This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.

Bug 3320 - Using IDs for nodes instead of / in addition to XPath
Summary: Using IDs for nodes instead of / in addition to XPath
Alias: None
Product: ITS
Classification: Unclassified
Component: ITS tagset (show other bugs)
Version: WorkingDraft
Hardware: PC Windows XP
: P2 normal
Target Milestone: AfterLC
Assignee: Felix Sasaki
QA Contact: Felix Sasaki
Keywords: changeDeclined, reviewerSatisfied
Depends on:
Reported: 2006-06-07 05:55 UTC by Felix Sasaki
Modified: 2006-10-03 14:42 UTC (History)
1 user (show)

See Also:


Description Felix Sasaki 2006-06-07 05:55:20 UTC
See comments from at Daniel Glazman and Werner Donne at

[[Daniel Glazman (invited expert in the CSS working group): "Why don't
you use IDs for elements (and keys for attributes), instead of the -
computationally expensive - XPath?"
My reply was: It would be good to do that, but you can do it only if you
have control over the whole data in the localization workflow. I heard a
similar comment from Iris Orris (Microsoft) at the Unicode conference,
but it seems to me this scenario works only for data inside big corporates.

- Werner Donne gave a talk on "Managing Multilingual Legislation With
XML", see
 . In his approach, he added "metadata" about versioning into XML
documents. The meta data was used for alignment of EC documents. Werner
said the ITS approach of global rules seems promising for such
versioning, since it has no or limited impact on the document(s). One
potential issue he saw (again, see Daniel above) was the computational
cost of XPath for large documents.]]
Comment 1 Felix Sasaki 2006-06-20 05:16:26 UTC
Added Werner to the CC list.
Comment 2 Felix Sasaki 2006-07-24 12:38:03 UTC
Action: working group to discuss and reply.
Comment 3 Felix Sasaki 2006-07-24 12:40:04 UTC
Action: working group to discuss.
Comment 4 Yves Savourel 2006-08-11 17:08:24 UTC
Proposed answer:

ITS does not use directly ID-based mechansim because it tries to cover more general cases: not all XML format use IDs. ID-only is also quite limitative if one need to use conditions (e.g translate element X if its parent has an attribute Y with a value Z. XPath 1.0 and later provides ID function that can be used to work with IDs if needed (See an example in the comment #1 of the issue #3463:

The cost of using XPath is obviously an aspect we had in mind. But from the test implementation we have and from the functionality we tried, the benefits of using XPath seem to outweighs its drawbacks.
Comment 5 Felix Sasaki 2006-09-06 14:17:54 UTC
Comment 6 Felix Sasaki 2006-09-27 12:38:16 UTC
Summary: The Working Group decided to decline the change.
Comment 7 Werner Donn 2006-10-03 09:37:55 UTC

The resolution of the Working Group is fine for me.