No.

A TextField is a component that handles rendering an <input
type="text" tag> and handling a subsequent form submission in the
context of a Form component (including all the related Ajax and
validation logic).

If you have a component that needs to render an <input> element for
some other reason (I assume some DHTML/Ajaxy thing): just do that.
Put <input> into the component's template. Come up with your own way
of assigning a name and/or id.  Tapestry won't care.

On Mon, Mar 9, 2009 at 11:20 AM, Andrea Chiumenti <[email protected]> wrote:
> Hello I'm currently integrating dojo into Tapestry5. Today things seem
> to go well, but I've seen one thing that doesn't fully convince me.
>
> In AbstractField we have
> @Environmental(false)
> private FormSupport formSupport;
>
> and then
>
> @SetupRender
>    final void setup()
>    {
>        // By default, use the component id as the (base) client id.
> If the clientid
>        // parameter is bound, then that is the value to use.
>
>        String id = clientId;
>
>        // Often, these controlName and clientId will end up as the
> same value. There are many
>        // exceptions, including a form that renders inside a loop, or
> a form inside a component
>        // that is used multiple times.
>
>        if (formSupport == null) throw new
> RuntimeException(InternalMessages.formFieldOutsideForm(getLabel()));
>
>        assignedClientId = renderSupport.allocateClientId(id);
>        String controlName = formSupport.allocateControlName(id);
>
>        formSupport.storeAndExecute(this, new Setup(controlName));
>        formSupport.store(this, PROCESS_SUBMISSION_ACTION);
>    }
>
>
> so a formSupport null will throw an excpetion.
>
> When I'll decorate components like TextField I'll be obliged to put it
> inside a Form component, when people not always need it.
>
> Should't be possible only to send a warning or something like that on
> formSupport null in 5.1.x?
>
>
>
> On Mon, Mar 9, 2009 at 7:11 PM, Howard Lewis Ship <[email protected]> wrote:
>> I'm working on the last major feature I'd like to see in 5.1:
>> template inheritance (*).
>>
>> With that, I'm ready to switch from implementing features to fixing bugs.
>>
>> I think it would be approrpriate to release 5.1.0.1, fix a bunch of
>> bugs, then see about getting a beta release out.
>>
>> Outside of Spring Web Flow and Portlet support (things I'd like to
>> work on in 5.2), the only other major feature I see people clammoring
>> for is dynamic updates on Selects, something I'm still noodling on.
>> There's probably a few other minor features that would be nice to
>> haves, but you have to draw the line somewhere!
>>
>>
>> (*) This is an idea I picked up from wicket, though I'm extending it
>> quite a bit. Parent component templates will be able to define
>> <t:extension-point>s, and child component template will be able to
>> provide <t:replace> elements that replace them (a unique id ties the
>> two together). This can be used an an alternative to using a layout
>> component, where each page will <t:replace> the default page content
>> inherited from the base class. I generally think that for pages, a
>> layout component is better, but for complex components (such as the
>> Grid), this will be a great new feature for allowing changes to a
>> container template without cutting and pasting the whole thing.
>>
>> --
>> Howard M. Lewis Ship
>>
>> Creator Apache Tapestry and Apache HiveMind
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [email protected]
>> For additional commands, e-mail: [email protected]
>>
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>



-- 
Howard M. Lewis Ship

Creator Apache Tapestry and Apache HiveMind

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to