Technical Mangroves

Startups - Just Stop Now

2011-10-09T12:43:00.001Z

Startups are sprouting around the globe in such numbers that there's even a industry creating websites listing the latest entrants. There's a bubble here, partly encouraged by the ongoing financial crisis. Even without that, the standard notion of a startup is a terrible business model to begin with. I'd like to suggest an alternative. First though, I should probably justify the rash assertions that this post started with.

There are thousands of startups vying for eyeballs on the web at the moment. Sites have been launched just to keep track of them - Beta Li.st, Killer Startups, Startup Search, there's even a directory of directories, Submit Startup.com. What's driving this bloom of entrepreneurial growth?

At first, the world financial crisis would seem to be a strange time for so many new businesses to be launching themselves into the world, hoping for funding and success. Yet it's exactly because of the currently disasterous markets that funding is finding its way to would-be Google killers and online pet-food stores. There's still money (quite a lot of it) in the system - but with gold at a nerve-wracking high, the stock-markets finding new lows and property in the doldrums, investors are looking for alternatives to bring in rewards. Add in a few out of work web developers and unemployable business graduates and the consequences are almost inevitable.

Of course it helps that a few bright sparks have gone on to blaze a path, raising millions or in a few case billions for their founders. Who could resist investing a little time and money when the rewards are so great? Yet a startup is a terrible business model.

..a startup is a terrible business model

The pattern is very well established. Have a great idea. Put together a dream team to implement it. Launch a cryptic website and sign up beta testers. Get some funding now you have a demo-able product. Preview the great concept with testers and improve the offering with agile development. Allow more people in from your list of testers. Encourage the site to go viral, get exciting reviews. Open the project up to the world and hit the big time.

It's true that when that laundry list works, it looks textbook simple. Sure, we are talking about a couple of years of your life just to get to launch, but that's just graft. No magic required, just working the business. Except for each Groupon, or Twitter there are hundreds, even thousands of never-heard-of-them companies dead or dying by the roadside.

For some, you can place the blame fairly on the execution. There was a great idea let down by never quite making it to market in a coherent shape. Some are just lousy ideas. A large number are perfectly reasonable, in a sea of perfectly reasonable businesses. Yet there wasn't the spark, the random good luck, the crystalising moment that thrust them to fame and fortune. Few would argue that a startup isn't a risky proposition. A gamble. So how do you improve your returns when you're dealing with tough odds?

The theory goes that, yes, a startup is a gamble, but if you follow the accepted patterns, you're minimising the risk. Better still, when all goes to plan, the rewards are enormous. You're tapping into a global space, hitting an emerging or unforeseen market and even a fraction of those billions of potential customers would still spell huge returns.

It's still a gamble though. Worse still, the gamble can only be called at the end of the game - and that's where I have the problem. Let's replay the list:

Have a great idea. Put together a dream team to implement it. Launch a cryptic website and sign up beta testers. Get some funding now you have a demo-able product. Preview the great concept with testers and improve the offering with agile development.

At this point, you're at least a year, possibly two down the line. You've baked the core of your idea into a solid code base, and have venture capital to get you through the launch. What? You've spent how much time, how much money and given away how much of your business to get here? Realistically that means that even the most successful launch is going to take a couple of years to just pay off the commitment you've made. That's a huge handicap to be working with. Given the dependency on a bit of luck, and the numbers involved, the vast majority of startups will just go nowhere, struggling for a painful year or two to pay back their investment. To get back to zero. Quite a few will just loose money and die.

So it's a gamble, with long odds and a big handicap. What's the alternative?

The answer lies in the one thing that you can do in a startup that will have the most consistent impact on results - execute fast. Not impressive, showy work, but simple, honest implementation. Get something out there that works, at least enough to demonstrate the idea as quickly as possible. That takes technical skill, a decent knowledge base and research - all of which are vital to get that idea to grow beyond a few nosey beta-testers. Good, quick execution cannot be distilled by following a multi-step plan you read on a business help site and it's the key thing that differentiates a working company from a bunch of guys with a clever idea.

So - drop the twelve step plan and demonstrate you can produce a functioning product (even if it's just a prototype) quickly when needed. It's fun going to the conferences, networking with other dreamers and charting out business plans, but until you can put something concrete in a customer's hands, you're wasting your time.

Fast, focussed execution doesn't beat the odds, but it allows you to play the game faster, more often. After all, you only need your number to come up once...

An Audience With...

2011-08-09T10:24:00.002Z

So you've got a Twitter account. And a blog. And a Facebook page. You're looking at Google Plus. You even manage to post the occasional video to YouTube.

But... isn't it all a bit one sided? Are you shouting into a void? There are a handful of comments that trickle in hours, days or even months after a posting. What about immediate, live, feedback? If you're trying to really engage people's attention, how do you make it instant?

Now I've got a theory. In ancient times, before the internet, before television, before radio and public address systems, even before newspapers and books an audience meant the group of people you were talking to at that moment. Talking live in front of a group is one of the great human skills, and one of the most natural ways to communicate an idea. Wouldn't it be great if you could hold an audience online?

