W3C home > Mailing lists > Public > public-xmlsec@w3.org > August 2010

RE: Streamable XPath Profile Review

From: Scott Cantor <cantor.2@osu.edu>
Date: Mon, 9 Aug 2010 17:06:44 -0400
To: "'Pratik Datta'" <pratik.datta@oracle.com>, "'Meiko Jensen'" <Meiko.Jensen@ruhr-uni-bochum.de>, "'XMLSec WG Public List'" <public-xmlsec@w3.org>
Message-ID: <01a901cb3806$c58db9b0$50a92d10$@osu.edu>
> Regarding complexity, I think it is ok to be complex. XPath 1.0 is complex
> and 2.0 is even more. What we are proposing is much less complex than
> either. I am assuming most Signature implementation will just be DOM based
> and not use streaming at all.  Streaming is only for very performance
> sensitive usages.

One point of input on this "subjective" aspect...I think the breaking point
is can you plausibly imagine people trying to use a clean room "not full
XPath, just what the profile needs" implementation inside a DOM-based
signature implementation so as to avoid pulling in/requiring an XPath
library?

(The streaming implementations will be one-offing everything anyway, that's
why I'm just thinking of the DOM side.)

If yes, we want to limit options as much as possible. If not (or if it's a
bad idea), then as long as it's consistent with XPath, I agree that it
probably doesn't matter that much.

-- Scott
Received on Monday, 9 August 2010 21:07:17 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:55:14 UTC