Re: Ringmark, Core Mobile, and the Financial Times

On Jun 11, 2012, at 11:21 , Charles McCathieNevile wrote:
> On Mon, 11 Jun 2012 11:13:34 +0200, Tobie Langel <tobie@fb.com> wrote:
>> On 6/11/12 11:05 AM, "Charles McCathieNevile" <chaals@opera.com> wrote:
>>> On Sun, 10 Jun 2012 23:03:36 +0200, Robert Shilston
>>> <robert.shilston@ft.com> wrote:
>>> 
>>>> 2) Ability to create home screen / chrome-less web apps without needing
>>>> 
>>>> native wrappers.  A combination of
>>>> https://people.mozilla.com/~anarayanan/webapps.html and the Fullscreen
>>>> API will help this.
>>> 
>>> Apart from lack of broad implementation (which is obviously a
>>> show-stopper for practical deployment of an app), do you have any
>>> comments about what widgets did wrong in this area?
>> 
>> Outside of the name? :-P
> 
> Seriously, I think the name turned people off - and while that seems silly it was also a real problem. I'd like to understand what other real problems existed in that technology stack, because I have found a lot of things it got right (other than deployment).

My list:

• the name, "Web Runtime Application Packaging" (WRAP) was much better;
• the timing, most of the work happened before most browser vendors realised that packaged Web apps could be useful;
• the XML, it's silly but some people care;
• mixing packaging and configuration, I think it would have been more successful if the configuration could have immediately applied to hosted apps;
• insufficient thinking about the runtime, because going offline in that way and opening the door to unsafe APIs has implications that are useful to think about before building a packaging format (the widget URI scheme was just one small plug in that hole);
• the security, XML may be acceptable to some people, but to many of those XML DigSig isn't.

-- 
Robin Berjon - http://berjon.com/ - @robinberjon

Received on Monday, 11 June 2012 10:44:35 UTC