Hold on - I write software - I can do that! Let's take the immediacy of Twitter, but turn it into a proper conversation that can be shared. Let's make it possible to arrange a chat with a colleague, friend or complete stranger that others can watch and enjoy. Let's add voting, so you know what your audience are thinking as you chat. Let's store it so you can come back to great conversations later, or link them from your blog, or send them via an email.

That's a cute idea. Let's call it Cewt Chat.

So far, this is barely a Beta test - let's call it alpha. Still, it shows the concept. If you enjoy chatting, want to engage with people, or want to elevate communication to performance art, this is a great way to do it.

Cewt. Over at http://cewt.co.

Sony PRS-505 E-Reader review

2009-08-03T15:25:00.002Z

It's rather strange to review a product that is more than likely to be obsolete in a few months, but in the case of the PRS-505 - an electronic book reader from the ubiquitous Sony - a late review is worthwhile. Let me explain.

Various companies have been worrying away at electronic books for a few years now. Initial devices used LCD displays, but couldn't quite get a form factor that held much appeal or made reading for long periods comfortable. Then, the invention of e-ink screens made it possible to produce a display that doesn't flicker, can be read in direct sunlight at any angle and that uses a fraction of the power of an equivalent LCD panel. Reading on e-ink is truly a revelation. There are however some disadvantages to e-ink, which I'll address later, but the most significant of these is that as a new and relatively niche technology it remains expensive. Regardless, the new screens have led to a new generation of e-readers that are slim, easy to carry and a pleasure to read.

A handful of companies have produced a range of readers, most of which have in common the 6-inch e-ink screen produced by Vizplex. The PRS-505 is one of these, introduced in 2008 and a steady if slightly slow seller for Sony. This has been due in part to the price, and in part to the difficulty in finding e-books. Then Amazon entered the market with the Kindle and everything changed. Other readers such as the Sony leave their owners to track down expensive and relatively bare book stores online using a normal PC and then plug in the reader to transfer files. In contrast the Kindle connects wirelessly to the Amazon store which is happy to sell many bestsellers at a good price (and a loss to Amazon). Amazon 'closed the loop' and made buying e-books as easy as... well, as easy as buying books from Amazon.

The Kindle remains an expensive device, but it's convenience is unrivalled and Amazon's domination of the book market has helped it sell tens, if not hundreds of thousands in a relatively short time. This has given a much needed kick to the e-reader manufacturers and suddenly players old and new are rushing new machines to market. At the same time, book stores are paying far more attention to e-books and the options for users are opening up.

Against this background, the two year old PRS-505 remains Sony's current model. It's almost certain that this Autumn will see it replaced with a new version that will take on some of the innovations that Amazon brought in with the Kindle. With it's rival offering more and Sony no doubt eyeing the future, the PRS-505 can now be bought for a bargain price - or at least a bargain price for an e-reader. With the cost of including the latest technology, the new models are pretty likely to come in at the older higher price and much the same price as the Kindle. So, if you're interested in picking up an electronic book, you have a choice. Pick one of the 'next generation' readers as they arrive in the shops, pay the price and hope that nothing else comes out to put it in shade. Or buy a last generation reader for half the cost and put up with it's limitations.

The PRS-505 can be bought at the moment for £150 from Play.com, or around $199 in the States. At this price, the only choice is the sleek silver model, but for a little more there are blue and red versions available. The price gets you the reader itself, a cable to connect it to your PC and charge it, a cover to protect it and a CD of free books.

Setting up is easy. Unpack the reader and clip it into the smart leather cover. Plug the USB cable into the bottom and the other end into your PC. The reader will charge itself over a few hours after which it will run to 8000 page turns before needing another charge. Note that the battery life is quoted in terms of page turns - the display consumes no power when it isn't updating, so it's only when you turn a page that the reader uses juice. In practice though, every menu selection equates to a page turn and the battery does loose a small amount of power each day. This results in the reader needing charging every three weeks or so - enough for a holiday or trip away.

The metal cased reader feels very well made, solid and a little heavy in the hand. In fact single handed reading is a little of a strain unless the book is supported. It has nearly twenty buttons on it's face, designed so that navigation always sits under a finger or thumb, whether the device is held in portrait or landscape orientation. There's an air of professionalism about it, somewhat reminiscent of the old Sony world radios.

In use, the display is easy to read for extended periods but has a couple of limitations. The first is that it is small - approximately half the size of an average paperback page. The second is that it is slow to refresh. Click on the 'next page' button and the screen goes momentarily black, then the new text emerges. Some have found this to be distracting, especially as you're turning pages more often than you would with a larger physical book. However, I've found it easy to get used to. Just remember that this is a book replacement, not a laptop or games console. The plethora of controls do their job admirably, and make a lot of sense in light of the slow screen refresh - the reader tries to let you chose the right operation with a single click, rather than having to zoom through menus as you would on a normal PC. Menus and lists line up with the numeric buttons on the right edge of the display and you can quickly make your way to the book you wish to read.

Books are available in a number of formats. EPub is emerging as the open standard and is particularly popular in Europe. Sony's own BeBB format is also supported by a number of bookshops. Both render well, and can be zoomed or rotated to read in landscape orientation. Adobe's PDF format is also supported but makes for a disaster when reading, contrary to expectations. The issue here is that PDF expects a fixed page size and specifies exactly where the text should go on that page. When faced with a reader that has a small screen, there's simply not the information needed to re-flow the text correctly, so either all formatting goes out of the window, or the reader struggles to display a large page on a small screen. This is a problem common to all small format e-readers at present and should be kept in mind before you buy one expecting to read company documents on the move. PDF format is also used for some commercial books, but really is a last choice if there are no other options.

