Hi Simon, the RI is definitely rendering "_id". I have been questioning this before, answer was there are versions of the HTML spec where the _ is a valid starting character. I don't remember which versions this were - XHTML?
regards, Martin On 10/24/07, Simon Kitching <[EMAIL PROTECTED]> wrote: > ---- Martin Marinschek <[EMAIL PROTECTED]> schrieb: > > @warnings in the produced HTML: I can't follow you there - both the RI > > and MyFaces produce ids of the form "_id42", so I don't see how this > > would be different between the implementations? > > This is an interesting point. The HTML401 spec does say that an ID attribute > must start with ([A-Za-z]) > http://www.w3.org/TR/html4/types.html#type-name > > However UIViewRoot.UNIQUE_ID_PREFIX is specified as "_id" in the official > Sun javadocs (ie the spec). > > I guess the html renderkit should be transforming this "component id" into a > valid html value, and transforming it back before the "decode" step. Does > the renderkit get invoked before decode? > > Martin, I presume you're saying that sun are really writing "_id" into the > html output, like MyFaces does? If so there any reason why MyFaces would > need to stay compatible with Sun here? I can't think of one.. > > Regards, > > Simon > -- http://www.irian.at Your JSF powerhouse - JSF Consulting, Development and Courses in English and German Professional Support for Apache MyFaces
