- From: Ian Hickson <ian@hixie.ch>
- Date: Wed, 4 Oct 2006 01:13:32 +0000 (UTC)
- To: Paul Topping <pault@dessci.com>
- Cc: David Carlisle <davidc@nag.co.uk>, www-math@w3.org, dev-tech-mathml@lists.mozilla.org
On Tue, 3 Oct 2006, Paul Topping wrote: > > While backward compatibility across versions of MathML is obviously > desirable, I believe that the creators of future versions of MathML will > feel free to change the definition of something, deprecate it, or remove > it in the new version. Backward compatibility can and should be provided > in the MathML renderer but it can only do so if it can identify the > version of the MathML content it is processing. IMHO backwards compatibility is not just "desirable", it is absolutely and fundamentally critical. Specification authors should always make their specs backward and forward looking. Yes, this constrains what the spec authors can do. That's life. BTW, UAs can't tell what version of content they are handling, even if the language provides a way to tell that, because authors regularly lie (just have a look at HTML DOCTYPEs around the Web, or SVG versions). -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 4 October 2006 01:13:46 UTC