The reader can also handle both photos and music, though both show the device's limitations. Photos are low resolution and grey. Music can be played as you read - if you enjoy the distraction - but dramatically reduces the battery life from weeks to hours. There are slots for memory cards (both SD and MemoryStick format), a headphone and charger socket and little else to distract from the single task the reader is designed for.

As a technological device, the reader is impressive. It's focused and well constructed. It will hold hundreds of books and runs for weeks between charges. As an alternative to paper based reading, it suffers from a small screen and high price. In that respect though it's no different from any of the current generation of devices. If you can see a strong benefit in making a large number of books fit into a small space, it makes a lot of sense. Obsessive holiday readers and those of us who rely on technical libraries will find it easier to justify the purchase.

The next generation of machines are likely to address some of the limitations of the PRS-505, though prices are likely to remain high whilst the technology matures. It would be nice to see a device with a larger screen taking up more of the front surface - the Sony is typical of the current generation, with nearly half of the available reading area lost to the controls and casing. I don't think we've yet seen the ideal control arrangement either, with manufacturers struggling between constrictive minimalism and messy abundance. Some commentators are keen to see greater interactivity, but I feel this is just wishing that you could get the functionality of a laptop into a device that is a tiny fraction of the size. The great strength of e-readers is that they have chosen a different set of compromises to those needed in a laptop, which makes them able to perform their task so much better than the idealised 'universal gizmo' many would wish for.

My view is that it's going to be a while before e-readers truly hit their stride, and that it'll be a couple of generations before the technology fully matures and is consumer ready. Whether you buy the tail of the last generation, or the cusp of the next, you're going to end up with a device that will be improved on again in the next twelve to eighteen months. The current progress is steady rather than spectacular, so saving money is probably a better bet than grasping for the bleeding edge. Buy the PRS-505? I did.

Now where did I put my book?

Filtering Replies Out Of Twitter Badge

2009-06-02T21:00:00.011Z

On my other other blog I've got deeper into Web 2.0 and added the standard Twitter Badge to show my latest Tweets. It's a handy bridge between quick updates and more detailed posts. However, one niggle is that the feed also shows @replies to other people, which make very little sense outside of Twitter itself.

I've looked in various places on the web for a solution and not found one that is easy to use. So, in the spirit of someone who's meant to earn his living from writing software, I've written my own. The important features are:

It uses the standard Twitter Badge code and API

It's completely free standing - you're not relying on a third party website

It's easy(ish) to install. Just a small edit to the badge code

To install it (assuming you've already added the Twitter badge), open your Blogger layout from the Dashboard, and click 'edit'. Take a look and you should see that the code ends in two <script>..</script> tags. The first looks like this:


<script src="http://twitter.com/javascripts/blogger.js" type="text/javascript"></script>

After that and before the following <script> tag, add the following text (you can copy and paste it from the text box) :

<script type="text/javascript">function filterCallback( twitter_json ) {var result = [];for(var index in twitter_json) {if(twitter_json[index].in_reply_to_user_id == null) {result[result.length] = twitter_json[index];}if( result.length==5 ) break;}twitterCallback2(result);}</script>

Finally edit the last <script> tag and replace the text 'twitterCallback2' with 'filterCallback'.

The filter script takes the Tweets from Twitter and hides any that are @replies to other users. That means that if you have your badge set to display five tweets, and your most recent tweet was a @reply, then only four will be shown on Blogger. To get around that you can ask Twitter to send a higher number of tweets before replies are filtered out. Look for the part of the final <script> tag that reads count=5 (if your badge normally displays five tweets), and change the number to a greater value. Try values of 8 or 10.

The maximum number of filtered Tweets shown is set to five in the new script, so if you want to set that to a different value, change the bit that says result.length==5, replacing '5' with your new count.

This could be the basis of even more complex filtering - you could look for #tags or particular words to be skipped or included in your Blogger gadget.

Pixel Qi, ebooks and convergence

2009-05-29T15:56:00.001Z

Pixel Qi are making a lot of news at the moment with the launch of their new display technology. First concieved for the OLPC laptop, the display is a clever update of existing LCD devices. With its backlight on, it behaves just like any other LCD panel - bright, colourful and power hungry. Turn the backlight off though, and the display consumes much less power, displaying detailed black and white images that can be seen in full daylight.

Immediately bound for netbooks that can take advantage of low power and don't mind the relatively small size of initial screens, the new display promises much. However, it's not the universal panacea and the ghosts of convergence are rattling their chains. Yes, everyone's getting excited that laptops, netbooks and ebooks are all going to merge into one uber-gizmo that will allow you to work, travel, read and browse the internet seamlessly, painlessly and cheaply.

Wrong.

Convergence is a misleading distraction when it comes to technology. The dream of a swiss army knife device that solves everyone's needs in one easy package is alluring. Time and time again innovators attept to deliver it. It doesn't help that the public clamour for it too. Why, if everyone wants it, is it such a bad idea then?

