3 ms·
> because a style-description language would not normally be > expected to be Turing-equivalent. Wouldn't it? XSLT, the only other style-description language
by bzbarsky 16y ago
> because a style-description language would not normally be
> expected to be Turing-equivalent.
Wouldn't it? XSLT, the only other style-description language in anything like common use on the web, has been Turing-complete for a while now (possibly since it started existing; I'm not familiar with the history)...
- arethuza 16y agoXSLT is the transformation part of XSL (hence the T), meant for transforming XML into other formats, including things like XSL-FO which is the formatting part of XSL. As it is a general purpose language (with iteration, recursion and conditionals) it's not that surprising that it is Turing-equivalent. XSL-FO is the formatting part of XSL (i.e. the bit most like CSS) - I'd be pretty surprised if this was shown to be Turing-equivalent.
- chwahoo 16y agoXSLT performs transformations (i.e., takes input and produces output). This sounds to me like a "program" so I wouldn't be surprised that a language like XSLT might be Turing complete. HTML and CSS are much more "static" in flavor (hence the OP's use of the term "descriptive") and so it might be more unexpected that they can perform universal computation.