>> I think the issue is for someone relatively new to T5, they don't >> know/understand why <input t:type="textfield.../> and >> <t:textfield.../> are equivalent, but <div t:type="remove> fails. To >> them it just seems buggy and/or inconsistent. The basic mechanisms >> look the same, but they don't act the same. That causes confusion. > > I agree. I don't use the <t:componentName> syntax.
Me three. I've been doing web application development for it seems like a century now. The primary reason I'm porting myself from WebObjects to Tapestry is very similar to the reason I originally selected WebObjects. Beauty. Making a beautiful website means working with artists, and it means some rounds of feedback cycles with those artists. So that means something like the invisible instrumentation stuff in Tapestry. All of the other frameworks out there are too prone to merge code with the HTML, resulting in, well, muck. Me and my artist intend to make full use of this, so that mostly, he can pass me HTML he's gotten working in a browser, I can add t:type="foo" chunks to parts of it to make it dynamic, and it's all good. So rather than t:type="foo" being _optional_, t:type="foo" is the _idiom_. That is, the standard way to do things so that your templates look nice, are understandable by artists. When I build components, I take the extra five minutes to turn them into web pages so that they're self documenting... > T5 already has some magic elements, a.k.a. directives, such as t:body, > t:parameter, t:block, t:content and t:remove. I don't want to > introduce magic component types as well. Agreed, but does it seem unreasonable to have the directives work with invisible instrumentation as well? Even if it had to be t:directive="remove"? I'm willing to do the work to make that happen if someone gives me a hint of where I should do that. Though couldn't t:type just match against the list of directives either first or last? Pierce
smime.p7s
Description: S/MIME cryptographic signature