The problem is that there is a disconnect between the ideal and the reality. In our imaginations, the converged device does everything just as well, if not better than our existing machines. However the vast majority of users (and quite a lot of technologists) don't really understand what it is that makes their devices so successful. When it comes to reality, the converged device must make compromise. The effect may be obvious (low quality lenses in camera phones) or it might be more subtle (the paucity of interaction on iPhone games), but it lowers the bar, and largely weakens the user experience rather than enhancing it.

When it comes to ebooks the ideal is a device that is small, light and virtually indestructable. It should rarely need charging and be fantastically simple to operate. To really gain acceptance, it should also be as cheap as possible. It should be 'invisible technolgy' - something that doesn't get in the way of the job in hand. Carry it anywhere or just pick it up and read. These goals are all achievable if you design for them - and the future of ebooks is likely to see devices refining these ideals.

With a netbook, the goal is different. We're talking about a device that gives you convenient, portable access to the stuff you do at work, the documents on your PC and the internet as a whole. Music, video, photographs, games, email, instant messaging.. the list goes on. This is a class of PC, albeit designed to be just powerful enough to do the mundane things and no more. Those specifications allow us to design a device that is cheap, portable and with reasonable battery life.

It's very tempting to converge the two. Why not keep evolving the netbook until it's battery life is long enough and it's screen good enough to let it function as an ebook? Why not add a keyboard to an ebook and let it browse the net and take notes as you go? The answer is that from one direction you get an ebook that is expensive, over complicated and just physically the wrong format for easy reading. From the other, you get a netbook that is underpowered, awkward to use and lasts hours, not days on a charge.

Rather than convergence, divergence will provide us with revolutionary new devices. Ebooks will dip below $100 in a very short time, and will be given away by bookshops eager to sell books and subscriptions to newspapers and periodicals. Netbooks will become powerful enough to be used as a desktop replacement in many offices and media centres in homes. Attempting to bring the two together will rob the end users of two very useful tools for the false promise of a single device that doesn't quite do two jobs well.

None of this is bad news for Pixel Qi - their displays could be used in both netbooks and ebooks. In the former, the backlight will usually be on as the user browses, watches video and edits documents. In the latter, it'll usually be off as the machine eaks out battery life, slumbering whilst the user reads largely static text. In both cases though it's the rest of the device that will make for an interesting product, not the screen technology.

Wii System Menu 4.0

2009-03-26T09:10:00.002Z

Many commentators in the Gaming world have suggested that Nintendo are loosing the console battle from in front by not effectively competing with the hugely successful XBox Live and Playstation Network services. The implication is that whilst the Wii has sold in stellar numbers, the next generation of gaming experience belongs in particular to the XBox. Eventually the underpowered and easily dismissed Wii will be supplanted by the technically superior XBox and even the underperforming Sony machine.

However, things could pan out somewhat differently. The point to realise here is that so long as your console is online and capable of receiving updates the gaming experience can be changed at the drop of a hat. The perceived weakness of the Wii in not having an internal hard drive to store games has been effectively wiped out in a single (and cheap) upgrade to the new Wii System Menu 4.0. Now gamers can download games directly to high capacity (up to 32Gb) SD cards. Who needs expensive (and occasionally unreliable) blue-ray and hard drives when you can buy an SD card for pocket money?

The other observation has to be that rarely in the software world does the first iteration of a concept win the race. The Apple Mac user interface was ground breaking, but Windows very nearly obliterated the competition. Wordperfect and Lotus Notes owned the market, but Microsoft Office consigned the innovators to history. In the case of XBox Live, there is very little advantage to an early technical lead when your competitor has more machines sitting in people's rooms. Nintendo has the luxury of a huge user base and plenty of money in the coffers. They can observe what works well and more importantly (*cough* Playstation Home) what doesn't.

It's by no means a given that the Wii will continue to succeed. Third party support is most politely described as patchy, and there is a limit to the number of different input devices you can force people to buy. However, Nintendo has lots of room to experiment and highlights like Mario Kart Wii show that online gaming works just as well on the little white box as on more expensive machines. The perfect combination of online payments and software upgrades puts the Wii on equal footing with any of its competitors when it comes to involving the users, and evolving the experience.

All this can be said whilst ignoring the hyperbole friendly OnLiveconcept (which has led to some hugely optimistic futurology in the blogsphere - deserving of a separate debunking post). If games don't need to run on the box next to your television, even the hardware advantage the other consoles have over the Wii will be gone.

Now who wants to call the race?

New Tech Excites Journalists

2009-03-19T13:46:00.002Z

Bruce On Games returns to a favourite meme - that computer games are overtaking film as a medium.

the Byron review favoured BBFC instead. Yet the F in BBFC stands for Film, just as the F in BAFTA stands for Film. And games are nothing like film. Film is an old medium in decline. We are the best medium with the technical advantages of interactivity, connectivity and non linearity. We are the future and they are the past. We really don’t need their institutions on our back and we have the power to create and use our own.

I would imagine that if you trawled the archives from around 100 years ago, you could find newspaper articles proclaiming that books would be made redundant by film. Plus ça change..

