Robin is an innovator and entrepreneur. 1st prize winner in the Calling All Innovators competition 2009 in the Internet Innovation category for TechBuzz widget which Robin wrote. He co-founded Mippin.com (then called Mobizines) in 2004 which won Forum Nokia developer of the year for 2006/7. He founded a new startup, Alibro Ltd in Oct 2009, as a vehicle to further EyeMags.com. In 2010 Robin was co-founder of Promoht Ltd, a platform with a promotional content and targetting focus created for use by mobile operators world wide- a sensible alternative to mobile advertising. This system is live in 6 countries for high end phones.
robin.jewsbury | 05 October, 2008 21:12
I wrote an post in March saying native were on the downturn and webapps were the answer to everything. Since then several things have happened:
The reality is now that we can no longer simply argue about webapps vs native apps. Now there is a sea-change going on where I believe the industry is positioning itself to make the Internet truely work well with the Mobile computing platforms. As always different people are solving the issues in different ways. Google believe in thin client/cloud computing but have realised that this does not work when it comes to mobile computing. So Gears now enables the browser to become fatter and more connected to it's host platform. The great thing is they're enabling easy access to upload mutliple files (eg pictures from a phone to a server) and local phone functionality such a location. Because they need ubiquity for this technology they also realised the only easy approach was to launch their own Chrome browser - and furthermore they have made it clear this will eventually run on the Android mobile platform.
Meanwhile Nokia have been working on the WRT technology for some time. WRT allows a webapps to be installed locally on the phone, but for me its just another approach to fattening the browser on the phone. For me the release of the BBC iplayer app is a definitive moment for this technology. Local installation is ideal where user want to repeatedly come back to the same thing. The important point is that this is pure web technology enabled as an installed application and found from a phone menu. Also there have been reports of enabling Javascript access to core phone functionality (eg GPS location) just like Gears is doing.
What Apple is doing is interesting. I think they are playing with us poor developers. They've made us learn their ObjectiveC platform for Native apps which was released in June 2008. Meanwhile they always had their DashCode functionality for the Mac which is almost identical to Nokia's WRT apart from the fact Dashcode can contain native code. Will we have to wait until June 2009 for DashCode for iPhone? My cynical brain is telling me they did not want too many apps this year so they made the programming hard - by making the programming hard it also increases the quality of the apps produced (only dedicated developers can produced them). Dashcode for iPhone will appear when Steve Jobs thinks it will help him most.
Its unclear what Microsoft are considering under the name Windows-Cloud but my guess is that they will present something similar to Google Gears and/or WRT/DashCode. I suspect there will be a native element to Windows-Cloud, just like Apple, Microsoft need to exert some control and make their developer's more exclusive.
Microsoft are about to create a new technology and with it we will get all the hype surrounding it. The reality is they will make it exclusive to Microsoft with connections to surface computing, Silverlight and Popfly, but it will essentially (I think), amount to the same features of parts Google, Nokia and Apple are already developing. Its just a shame that we may end up with 4 different approaches to solving the same problem. I can, of course, see possibility for interworking and convergence. In particular Dashcode without native code and WRT should interwork and Google Gears could interwork using a Javascript abstraction layer. The Microsoft version may not interwork but may be they would be happy with that.
Meanwhile I still believe in webapps but I may slightly change my definition of a webapp and include DashCode(non native)/WRT apps.... they are almost the same thing. I see the undoubted success of the Apples native iPhone apps an aberration. Apple will cleverly introduce Dashcode apps at some stage - the consumer will not spot it, and most developers will just think Apple made their life easier - not realising they could have done that this June.
Perhaps we're all in Cloud Cuckoo Land until its clear what Microsoft, Apple, Google and Nokia are really doing with all these technologies.
Commentsrobin.jewsbury | 06/10/2008, 16:06
I agree. There is a reason why each technology exists. The BBC S60 iPlayer app is ideal as an app because its so good you would want to come back to it again and again. But this installable app is based on webapp technology so the arguments get blurred.
And you are also right about fragmentation but the extent of the fragmentation with HTML/Javascript/CSS will never be as great as with Native apps today. So there are advantages here and clever developers will build abstraction layers which will reduce some of the fragmentation.
Sorcery-ltd | 06/10/2008, 23:26
Interesting, while I agree that the web technologies will probably never fragment as badly as native options today, from where I'm sitting I see the native options getting less fragmented. Actually I think there may even come a day where you can write standard C/C++ and a Qt UI almost everywhere.
Fundamentally I dislike the web app concept for mobile platforms, it's wasteful of power and bandwidth, but the advantage of having complete control of the version all your users run and the ability to update that immediately everywhere is too difficult to ignore.
I wonder if the compromise might be something like a web scripted UI that is fully updateable online but cached, running with a native backend that is controlled by the developer, not the device manufacturer or browser vendor, using something like the Qt Webkit integration (if you haven't seen it take a look - it's pretty amazing).
Mark
cell phone news | 02/04/2009, 03:31
hi does anyone of you how to make webapp excutable standalone especially php apps.
and thank you for your article Robin.
Webapps are not everything
tote_b5 | 06/10/2008, 15:00
Hi,
Although I can see (most of) the advantages of webapps I do not think they're the ultimate answer to everything. First off, the technology might be advanced (it is NOT), but tariffs are still so high that users are not willing to use webapps doing extensive data transfer.
Second, there are simply tasks that cannot be done with webapps architecture - usually those that rather require local (i.e. on-device) computing power than afford data being transmitted to back-end server and waiting for result coming over the network.
Thirdly, fragmentation will reach webapps solutions as well. Never think that those big players that you mentioned above will ever agree on a standard to get contacts, access gps, manipulate calendar, etc. so that you can write HTML/JavaScript/CSS/etc. code relying on a single version of the relevant APIs. And eventually it will make the life of developers just as hard as it is today for native apps.
All in all, again, I can see why it is worth writing webapps - I'm just saying that it's not the ultimate solution and imho it will never be.