Re: A simpler, webbier approach to Web Intents?

On Wed, Mar 21, 2012 at 1:42 PM, GAUSMAN, PAUL <pg2483@att.com> wrote:
> ·        Could a web page or web app include a list of services to be added
> to the list of available services for various Intents?

We've proposed the <intent> tag for this purpose. That's still under
discussion. For Chrome we will support adding intent support to the
manifest for web apps. (That support is in the dev/beta Chrome 18
now.)


> ·        Can intents services be embedded in ads that come from an ad
> network and are inserted into web pages?

The spec discourages UAs from allowing installation from iframes.

> ·        Can an intent service be pre-selected by the page or by client
> settings so that the choice does not have to impact the user each time?

This kind of defaulting behavior will be an important component of
user agent behavior, yes.

> ·        Can a service provider have an intents services function which
> could be pushed to a browser?

I don't understand this question. Do you mean pushing service
registration changes to the browser?


>
>
>
> Thanks!
>
> -Paul
>
>
>
> Q me
>
>
>
>
>
> From: Paul Kinlan [mailto:paulkinlan@google.com]
> Sent: Tuesday, February 21, 2012 11:14 AM
> To: John J Barton
> Cc: Rachel Blum; Greg Billock; public-web-intents@w3.org
> Subject: Re: A simpler, webbier approach to Web Intents?
>
>
>
> Kind of.
>
>
>
> The user would accumulate entries on the sites that they visit, and these
> are the apps that are then registered, this is where the user grants the
> ability for the service to fulfill the intent.  The user to would not have
> to click install in the IDE at all, the user in the IDE would just request a
> service from the browser.
>
>
>
> There are other ways of finding services too, the current shim maintains a
> registry of publicly accessible services that support intents and can offer
> them to the user to invoke.
>
>
>
> P
>
>
>
> On Tue, Feb 21, 2012 at 3:53 PM, John J Barton <johnjbarton@johnjbarton.com>
> wrote:
>
> Ok, so I gather the idea is that a user will visit some sites that
> have top-level pages with intent tags and, as a side effect, they will
> accumulate entries in the registry. Then they would go to the IDE and
> click "install" or something. Then the IDE would then request the
> services it is programmed to consume and the user would select the one
> they want from the list.
>
> Is this correct?
>
> jjb
>
>
> On Mon, Feb 20, 2012 at 9:38 AM, Paul Kinlan <paulkinlan@google.com> wrote:
>> Sorry to add to the confusion.  I only stated that registration couldn't
>> happen via an iframe and it shouldn't.
>>
>> That is not to say we shouldn't have an iframe solution, I think we should
>> and it shouldd be discussed and worked on here because I see Web Intents
>> solving the negotiation problems that we see in all of the raw postMessage
>> solutions, which is something the developer should not have to handle.  I
>> believe Darin Fisher suggested something similar on this group a while ago
>> [1].
>>
>> I spoke to someone a while ago about Orion and this was something we
>> didn't
>> have support for in Web Intents yet.  The closest we get is "inline"
>> disposition which displays the launched app in the context of the old app.
>>
>>
>> [1] http://lists.w3.org/Archives/Public/public-web-intents/2011Dec/0050.html
>>
>> On Mon, Feb 20, 2012 at 5:22 PM, John J Barton
>> <johnjbarton@johnjbarton.com>
>> wrote:
>>>
>>> Oh, sorry. I ended up here after I asked on public-webapps about a
>>> standard way to establish communications between cross-domain iframes