The problem here is taking a personal experience and extrapolating it (with the aid of natural sales trends) into a vision of the future. Bruce finds games more involving than films and has a stack of features such as non linearity to justify his personal preference. However, the very features he holds high are far from a replacement for the experience provided by the old medium. Non linearity makes telling a compelling story very, very much more difficult; connectivity reduces a finely tuned experience to a clamour of 'me too' mediocrity and so on.

That's not to say that games aren't deeply involving, or that the possibilities offered by technology aren't exciting. It is certain though that they're not the stake in the heart of film either.

Should games have their own institutions? Possibly, but they should not loose sight of the fact that the old media remain far more sophisticated and evolved than games have yet become. We're still seeing baby steps in connectivity and interactivity which may be technically very impressive, but are still remarkably shallow experiences. If you doubt that, take a look at games little more than a year old and witness how outdated they can feel.

Psion Must Be Kicking Themselves

2008-09-16T15:53:00.000Z

As the Eee-PC destroys all innovation in the laptop market, drowning us in a deluge of mee-too imitations, Psion must be wondering what all the fuss is about.

Nearly ten years ago The Psion Series Seven did all that the Eee-PC now promises. It included a stunning set of useful applications, instant on computing and a ingeniously designed case that showed that clever design isn't just Apple-skin deep. Not only that, but it ran for nearly a full day on a set of batteries, rather than the Eee's paltry few hours. Psion had been supplying hand-held computers since the mid eighties, and the Series Seven (and Series Five before it) showed all they had learned to great effect.

Sadly at the time the fad of the day was the Palm. A glorified address book with an innovative but deeply constrictive user interface, the Palm was seen as the future of truly portable computing. Compared to the rich application set supplied with the Psion, which could interoperate with the standard Office toolset and provide document viewing and editing on the move, the Palm was a retrograde step. Unfortunately the Palm became a fashion item before the days of the iPod and iPhone, due in no small part to the ease with which new users could start doing trivial tasks.

The Palm spelt the end of the Psion, which would never see the days when greater digital literacy would lead users to the niche it had defined for itself. Computer literate users now want small cheap computers which provide the standard set of applications usually encountered on their desktops, but without the bulk and expense of a powerful laptop. It's just a pity Psion are no longer in a position to supply their hard won expertise and well designed products.

If I Tell You...

2008-08-27T15:37:00.001Z

A friend spent a good few years of his life and a seven figure sum on a project that ultimately died, taking his grand vision and start-up with it. The signs that his idea was doomed were there from the start, but it is the role of the visionary to ignore their detractors. It's also the role of a friend not to labour the point that you think their idea is bloody stupid.

Perhaps the key indicator of ultimate failure was the secrecy involved. The project survived for five years off VC investment, government grants and low wages for keen staff. During that time I would periodically bump into the guy behind it all and ask him how it was going. Then I would ask what it was that they were doing. The second question always got the same vague answer. As a 'software guy' I could (theoretically) implement anything that is sufficiently well described to me. So, when asked what it was my friend and his band of followers were doing, he would carefully explain that he couldn't possibly tell me. If he did, I might steal his idea.

That's the warning sign. Watch Dragon's Den on BBC television (if you can stand the patronising narrative) and you'll see the same thing. If all you have is an idea, however smart it might be, you don't have anything. The thing that makes an idea succeed is putting in the huge amount of effort required to implement it well and get it out there. If I, as a single developer, could threaten to re-implement your idea that you have spent years working on with a team of skilled people, then either the idea is not strong enough, or your team is not productive enough. Either way, failure looms on the horizon.

A successful product comes from the understanding that not only the core idea must be sound, but that actually implementing it is the bigger challenge. Putting an idea into practise is by far the biggest investment in a new product, and the 'cost' that prevents others from just stealing it. If the only value of the finished product is the original idea, then either the initial concept was not sophisticated enough, the development was not thorough enough, or you'd better have some searingly hot IP lawyers sharing your venture capital.

The Font of All Knol

2008-08-21T09:27:00.009Z

Google's Wikipedia beater, Knol has lurched into the public view with all the finesse of Google's Beta releases. It's harsh to complain about the relatively few small bugs in this otherwise very well presented web service. Like most Google products, Knol is slick and professional. The handful of distressingly disruptive bugs that existed at launch have been quickly addressed. Knol appears to be another successful Google 'experiment'.

However, in directly engaging users to produce content I suspect Google have come up against something slightly unexpected. Their algorithmic focus on efficient search, indexing and user profiling has perhaps left them with a blind spot when it comes to social interactions. Certainly, when user content is the be-all and end-all of a web site, you're no longer in the realms of inert data. Instead you're in the land of social software and at the mercy of the annoying behaviour of a crowd of untamed users. In short, a slick service can still suffer from big useability problems.

From the idealistic point of view of providing useful information, Knol suffers from organisational issues. Individual Knols (Google's term for an article about a specific subject) exist in isolation. Without subject areas, categories and a convenient way to interlink related Knols, articles have to either attempt to cover a subject in one long, breathless page or take small bite-size chunks and leave the user to make the connections. Given that many authors can tackle a subject area, the task of connecting related material is as big a research effort as writing a Knol itself. At present the user is left with that research to do for themselves. The end effect is that rather than providing easy access to information, Knol challenges the user to make sense of the many clamouring voices.

