Thanks for writing this up Frank.
One concern about this:
> we could convert the fieldset to <fieldset name="twoFields" id="twoFields" jsfc="h:panelGroup" elementName="fieldset"> and the legend to <legend id="legend" jsfc="h:panelGroup" elementName="legend">
Is that it requires duplication of the element name.
What did you think of the idea that Imre and I discussed - ie. doing
something like:
<legend id="legend" jsfc="h:element">
Or maybe some simplification of this, eg:
<legend id="legend" jsfc="true">
<legend id="legend" jsfc="jsfc">
If we go this route we would need to have some more discussion about how
to handle component-level attributes (eg. rendered, id) vs pass through
attributes (eg. everything else).
Andy
On 6/4/12 12:20 PM, Frank Caputo wrote:
> Hello EG,
>
> this email has 3 SECTIONS. Find the proposal at the end.
>
>
> SECTION: Answer to Ed's question.
>
> I can answer the question with a simple example. Suppose the following simple form where you have a dynamic fieldset and a dynamic legend rendered on ajax requests.
>
> <h:form>
> <fieldset name="twoFields" id="twoFields">
> <legend id="legend">#{bean.legend}</legend>
> <h:inputText value="#{bean.text}">
> <f:ajax execute="@this" render="legend" event="keyup"/>
> </h:inputText>
> <h:inputSecret value="#{bean.secret}">
> <f:ajax execute="@this" render="twoFields" event="blur"/>
> </h:inputSecret>
> </fieldset>
> </h:form>
>
> If we added an attribute elementName to the panelGroup, we could convert the fieldset to <fieldset name="twoFields" id="twoFields" jsfc="h:panelGroup" elementName="fieldset"> and the legend to <legend id="legend" jsfc="h:panelGroup" elementName="legend">.
>
> If we wanted to do this with composite components, we would have to write 2 composite components, which would be quite complex, to achieve this simple behavior.
>
> Since the generic dynamic element behaves exactly the same like the panelGroup, I still propose to simply add the attribute elementName to h:panelGroup.
>
> I think, this is a very simple change, having the passthru attributes, and gives the HTML developers a good experience.
>
>
> SECTION: alternative
>
> We could do copy and paste and invent a new element which behaves like h:panelGroup with an additional attribute elementName.
>
>
> SECTION: proposal
>
> 1. Add the attribute elementName to the VDL doc of h:panelGroup with the following constraints:
> - name: elementName
> - required: false
> - request-time: false
> - type: javax.el.ValueExpression (must evaluate to java.lang.String)
> - description: The name of the rendered HTML element. If layout is also specified, elementName takes priority.
>
>
> Cheers,
> Frank
>
>
> Am 31.05.2012 um 00:05 schrieb Edward Burns:
>
>
>> Hello Experts,
>>
>> I've had a hard time following the discussion on passthru elements,
>> mainly because I was trying to *not* follow it because I was trying to
>> make progress on passthru attributes. To save all of us time, I
>> nominate Frank Caputo to craft a cohesive and self-contained proposal,
>> IF he can answer this question. Why do we need pass through elements in
>> the first place when you can do something like this?
>>
>> Using page:
>>
>> <my:componentUsingHtml5 a="b" c="#{d.e}">
>> <f:actionListener binding="#{d.f}" />
>> </my:componentUsingHtml5>
>>
>> Defining page:
>>
>> <cc:interface componentType="javaTopLevelComponent">
>> <cc:actionListener ...>
>> <cc:attribute >
>> ...
>> </cc:interface>
>>
>> <cc:implementation>
>>
>> arbitrary html 5 capable content goes here.
>>
>> </cc:implementation>
>>
>> I'd say that if the composite component feature is not up to the task,
>> then we should improve it rather than inventing more syntax and giving
>> our detractors more reason to say we're too complex.
>>
>> Sincerely,
>>
>> Ed
>>
>> --
>> | edward.burns_at_oracle.com | office: +1 407 458 0017
>> | homepage: | http://ridingthecrest.com/
>> | 31 Business days til JSF 2.2 Public Review to EG
>>
>>
>
>
>