RE: [xml-dev] ID/IDREF is evil

["Cox, Bruce" <Bruce.Cox@USPTO.GOV>]

Since our manuals have been managed as multiple files from the start, any potential cross-reference target has been given a GUID (type=ID), guaranteed to be unique across the entire manual, not just the chapter in which it occurs.  The editors are very systematic about what they will and will not reference, so we were able to build a widget in oXygen that generates the GUID when they create the hyperlink.  I don't have much experience with Schematron or tools that support it, so I'm a little foggy on how, in practice, and editor would actually build a link without having to leave their place in a chapter.

Bruce B Cox
OCIO/AED/Software Architecture and Engineering Division
571-272-9004


-----Original Message-----
From: Michael Kay [mailto:mike@saxonica.com] 
Sent: 2014 February 4, Tuesday 10:59
To: Cox, Bruce
Cc: Costello, Roger L.; xml-dev@lists.xml.org
Subject: Re: [xml-dev] ID/IDREF is evil


On 4 Feb 2014, at 15:51, Cox, Bruce <Bruce.Cox@USPTO.GOV> wrote:

> Hey, Roger, I see you're still at it.  I've been busy with other things and haven't looked at xml-dev for quite some time, but I'm going to get back into it, if only for the entertainment value.  I'm retiring in 24 days, so I'm going to want some brain teasers.
> 
> I've heard others advise to abandon ID/IDREF, but yours is the first rationalization I've heard, and you're the only one I've heard suggest Schematron as the replacement.  If you look at mpep.uspto.gov, or tmep.uspto.gov, you'll see what I've been up to the past three years.  These procedure manuals are fairly bursting with internal cross references, which we manage with ID/IDREF.  In practice, how much more efficient would Schematron be for this application?  Would we actually see performance gains?
> 

As far as I'm concerned, performance isn't the issue. The issue is that you can't break your document into pieces without switching to a different validation technology.

I see this with the W3C specs I manage, The DTD for these makes heavy use of ID/IDREF. As the specs have grown we have split some of them into pieces for ease of management (e.g. change control, authoring), and this makes the documents that one actually authors invalid against their own DTD. 

Michael Kay
Saxonica