In making the reader work for their information, Knol doesn't provide easy access to knowledge, nor does it welcome the casual visitor with a browsing experience that might draw them into the site. Coming across a Knol doesn't encourage the reader to explore further, stumble upon related articles or compare with alternative viewpoints. The breadth of a subject area remains hidden and the visitor remains unengaged. In this case the sum of all Knol is less than the sum of it's parts.

The other issue that Knol appears to be dealing with is the far more base desires of the (content generating) public. Knol provides a public stage that has some authority just through association with Google. The theory goes that a Knol will have some priority in Google searches as they attempt to keep browsing users within the Googlesphere. Browsing users equate to all important revenue as Google can serve up adverts so long as those users remain on Google's own web pages. Knol authors can also benefit from ad revenue through the Adsense programme.

Unfortunately, whilst Google struggles to 'do no evil', their users are not necessarily so morally inclined. The profit and self-promotion that Knol offers encourages people to try to game the system. As a result, the first few weeks of Knol's public launch have seen numerous articles ripped wholesale from other sources in an attempt to gain ad revenue. Popular search terms have been quickly Knolled (if you forgive the verbification), largely by copying Wikipedia and IMDB articles verbatim. Other entrepreneurially minded individuals have mass produced brief product reviews that conveniently link to affiliate sites selling those products. Knol's position within the Google universe encourages spam.

No one could claim that this sort of misuse has come as a surprise to Google. From launch, Knol was able to identify similar content that existed elsewhere on the web. It also provides a comprehensive system for complaining about content. What does seem to have surprised the Knol team is the sheer volume of articles that are less than altruistic in sharing the author's knowledge. Policing Knols appears to be giving them quite a challenge. At the time of writing it seems that Knols abusing the system can remain in place for quite some time before they are removed.

Perhaps most significant is the fact that Google's main search index of their own site does not appear to have been updated in a couple of weeks. Search Google for a topic that is covered in a Knol and you're fairly unlikely to find it. Search on the Knol site itself and you'll have more luck, but it seems that Google have lost a little confidence in their project for the moment. Reading between the lines you might assume that until they are confident that Knols are going to be reliably informative rather than spam, Google are delaying bringing Knols to the attention of the wider searching public.

I have to say I like the site. I've experimented in writing a couple of articles (you can read my enthusiasm for Canon's 450D here). As an author the experience is good, and Knols are very nicely presented to the reader. When Google overcome the current issues - and I have no doubt that they will do so - Wikipedia will have a strong competitor. Right now though the ubiquitous Beta status of Google's product is perhaps more deserved than usual, and not for the quality of the software beneath the site.

From the Journals of 'I Told You So'

2008-08-12T15:20:00.002Z

Ohhh look. Over eighteen months ago I blogged about the BBC's experience with mobile content. Maybe I didn't hammer home the important fact that they had pretty well established that no-one watched it. Mobile content excites geeks ('look what I can do') more than it excites the great unwashed public ('why would I want to do that?').

Now, apparently to a few peoples' great surprise, Korea has come up with some figures that show that... mobile content has statistically (as well as commercially) insignificant takeup. Here's The Register's take on it

Korea has long been envied by Western tech firms as the mobile platforms there have apparently penetrated deep into cell phone owners' lives. More than just text messaging (which we have grown to rely on in the UK, but which also resists any attempts at extracting more revenue per interaction), the Koreans play games, socialise and consume content at phenomenal rates. You can smell the profit from half way around the globe.

So when we find out that Koreans don't watch miniscule video on the move, can we finally lay to rest the idea of 'mobile device as (expensive) TV'? It didn't work in the 70's with Sinclair's 'pocket' TV. It was no better in the 80's with fat TV-watches. It didn't work in the 90's with Sega's Lynx. Note that all of those failures had the advantage over more modern iterations of using freely available broadcasts. Their failure didn't hinge on prohibitive subscription models and digital rights issues.

More recently Sony's PSP has died on it's ass when it comes to mobile content and Nintendo have the good sense to leave video playback as a (very) optional extra for the DS. The only glimmer of success comes from Apples' camp where video rides the shirt tails of something people actually want - a phone and a music player. In practise though, I'd lay good money on the video functions being used a handful of times "ooh, look what my phone can do" - and then being forgotten in the vast majority of cases. As a user experience, it just doesn't work.

RIP mobile video. Please.

Xerox - Research And Development

2008-05-01T14:47:00.002Z

The BBC recently visited Xerox's famous Palo Alto Research Centre and reported on the innovations being pioneered there.

A few years ago I worked for Xerox in Cambridge, when they still had a research centre here. The breadth of research and willingness to explore tangential ideas was deeply impressive. On the other hand, the ability to move blue skies ideas towards commercial reality was almost entirely absent. I can only assume that's why the office was eventually shut down.

At the time though my exposure to the rest of the company suggested that the problem was endemic. If I remember rightly the culture elevated star researchers to high status, but provided them with virtually no expertise on the all important development side. Simultaneously it made it harder for other employees to put forward their own ideas as they stood in the shadows of the stars. The result was that a small number of people would wander aimlessly around producing truly exciting demos that would limp towards the real world with all the urgency of a decapitated sloth. Indeed, in the couple of decades that Xerox had a presence in Cambridge, I believe they produced on average approximately one patentable idea a year. For a large office devoted to research that seems to be a pitifully small return.

