- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Sun, 22 Nov 2009 12:25:10 -0600
- To: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
- Cc: Shelley Powers <shelley.just@gmail.com>, "Ennals, Robert" <robert.ennals@intel.com>, "public-html@w3.org" <public-html@w3.org>, "Carr, Wayne" <wayne.carr@intel.com>, "Tran, Dzung D" <dzung.d.tran@intel.com>
On Sun, Nov 22, 2009 at 11:40 AM, Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no> wrote: > So, contrary to what you said about namespaces not being a good > solution for introducing new experimental features, it would actually > be simpler to introduce <caption> for other elements than <table> via a > real/prefixed namespace. For example a transitional namespace for new > HTML5 features. The following already works.: > > <figure><t:caption></t:caption></figure> > > As time goes, and old user agents fades out, one could drop the prefix. It's contrary to nothing, and furthermore probably incorrect. 1) What you are talking about is not an experimental feature, it's a standardized feature. There should be only one way to do it that everyone follows; the problem is finding an acceptable way. 2) The only reason namespaces might work for this particular case is as a gross hack that simply temporarily obfuscates the name of the element. That has nothing to do with namespaces - calling it <tcaption> or <x-caption> or <legend> accomplishes the same end. 3) I'd have to test, but depending on how older user-agents handle element names with colons in them, it still might not work. 4) I'd highly question any notion of 'simpler' that rates "use namespaces" higher than "use a new element name" when the sole problem is "the name we want is unusable". ~TJ
Received on Sunday, 22 November 2009 18:26:05 UTC