Converting IDOMNode to IDOMHTMLInputElement fails

2012-11-19 Thread Pelota
I have an IDOMNode
nsCOMPtr_IDOMNode elementNode;

from which I can get the node name without problems ("INPUT", "FIELDSET", ...)
nsString strName;
elementNode->GetNodeName(strName);

Since Gecko 17 I have the problem, that I cannot get the HTML Input Element 
Interface for this node:
nsCOMPtr pHTMLInputElement;

pHTMLInputElement = do_QueryInterface(elementNode);
if(pHTMLInputElement) {

This works fine with < Gecko 16. Any Ideas or known bugs? Did something change 
between Gecko 16 to 17?

___
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform


Re: Converting IDOMNode to IDOMHTMLInputElement fails

2012-11-19 Thread Kyle Huey
The IID of nsIDOMHTMLInputElement changed from Gecko 16 to Gecko 17.  You
are probably compiling against headers for Gecko 16.

- Kyle
On Nov 19, 2012 7:10 AM, "Pelota" 
wrote:

> I have an IDOMNode
> nsCOMPtr_IDOMNode elementNode;
>
> from which I can get the node name without problems ("INPUT", "FIELDSET",
> ...)
> nsString strName;
> elementNode->GetNodeName(strName);
>
> Since Gecko 17 I have the problem, that I cannot get the HTML Input
> Element Interface for this node:
> nsCOMPtr pHTMLInputElement;
>
> pHTMLInputElement = do_QueryInterface(elementNode);
> if(pHTMLInputElement) {
>
> This works fine with < Gecko 16. Any Ideas or known bugs? Did something
> change between Gecko 16 to 17?
>
> ___
> dev-platform mailing list
> dev-platform@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-platform
>
___
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform


Re: Converting IDOMNode to IDOMHTMLInputElement fails

2012-11-19 Thread immonikolausneuhaus
Additional infos:
- Tests on Windows 7
- Firefox 17.0
- do_QueryInterface(elementNode, &error); yields NS_ERROR_NO_INTERFACE 
(0x80004002).

Why didn't I get here the interface since Gecko 17?
___
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform


No DOM Bindings Meeting this Week

2012-11-19 Thread Kyle Huey
Due to the holidays there will be no DOM bindings meeting this week.  We
will resume next week.

- Kyle
___
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform


Re: UA string: "Touch" or "Tablet" again

2012-11-19 Thread Jonas Sicking
On Tue, Nov 13, 2012 at 2:46 PM, Dao  wrote:
> On 13.11.2012 12:24, jim.math...@gmail.com wrote:
>>
>> On Tuesday, November 13, 2012 4:49:14 AM UTC-6, Jonas Sicking wrote:
>>>
>>> Note that putting "touch" in the UA is somewhat different than
>>> traditional UA sniffing. It's actually capability testing which is
>>> what we are encouraging people to do. Using HTTP-headers is the way
>>> you do server-side capability testing, comparable to using JS and
>>> API-detection to do client-side capability testing.
>>
>>
>> Isn't this what we want to avoid? For the user with a laptop that
>> supports touch they are going to get thrown into touch centric (mobile)
>> interfaces on web sites. I have had this happen on the tablets I test
>> with thanks to the touch interfaces
>
> This is a misguided simile. Once upon a time, the touch interfaces were
> actually only present on mobile devices. This is never going to be the case
> for the Touch token. The Touch token won't be present on mobile devices;
> it's unclear to me why you think anyone would treat it like the touch
> interfaces.

If indeed we can't send the token on mobile devices (because it would
cause them to send us the somewhat-larger-screen-size tablet UI?) then
I agree with you. My proposal only makes sense if we can send the
"touch" token for mobile as well.

/ Jonas
___
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform


Re: UA string: "Touch" or "Tablet" again

2012-11-19 Thread Jonas Sicking
On Tue, Nov 13, 2012 at 8:37 AM,   wrote:
> On Tuesday, November 13, 2012 2:49:14 AM UTC-8, Jonas Sicking wrote:
>> Websites generally send dramatically different content for touch-based
>> UIs. Different enough that they'd want to send a different piece of
>> main content.
>
> Just to be clear, this is NOT generally true in what I've seen. Websites send 
> dramatically different content to small screen devices, but not to touch 
> based ones or tablets.

If this is indeed the case, maybe we should add something to the UA
string which indicates the screen size.

/ Jonas
___
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform