Thursday, 10 June 2010

"W3C widgets still 2 years away" says well known app developer

Yesterday I met the head of a big mobile application developer here in London. When talking about widgets he said "well they told us about them some time ago and they're still 2 years away".

I was quite shocked by this and partly because I think there is an element of truth in it. When I think about it, I have been working with WRT technology for nearly 2 years already and with W3C widgets for a little less time and I can agree that 2 years ago I expected them to be mainstream by now; they are not. In fact the only appstore that can install WRT widgets is Ovistore. All other appstores have not added the mime type they need to do to enable their download. The work to add this mime type would take just a few minutes, with admittedly a UI change to say they can accept WRT, plus some testing. Not very significant effort but no-one has done it. I've asked GetJar to do it, I've asked one of the big appstore suppliers to Operators in the far east to do it, I've asked Sony Ericsson to do it (after all they have S60 phones with WRT capabilitiy but their appstore does not accept that type). All I get is we are considering doing it. Why? I am not sure why they refuse to do it - to some extent is lack of inertia or even foresight that this technology will be massive. The problem is there is no Steve Jobs to tell the world about the technology No-one giving it any momentum whatsover.

Meanwhile have the backing of Vodafone, China Mobile and Softbank. This could not be bigger. It's beginning to move. You can submit widgets to JIL which will eventually end up on these operator portal (although currently its only Vodafone360 shop where they go). So it is beginning to start moving, if a little slowly.

Is it still 2 years away from being mainstream? I hope not!


  1. My years of software development have taught me that a change always seems simple and quick if you can't see the code and processes which require changing. Assumptions like "would take just a few minutes" and "Not very significant effort" are almost always guaranteed to be wrong.

    Here's a quick, off the top of my head list of things which would actually be required:
    - App/widget approval processes and documentation updating
    - Developer instructions and support
    - End user support
    - Training and documentation/instructions for those providing support
    - Review of purchase processes and possible design changes
    - UI changes
    - Internationalisation of all of the above.
    - Testing, testing, testing!

    Remember that most app stores are run by large companies and therefore often fall foul of long complicated approval/sign-off processes and resistance to change. This will make all the above more complicated and time consuming too.

    It's catch-22. Why should an app store go through all the time, effort and expense of being able to support widgets if there aren't sufficient numbers available to make it worth their while? And why would developers make widgets they can't distribute through an app store?
    There's also the issue of the strong selling point of widgets being that they can run cross platform. One thing app stores are looking at and trying to get is exclusive applications which consumers will by devices for to get those apps. If apps become widgets that can run everywhere MNOs and OEMs lose that differentiation.

    Yes, it just needs a Jobs type figure to just support it and then everyone will follow. Until then—who knows.

    I'd say we're at least 2 years away from mainstream.

  2. Unfortunately I agree here. Though, being deep inside of this technology and the ongoings I see a lot of push, especially form the Vodafone side. But enabling widgets to also run on iPhone, Android, etc. using technologies like PhoneGap demostrate well that the core of the future apps is the widget.
    We at uxebu do our best to also close the gap as soon as possible, but that is not a one day job unfortunately. We have a demo app, that proves that widgets can be cross platform
    But I am convinced sticking to widgets is the right thing to do, they are the future! What they will be called then is another thing :)
    /me like "html5 apps" (thx ppk)