The issue of course is that great intellectual talent does not automatically infer great business talent, nor even the ability to manage the team of scientists and engineers necessary to take an idea beyond the proof of concept stage. At the time it seemed somewhat ironic that Xerox had opened a facility in Cambridge which has exemplified this sad fact time after time in exciting but doomed high tech startups.

Since then I've met with more than one company who have been brimful of confidence that their great idea is so mind-bogglingly clever that it will make them all very rich and famous. It shouldn't come as a suprise that the ones who do well are the ones who have talent for making simple ideas work, or complex ideas work simply.

Javascript

2008-04-28T09:29:00.003Z

Pulled into battling the multi-standardised demons of Javascript for my latest client, I find myself referring to the usual haunts for solutions to problems that shouldn't exist. Javascript started as a simple language for supporting lightweight interactivity on a web page - clicks and prompts and updating the document in response. Unfortunately it was caught in the browser wars and since has become the cornerstone of application grade user interfaces never dreampt of when it was conceived.

The suprising thing is that almost without exception, I haven't found a single Javascript resource that always gets the right answers. Take the simple problem of locating the mouse. Search on Google for mouse position javascript and the first page of links all take you to solutions that will fail under one circumstance or another.

Bad examples include:

Retrieving mouse-x and mouse-y in two separate functions, both of which have to traverse the same document model.
Relying on browser detect functions to decide how to access properties
Ignoring browser types completely and only functioning in IE
Only solving the first half of the problem - finding the mouse on the page. If you don't know where your document element is, you have no context for the rest of your code.
Failing to handle the case where the document has been scrolled
Failing to handle the case where your event occurs within an IFRAME

Of all of these, Quirksmode only falls into the last trap. If you're handling an event from an IFRAME, you should be careful to account for the scroll offset of the frame itself and not the document that the event handler resides in. This function does exactly that:


function mousePos(evt, relativeTo) {
  var xPos, yPos;
  if (evt.pageX || evt.pageY)     {
    // Netscape
    xPos = evt.pageX;
    yPos = evt.pageY;
  }
  else if (evt.clientX || evt.clientY)    {
    // Probably IE
    var docn = evt.srcElement ? evt.srcElement.ownerDocument : evt.target.ownerDocument;
          
    xPos = evt.clientX + docn.body.scrollLeft + docn.documentElement.scrollLeft;
    yPos = evt.clientY + docn.body.scrollTop + docn.documentElement.scrollTop;
  }
    
  var locn = {x: xPos, y: yPos};
  if( relativeTo ) subtract( locn, relativeTo );
  return locn;
}

function sutract( point1, point2 ) {
  point1.x = point1.x - point2.x;
  point1.y = point1.y - point2.y;
}

Note that the function takes the event object and an optional parameter for the position of the element that you want to find the mouse location relative to.

Now, all you need to know is where that element is relative to the document it resides in, and you can get a fix on your mouse event. Here's the code, which takes account of elements that can be scrolled.


function findPosition( obj ) {
  var xPos = yPos = 0;
    
  var start = obj;
  while (obj && obj.offsetParent) { 
    xPos += obj.offsetLeft;
    yPos += obj.offsetTop; 
    // Don't include scroll offsets for myself, or my parent document
    if( obj != start && obj != obj.ownerDocument.body && obj != obj.ownerDocument.documentElement ) { 
      xPos -= obj.scrollLeft;
      yPos -= obj.scrollTop;
    }
    obj = obj.offsetParent; 
  } 
  
  return {x: xPos, y: yPos};
}

AJAX and the Presentation Layer

2007-08-03T09:55:00.001Z

When Java servlets were first deployed, the answer to all our online development woes seemed to be at hand. Of course, that was a long time ago and the question has become far more complex. In turn so have the answers. Servlets, JSP, J2EE Containers, Struts, Tiles, Spring, Hibernate - frameworks and APIs have evolved to make development more manageable and efficient.

One of the key steps in that evolution was away from the question of 'how do I provide a service' and on to the question of 'how do I provide a maintainable service'. The initial toolsets allowed complex online services to be developed with an amorphous mass of spaghetti code. Even the most diligent of architects found it hard to structure the components of such systems so that future fixes, updates and additions could be made with a minimum of disruption. As the problem became better understood, frameworks began to divide the software into layers, then tiers. Crucially, persistence, business logic and presentation were separated so that each could be worked or reworked independently of the other.

The importance of this can't be over emphasised (that's my excuse for the naff history lesson, for those of you who aren't paying attention). In my time I've come across some severely broken software and some consequently severely broken developers. Usually the outcome of some unsuccessful attempt at providing a rich online service is that the tools and APIs used are declared to be unfit for purpose. The result can be seen in the enthusiastic stampede from one platform to the next (hello, .NET, you benefited from this). The thinking is "that was way too hard, maybe this new tool will make it easier". Unless that tool encourages or even enforces stricter architectural separation nothing actually changes. Online services are hard to design, hard to implement and can be impossible to maintain.

Things were just beginning to settle down when AJAX came along. It hasn't made things any easier.

I'm left feeling that the rush to provide technical demos, helpful tutorials and wizzy Web 2.0 services is going to cost some developers dearly. New languages are already benefiting from the death of poorly designed web services - surely ditching Java and .NET in favour of the latest flavour du jour will solve all your problems? No, of course not.

