- From: Liam R E Quin <liam@w3.org>
- Date: Mon, 16 Jan 2012 17:42:09 -0500
- To: "C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>
- Cc: www-xml-schema-comments@w3.org, Henry Story <henry.story@bblfish.net>, Noah Mendelsohn <nrm@arcanedomain.com>
On Mon, 2012-01-16 at 14:13 -0700, C. M. Sperberg-McQueen wrote: [...] > In both cases, of course, it has rhetorical importance; supporters > of XSD use it (as NM seems to be doing here) to suggest that > certain uses of XSD are misleading, incorrect, improper, and / or > impure, while critics of XSD use it to criticize XSD for not making > it possible to validate things that users in fact would like to be able > to validate, and which do not in fact pose any particular difficulty > for validation (beyond the fact that XSD doesn't provide the > necessary primitive notions). Correct and complete... where's Bertrand Russell when we need him? More to the point, I believe the way forward for validation is pipeline validation, with multiple (possibly transforming) steps. Note that there is likely to be an additional constraint on the decoded binary value, that the checksum matches. One can imagine a schema language capable of expressing that directly. One can also imagine a schema language saying that a dateTime is valid only if it has occurred in the past (leap-seconds included), or that a stock ticker symbol must be currently traded on the stock exchange named by @xid... I've for sure encountered real business needs for such things. But they are beyond the remit of lexical validation, and I do not think it reflects badly on XSD that there are things it cannot do. Regards, Liam -- Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/ Pictures from old books: http://fromoldbooks.org/
Received on Monday, 16 January 2012 22:43:15 UTC