Damian Steer from University of Bristol and Alex Dutton from University of Oxford taking this session.

Originally mobile phones were just ‘phones’! Ten years ago, phones had started to support other functions – SMS, WAP, WML – but very limited interactions – ‘web fidelity’ very low. Things continued to develop – ‘smartphones’, Blackberries etc. Supported better interactions – java becomes standard development platform for these devices. Could read email on these devices, and to some extent could browse the web – e.g. using Opera Mini – although a lot of ‘backroom’ stuff to make this work – but it was browsing the real web.

Then the iPhone appears. Damian says generally nothing new about the iPhone (except possibly ‘gestures’) – but it was really really usable, and had a really good web browser… (doesn’t mention the unlimited data plan – would have thought this was also a factor?)

iPhone has much higher usage on webpages than it’s marketshare – that is iPhone users browse the web disproportionately more than users with other devices.

Mobile Safari on iPhone built on WebKit. WebKit also used for other platforms. There is no IE6 for mobile devices – that is terrible browser that is very widely used and needs to be catered for – nice to have clean slate.

‘One Web’ concept – don’t do something special for mobile devices.

Traditional usage absolutely dwarfs mobile usage – so think very carefully about when users going to using or need mobile access – focus for what you deliver to mobiles on these scenarios.

Why do ‘web applications’ rather than platform apps (e.g. iPhone app)? Don’t you lose lots of functionality – can’t access stuff like motion sensors in phones etc? Damian says very fragmented market – if you develop web apps you can reach wider range of devices. Also only a few things you can’t access via web apps as opposed to a platform based app – so unless you really need to use motion sensor build a web app instead. Note you can access geo-location information within the browser, and provide ‘offline’ functionality.

Some stuff you can’t do so well in web apps – gestures (supported at a low level, and can be fiddly).

So there are some things native apps are better – e.g. gaming, 3D etc. Also example of Guardian iPhone app – could all be done as web app, but native app is very careful with resources and a key factor – you could sell the app easily and make revenue via the app store!

Damian now showing some web app stuff developed at Bristol – example of  ‘when is next bus’ – can locate and find next bus times – all done as web app. Can contact Mike Jones (@mrj1971) for more information…

Now moving to Alex Dutton from Oxford. JISC funded development of mobile services under Institutional Innovation programme. Developed ‘Mobile Oxford’ – went for easy wins – readily available data – location etc. Getting University contact information (originally screenscraped, but can now do via API and LDAP), search library catalogue via z39.50 etc. Also working with Oxford Brookes to add their data.

Mobile Oxford search box – uses regular expressions to try to work out what type of search is being done and return relevant information – e.g. look for ISBNs, Bus stop numbers etc.

Every page in Mobile Oxford can be returned as HTML, XML, JSON and YAML – e.g. add “?format=xml” to URL, or also support content negotation. So others could build services based on the Mobile Oxford data.

Other work in progress – new developments coming – e.g. Sakai (VLE) tools…

Oxford do quite a bit of ‘client sniffing’ – i.e. work out what client is accessing the page, and return appropriate content (so e.g. if javascript not supported, don’t use it etc.)

Oxford releasing the work they have done as an Open Source project ‘Molly’ – http://mollyproject.sourceforge.net/ – encourage other institutions to join in, use and contribute to Molly.

Finally Damian mentioning that more developments coming to web apps – Augmented Reality perhaps a notable development that is coming.