- From: Les Cuff <lez@fastfwd.com>
- Date: Sun, 28 Jun 1998 00:06:37 -0230
- To: "Justin Wells" <reader@semiotek.com>
- Cc: <www-dom@w3.org>
Justin is so very right. The text object MUST not be able to affect the surrounding neighbourhood cluster. It would be apartheid-esque to suggest that only text objects will be permitted at any specific layer of the document. The text object (all objects) must be able to determine their own container class. What if (for argument sake) objects within the same object had different modification, replication and access rights? Is it worth forcing the text object to deal with "go away exceptions" from hostile neighbouring content? Can we drop the idea that "everything in a document has equivalent access and reproduction rights?" Is that assumed (forgive my catch up...) Moreover, I argue for a document naming service, based loosely on DNS except accomodating multiple sites with same address. (Bearing in mind that the Canadian Name space and the American Name space are two distinct things within a single English Language.) Personal name space is the essence of the 'visitation history'. Honestly, if I say Verse 3 in BIBLE, serve the page from my nearest bible. Let the negotiaton for rights reuse and access occur between document servers and have an abstract naming scheme so I can reference a bible without knowing which flavour of bible on which Drive mapping a given user is constrained by. Give the consumer ultimate authority. Geesh. Let's drop CASE SenSiTIviTY from the document name space too, please. Sorry for my absence, Les Cuff nf.ca
Received on Saturday, 27 June 1998 22:39:19 UTC