the only thing I’m interested in. And because that green Ruby line has a gradient to it, the time to Point A is unfortunately not zero (bring on the Matrix – “I know Ruby…”). The time to get to the Prototyping threshold for VbScript has already passed. I’m there, I’ve learnt it.

Uncomfortable though it is, I’m therefore going to dump Ruby. She’s very fine, but I’ve got too much to do, and not enough time to do it. VbScript will have to do…

Museum directory v2.0

In my previous post about the “museum directory” I built at UK Museums on the Web mashup day, I mentioned a museum address CSV file from the 24hr Museum which I planned to put use at a later date.

The original source I had contained *really* dodgy data and only about 380 institutions – I’d done some seriously horrible hacking to get it out of various APIs – but the new feed is derived from the 24hr Museum “Direct Data Entry” (DDE) system. This contains around 3,800 entries and is therefore much more interesting as a dataset.

24hr Museum have asked that I don’t expose the KML file at this time, so what you see is the museum directory as it was in version one but with more, and more accurate, data. Version 2 is pretty much the same code and approach as the original – read about how I did it here.

The new data set still didn’t contain geo references, so I had to re-hack the original postcode script to query the Google AJAX API on a running basis and write the lat/long back to the database. That hurt a bit – nearly 4,000 queries takes a long time, especially when postcodes weren’t found. This slightly manual approach, together with some discrepancies in the CSV I had to deal with by hand led me to add the disclaimer on the page about accuracy – nothing to do with the original data…

Anyway, enough tech rubbish. Go play with version two and let me know any ideas or thoughts you have. I’m already thinking about the next version which is gonna be a whole lot more exciting, functionality-wise…

Thought clarification: JUST DO IT but FOR A REASON

A long and interesting thread broke out on the Museums Computer Group mailing list today about how museums could use Facebook to their best advantage. As I said on the thread – although the question about how Facebook deals with organisations vs individuals is interesting, the key question to me is what we’re trying to get out of having a presence on social networking sites.

Although I spend a lot of time going on about how we should “just do it” (good tagline, that. Shame it’s been claimed by a global corporation of dubious ethics..), I’m also well aware that museums aren’t immune from the hype curve either. The suggestion we should “do something with Facebook” throughout the thread is terribly reminiscent of many requests I’ve had to “do web 2.0″. The conversation usually goes like this:

——————

Web team office, early morning. Somewhere a phone rings.

Web Team: “good morning, this is your friendly web team. how can I help?”

Important Person, usually somewhere high up in the organisation: “we need a blog/discussion board/wiki/podcast/facebook account/mobile website/[insert other new tech thingy here]”

WT: “why?”

IP: “because I read an article in the Guardian on Saturday and it’ll improve our productivity/sales/grooviness. Besides, it’s free”

WT: “what do you want to say on your blog/discussion board/wiki/[...you get the picture...] ?”

IP: “why does that matter?”

WT: “who is your audience?”

IP: “the kids, of course. da street. da yoof. innit?”

WT: …

IP: “right, I’ll hope to see some serious re-alignment of our visitor figures by, say, a week Wednesday. I is expectin’ big fings in da hood. Bitchin’. ”

——————-

There’s a fine line of course between what I push for – technology growth, user understanding, fast to market, flexible applications – and the Important Person’s vision. This is a subtle game, and one which often causes concerns.

I see it like this:

> the mashup environment is about playing with technology – it is therefore partially technology driven (a bad thing) but also understands and build on content and data from disparate sources in the hope that the thing which pops out at the end is useful (a good thing). It relies on a Darwinian process to determine what works and what doesn’t: if your users like it, they’ll take to it and it’ll succeed.

> the drive to make things happen – the push which I believe museums should be making to be more leading than lagging – should always come out of user centred design. Websites should come from a user need. Ultimately, they should fill a hole in people’s lives. The bitter pill to swallow is that the needs of the institution aren’t always the needs of the user, and that’s where conversations like the one above start to cause pain.

Sometimes the needs of the institution do match (or can be bent so they match) the needs of the end user – this is when the best things happen. Take for example the fabulous English Cut blog – a fascinating look into the otherwise closed world of the Savile Row tailor. Hugh Mcleod helped put this together and he writes wonderfully about the value of the “micro smarter conversation” vs the value of the “macro brand metaphor”.

This is where web teams need to be incredibly savvy about what is out there and how to make this stuff happen. Actually, the conversation above should have a moment where Web Team gets in quickly with “Good plan, Mrs Important Person. How about a personal blog written by X about the way in which we Y”, thereby cutting off any possibility that you’ll “just do it” in the wrong direction with some god-awful corporate nonsense.

So….should museums be on Facebook? Yes, probably, if that presence does something interesting and motivating for users. Should museums be on Facebook just because it’s there? Obviously not.

Google Mashup Editor: first impressions

Unfinished shedI just got my alpha sign-in from Google for their Mashup Editor. Immediately, this ruined any good intentions I have for finishing off my shed but hey, every sane person is in bed at 7am on a Saturday morning, so it’s time in lieu as far as I’m concerned.