Let Us Count The Ways

2007-07-25T09:24:00.000Z

Lately I've been using a Genealogy program to fill in the details of our family tree. For some time my Uncle has been diligently researching our roots and has painstakingly drawn out the chart by hand. Unfortunately as he adds new people and updates the information he has about existing family members, he has to start again and lay out each and every person. With just short of four hundred people to keep track of, it's a mammoth task - and a mammoth chart, around six metres long.

So for his 70th birthday I committed to putting all the information into a proper database so that it could be charted and updated more easily. The software in question is regarded as one of the best in the field. Yet, as a developer, it leaves me with that familiar itch; that feeling of "I could do a lot better than this".

The problem is that software like this is sufficiently complex that writing it from scratch to address an existing market is very hard to justify. The open source acolytes don't have the stamina to solve the problem for a relatively small audience and the commercial developers don't want to take the financial risk. As a consequence, there are many applications that are on version 6, 7, 8 and beyond that have evolved into sophisticated beasts that are loved by their users at the same time as being significantly flawed.

The flaws in this particular application range from simple little things to deep rooted architectural problems. At the simple end there are dialogs for entering a single value that don't put the focus on the input box. At the architectural end there is a database that is used to generate charts that almost inevitably need adjusting and moving by hand to tidy up the layout. Yet the charts are divorced from the database when they are generated, so adding or updating a record means that the chart has to be generated again and the painful process of fixing the layout has to be repeated. There's a help system that is so intrusive that it gets completely ignored - and when it is referred to it falls into the trap of describing what dialogs do rather than what they mean and how they can help the user achieve a task. There are technical bugs, such as an undo button in the charting application that doesn't undo all of the user's actions, skipping certain functions and leaving a mess on screen. Then there are signs of an application that has evolved too far, with different activities using different mechanisms to set up and control the software's output. The same or similar settings have to be applied in many different places in an attempt to get consistent reports.

Don't get me wrong - the software is clever and allows the job to be done. However, it could be done so much better with a clean and consistent user interface, a better integrated and more efficient database and a carefully considered design for the way charts and reports are generated. None of these things are too difficult - though many developers have some difficulty with user interface design for humans, and most developers find it hard to step back from a 'solved problem' and come up with a clean slate design that doesn't fall into the same architectural traps.

It would be wonderful to have the free time and finances to sit down and write a better application. Realistically though, it's not going to happen, so I'll keep on using this software whilst trying to avoid similar mistakes when developing code for my clients.

Good Old Beeb

2007-01-31T12:56:00.000Z

The BBC are going ahead with video downloads of popular programmes. I have to say I love this. In the US, tech and media companies tie themselves up in knots trying to lock their customers into paying again and again for 'protected' content. Meanwhile in quaint old Britain, our national broadcaster gets on and does the job. We've been able to listen online to radio shows for a few years now, and here comes video.

It's hard not to be impressed. Whilst start ups fiddle around with business models trying to charge for content on mobile phones, the BBC is putting on its "Been there, done that" t-shirt. In the last couple of years they worked hard on mobile content, and demonstrated that outside of over-excited funding presentations, take-up is pretty poor. Heavily promoted to just the right young, tech-savvy crowd, mobile-friendly Doctor Who content saw audiences that were a miserable fraction of the TV viewing figures. Really miserable.

Why should this be? Well, my guess is that the occasions when you've got long enough to take in even a short bit of video, you're probably sitting at home or at work within reach of a TV or PC. When you're actually away from either of these sources, you're probably not in a position to sit down, chill out and watch a microscopic abbreviation of something that works much better in a different format. Outside of London, few people have the 'privilege' of sitting for ages on public transport, and outside of VCs and tech journalists, not many of us spend that long on 'planes either. So - nowhere to watch, nothing to see.

And in the mean time, the BBC has learnt a lot of lessons and continues to push the envelope, providing that elusive commodity, 'popular content' and trying out technology and formats that will obsess Slashdot readers in months and years to come. That is, when the tech firms have got some funding and worked out how to protect their media.

The iPhone

2007-01-19T14:31:00.000Z

It seems to me that the cleverest thing about the cult-of-apple is that they have managed to present a series of patent land-grabs as good news for the consumer.

The most important thing about the iPhone is not the slick user interface or the geek friendly (but user-harmful) convergence. It's the fact that the damn thing is wrapped in patents. Indeed, if I'm interpreting Apple's business model correctly, the whole aim of developing new products like this is to find patentable concepts, rather than workmanlike products. Whether it takes off or not (I'm guessing yes in the States, no in Europe), the end result will be a series of tech-lookalikes that can't get near the clever user experience. So, what may be a genuine innovation in user interface will be restricted to iPhones only, with everyone else stuck with second rate emulations that will mainly miss the point. Imagine if Xerox had successfully patented the idea of the mouse, leaving the rest of us with unholy almost-mice: trackballs, knee pointers and such.

It's a product of the current business climate and patent friendly culture, so I can hardly object to it on the grounds that it's designed to hobble a particular technical niche to Apple. That's how you do business today.

What's This?

2007-01-19T14:08:00.000Z

Just a place to post reminders to myself about technical itches that need scratching, thoughts on current issues and other detritus.