[
Lists Home |
Date Index |
Thread Index
]
At 11:01 AM 5/8/2002 -0700, Dare Obasanjo wrote:
>Seriously though, you may pontificate all you want but the fact of the
>matter is that vendors typically do not completely implement standards
>interoperably in the software industry unless the standard is fairly
>straightforward to implement and have an organization with teeth
>enforcing conformance(e.g. Java). C++, C99, SQL, HTML, CSS, etc are all
>examples of "standards" that have never really been fully implemented by
>various vendors.
Yes, this is true. XQuery will never be as short and simple as the XML 1.0
spec. The W3C does not have the same teeth that Sun has with Java, and does
not enforce conformance.
I *would* like to see conformance test suites developed for XQuery.
>Considering that there are over 900 pages of documentation to read
>before one fully grasps what is actually an XQuery implementation (with
>more on the way), this doesn't count as straightforward to me.
The 900 pages is somewhat inflated, but I agree that there's a lot to read
and understand.
>XQuery Requirements 13
>XQuery Use Cases 88
These two are not normative for implementors.
>XQuery 1.0 (Syntax) 175
>XPath 2.0 (Syntax) 149
These two are essentially the same document. XQuery adds construction
features, XPath adds fallback. For an XQuery implementation, you can ignore
the XPath document.
>Functions & Operators 193
Essentially the function library. Yes, you need to implement this.
>Formal Semantics 222
>Data Model 59
These are normative.
>XQueryX (XML syntax for XQuery) 23
You do not need to implement this, and if you do, it is just a different
syntax for XQuery, with no new semantics.
So the number of normative pages to read for the language itself is 175 +
222 + 59 = 446. Add the functions and operators, and you are up to 639.
That's not more than 900 pages, but it is still a lot of reading.
>The main reason I bring this up is that I'd like the XQuery WG to think
>seriously about things like conformance levels and trimming features if
>they don't want XQuery to end up as yet another standard that vendors
>pay lip service to.
There are already notes in the XQuery document suggesting this is likely to
happen, eg:
XQuery is likely to have multiple conformance levels.
There may be a conformance level that does not include
static type checking. There may be a conformance level
that does not support Schema import, so that only built-in
types and node types may be used in declarations.
Jonathan
|