First impressions: true to Google style, they’re focussing on their developer base first and foremost and will probably worry about the GUI later. It’s sparse – essentially a code editor view and a couple of tabs for feeds and sandboxing. Compared to Yahoo! Pipes it’s very unimpressive to look at, but Pipes was always going to be a winner GUI solely on the basis of its innovative drag and drop interface.

Google Mashup Editor - editWhat’s more important, of course, is what it does..

At its heart, GME gives you access into a kind of metalanguage which is a mish-mash of html, javascript and the <gm: ***> namespace. It’s very reminiscent of the vague dabblings I’ve had with .Net. For instance, to define a repeating row based on some kind of input feed, you’d say:

    <tr repeat="true">
      <td><gm:text ref="atom:title"/></td>
      <td><gm:text ref="digg:diggCount"/></td>
    </tr>

As with any coding based environment, you make changes direct into the line editor, then save and preview – when you do this, the sandbox tab comes up and you get to see your creation. Or an error message…

Maps wikiThe samples are impressive and give some idea as to the range of ideas that can be implemented. The screen grab on the right for example is a “Maps wiki” which lets you add a point with comments and a rating onto a Google Map. You can see the GME code for this sample here. Not bad for 81 lines of code…

When you push publish, your application is pushed up onto the Google servers – interestingly, you can also choose to publish (and preview, while editing) as a Google Gadget which means anyone can embed the widget onto their personalised Google startpage. There’s a few mashups already online in the Google Mashup Gallery.

Crucially, the cross-over between gadgets on the web and gadgets on your desktop is blurring (see Google’s “Google Desktop” gadget list, for example) so expect this to become a very interesting space to watch.

What’s important with all of this – and I think this is interestingly reminiscent of the (first days of the) web – is that you can “view source” on any of these apps: much like Pipes where you can clone anything that anyone else has done, the starting point is often “I’ll copy what person X has done and extend it so it works for me”. This is a new and welcome paradigm which keeps appearing – MIT Scratch which I reviewed briefly a while back has a similar approach: once you publish, your work is automatically available to others. This will ensure that beginners always have code samples to learn from, and ultimately grows creativity exponentially.

I’ve so far spent all my time writing this rather than playing, but my first impressions are that although graphically less impressive, the potential here is far greater than Yahoo! Pipes: being able to save data, edit at code level, embed javascripting, tweak CSS, publish as a Google Gadget, etc. – these feel like the beginnings of something very, very cool.

I’m still waiting for my Popfly account (c’mon Microsoft, catch up..) – hopefully will be able to report back soon.

Right, must go and do some more playing. Er, I mean, finish off my shed…

Send your ideas partying

I stumbled across Paul Walk’s post on breakthroughs a couple of nights ago and meant to have a serious conversation with him during our regular ElPub gathering. Instead I was wearing a silly hat (Facebook login required) and no such conversation happened.

The point he’s picked up on is a good one. Stuff happens when you get together and talk – to quote from someone called Terry Frazier who Paul, er, quotes from in his post:

“Introverts, no matter how smart, rarely make breakthroughs – Breakthroughs do not happen in front of your face. They happen in the connections and gaps and networks that emerge from constant forward action and focus”

This seems to me an important point when trying to shape environments which innovate. Andy Powell from Eduserv (who was also lurking at the pub in a fine pair of virtual shoes) agreed – getting out there in front of peers and competitors was the only way in which thoughts could evolve and grow.

Perfection geneI’d take the point a little further – I reckon introverted ideas don’t go very far either.

One of my major shortfallings is that I have a problem with OPG – an overactive perfection gene. I blogged about this about 5 years ago and drew the graph on left whose asymptoticness (?) illustrates the problem. Voltaire (or Flaubert, depending on who you believe) got there first with the phrase “perfection is the enemy of the good” but I drew a pretty graph so I reckon I at least earn some credibility..

The point is this: you come up with an idea, you tinker, you fiddle, you fix, you draw up a business plan, you dream of word domination. 3 years down the line the thing is still a half formed whotsit on your hard drive / in your desk / on an A4 pad and sweet f/a has actually happened. I’ve got a house and a head full of these swines – perfectly formed, beautifully optimised, wonderfully branded. And going precisely nowhere.

So, what’s the solution? Well, it’s really painful when you’ve got OPG, but force your idea out the door; give it just enough makeup to make it respectable. Make your introverted idea work for itself; get feedback, provide links, test it. You’ll realise before long that:

1. Lots of people have had the idea before. Get over it.

2. If it’s good and people like it, they’ll use it. If it isn’t, it’ll fall on its arse and you’ll learn something.

3. Google probably won’t care how optimised it is when they come knocking. So don’t spend time perfecting – spend time getting a solid user base instead.

This now sounds like a sermon, which it isn’t, more a mantra to keep repeating to myself. After all, I’ve got Stufflinker just sitting there, and IdeaSurface which never really went anywhere…oh, and Moving Words (IE only, never quite got round to the Firefox version) and a bunch of tunes I wrote, and Captionizer and Radish and…

Maybe this is why I think the whole mashup and data freedom thing is important, and why I’m such a big fan of making things happen. It’s better that your ideas are off partying and getting knocked down than just stuck in your head: protected, yes, but also fairly useless.