• To: "David Carlisle" <davidc@n...>,<gtn@r...>
• Subject: XML ID - Why is it needed ?
• From: "Jens Jakob Andersen, PDI" <jens.jakob.andersen@p...>
• Date: Mon, 12 Nov 2001 10:36:12 +0100
• Cc: <xml-dev@l...>
• Thread-index: AcFrXA+e3K9h5eTARQi16rKHgbRbQgAABlMw
• Thread-topic: So maybe ID isn't a problem after all.

Hello Xternal Material Loungees

I've been watching the XML ID discussion, and I think that I have a few
central questions, since the core of XML ID just eludes me.

1. Why do we need XML ID ?
2. A ID number/tag in a given XML document will not describe the
uniqueness of this XML documents content, it will just be a globally
unique identified (GUID) for this particular document (and then how do
we define uniqueness ? Across the world, or just across a given local
environment, such as a database?)
3. I guess that it wouldnt make sense to add a GUID to each and every
XML document (look at the HTML of the world today.)
4. If the ID is needed for i.e. middleware to handle documents in
transit, the middleware should just add this id to the transport
container (think the OSI model,
http://webopedia.internet.com/quick_ref/OSI_Layers.html). (Hey, here is
a great model on how you handle stuff such as ID's in a non-intrusive
way, hmm...)

5. As we say on the 2CV mailing list, "If it aint broke, dont fix it". I
think that if the current XML spec isn't broke, we shouldnt change it
unless we have some really really really good reasons.

With Xtreme Many Lookups

JJ
 

• Prev by Date: Re: So maybe ID isn't a problem after all.
• Next by Date: IDs considered harmful or why keys might be better than IDs...
• Previous by thread: I need a research position to continue my project
• Next by thread: IDs considered harmful or why keys might be better than IDs...
• Index(es):
  • Date
  • Thread