RE: clarification of p:pipeline-library and p:import from Vasil Rangelov on 2007-10-24 (public-xml-processing-model-comments@w3.org from October 2007)

From: Vasil Rangelov <boen.robot@gmail.com>
Date: Thu, 25 Oct 2007 02:19:11 +0300
To: <public-xml-processing-model-comments@w3.org>
Message-ID: <471fd2f2.0bbf5e0a.0c94.54ce@mx.google.com>

>can WG comment if there is a default value to namespace attribute on
>p:pipeline-library ?
>
>e.g. namespace="http://www.w3.org/ns/xproc"

That should be obvious. The "p" prefix and its namespace URI is defined earlier in the document. If there's an error in an example though (to which you don't point), it should probably be corrected.

>should we allow filter option on href? or provide a globbing syntax?

What's the benefit of that? Clever implementations should load each imported step definition only once it is requested and not found in the current document, so that they use up less memory. Thus, using such filters on imports makes no sence.

Received on Wednesday, 24 October 2007 23:19:25 UTC