>>> and a web page. Allowing plugin authors and plugin users to succeed
>>> without negotiating with the framework authors is powerful advantage.
>>> The use case, plugins for development tools, closely resembles
>>> web-intents. The plugin API could be as simple as image exchange like
>>> the memegen example (hence my interest in web-intents) or more
>>> complicated exchanges requiring two way operations (hence my questions
>>> about using postMessage).
>>>
>>> But iframes cannot participate in web-intents so we are out of luck.
>>>
>>> jjb
>>>
>>> On Mon, Feb 20, 2012 at 8:58 AM, Paul Kinlan <paulkinlan@google.com>
>>> wrote:
>>> > why?
>>> >
>>> >
>>> > On Mon, Feb 20, 2012 at 4:38 PM, John J Barton
>>> > <johnjbarton@johnjbarton.com>
>>> > wrote:
>>> >>
>>> >> Ok, thanks for the info. Sounds like we will have to look elsewhere
>>> >> for a cross-domain application integration framework.
>>> >> jjb
>>> >>
>>> >> On Mon, Feb 20, 2012 at 5:02 AM, Paul Kinlan <paulkinlan@google.com>
>>> >> wrote:
>>> >> > No.  The shim blocks all registrations from occurring inside an
>>> >> > iframe.
>>> >> >  I
>>> >> > would expect the native implementation to do the same.  If the spec
>>> >> > doesn't
>>> >> > mention this, it should.
>>> >> >
>>> >> > On Fri, Feb 17, 2012 at 11:55 PM, John J Barton
>>> >> > <johnjbarton@johnjbarton.com> wrote:
>>> >> >>
>>> >> >> On Fri, Feb 17, 2012 at 3:41 PM, Rachel Blum <groby@google.com>
>>> >> >> wrote:
>>> >> >> >
>>> >> >> >>
>>> >> >> >> > Why wouldn't I race James to create
>>> >> >> >> > WorldsMostAwesomeWebIntents
>>> >> >> >> > page
>>> >> >> >> > full of <intent> tags?
>>> >> >> >> > Won't people be motivated to create ad supported lists? Won't
>>> >> >> >> > users
>>> >> >> >> > be
>>> >> >> >> > bombarded with <intent> pages?
>>> >> >> >> > I guess these are problems you'd love to have.
>>> >> >> >
>>> >> >> >
>>> >> >> > Let's keep in mind that (unless I misremember) intent tags have a
>>> >> >> > same-origin restriction on the action path.
>>> >> >> >
>>> >> >> > So you'll actually need to do a bit of work beyond just
>>> >> >> > collecting
>>> >> >> > tags
>>> >> >> > on
>>> >> >> > your page - unless you choose to provide no-op intents.
>>> >> >>
>>> >> >> I guess a page full of
>>> >> >> <iframe src=<page-with-intent-tag>>
>>> >> >> wouid work, right?
>>> >> >>
>>> >> >> jjb
>>> >> >>
>>> >> >>
>>> >> >> >
>>> >> >> > Rachel
>>> >> >>
>>> >> >
>>> >> >
>>> >> >
>>> >> > --
>>> >> > Paul Kinlan
>>> >> > Developer Advocate @ Google for Chrome and HTML5
>>> >> > G+: http://plus.ly/paul.kinlan
>>> >> > t: +447730517944
>>> >> > tw: @Paul_Kinlan
>>> >> > LinkedIn: http://uk.linkedin.com/in/paulkinlan
>>> >> > Blog: http://paul.kinlan.me
>>> >> > Skype: paul.kinlan
>>> >> >
>>> >
>>> >
>>> >
>>> >
>>> > --
>>> > Paul Kinlan
>>> > Developer Advocate @ Google for Chrome and HTML5
>>> > G+: http://plus.ly/paul.kinlan
>>> > t: +447730517944
>>> > tw: @Paul_Kinlan
>>> > LinkedIn: http://uk.linkedin.com/in/paulkinlan
>>> > Blog: http://paul.kinlan.me
>>> > Skype: paul.kinlan
>>> >
>>
>>
>>
>>
>> --
>> Paul Kinlan
>> Developer Advocate @ Google for Chrome and HTML5
>> G+: http://plus.ly/paul.kinlan
>> t: +447730517944
>> tw: @Paul_Kinlan
>> LinkedIn: http://uk.linkedin.com/in/paulkinlan
>> Blog: http://paul.kinlan.me
>> Skype: paul.kinlan
>>
>
>
>
>
>
> --
> Paul Kinlan
> Developer Advocate @ Google for Chrome and HTML5
>
> G+: http://plus.ly/paul.kinlan
> t: +447730517944
> tw: @Paul_Kinlan
> LinkedIn: http://uk.linkedin.com/in/paulkinlan
> Blog: http://paul.kinlan.me
> Skype: paul.kinlan
>
>

Received on Thursday, 22 March 2012 06:48:29 UTC