- From: Philip TAYLOR [PC87S/O-XP] <P.Taylor@Rhul.Ac.Uk>
- Date: Tue, 13 Aug 2002 12:27:30 +0100
- To: www-html@w3.org
Sorry, accidentally sent last message before pasting : it should have read : -> Lachlan Cannon wrote: > Please correct me if I'm wrong, but I believe the point of the set-up > being this way is so that > > a) markup tags are well thought out before being implemented -- you'd > want to have your tags well thought out before modifying the dtd, to > prevent having to do it again. > > b) everyone and their dog won't go out and make up random, and in a > lot of cases pointless tags, and overlapping ones with different names. > Why would you spend 5 hours creating a DTD integrating animalML with > XHTML when someone else has done it? Re-use of DTDs saves bandwidth > (assuming they're cached), effort, time, and it means that applications > don't have to be developed for 1000s of different formats -- one of the > points of XML is saving the costs of transferring data between different > incompatible formats. > > A barrier to entry is nearly always a good thing, as long as it's not > too big. By creating this barrier to entry, documents won't be littered > with meaningless, or stupid tags (eg <fUnKyT3xT>, <myC00lstuFF>). People > who are interested enough to learn how to work with are are more likely > to consider the semantics of a document. > > Again, please correct me if I'm way off here. I can't say whether you're way off or whether I am way off, and perhaps neither or both of us are, but ... Coming from a typesetting-oriented markup background, my experience is that every document is unique. For this reason, I eschew LaTeX, use TeX, and create tags on the fly to reflect the actual structure and content of the document in-hand, rather than re-using (and abusing) the rather restricted set of tags which LaTeX would like me to use. When I come to mark up documents for the web, I find that exactly the same situation obtains : there are rarely any two documents for which the same tag set would suffice. Thus I would like the flexibility to mark up a web page with an arbitrary set of tags of my choosing, and I accept that the price to be paid for this is that I should define a mapping between /my/ tags and a minimal tag set defined as necessary and sufficient by the W3C. Since the two sets are clearly not in 1 : 1 relationship, I must additionally use CLASS (and, if necessary, ID and other) attributes to express the variations between each of my tags that map to one canonical W3C tag. And finally I must define a set of presentation rules such that the content delimited by my tags is presented in an appropriate way (which can, of course, be over-ridden by the viewer/reader/listener/..., should he or she so choose). Now, if I were defining a tag-set which I knew I would use over and over again, then there would be some (maybe considerable) merit in also defining a syntax for that tag set such that a validating browser could detect if I'd used a tag in an inappropriate context. But if the tag-set is essentially use-and-throw-away, then the overheads of defining that syntax /far/ outweigh any small advantage that might be otherwise obtain. And therefore what I am looking for (and maybe proposing, if it does not already exist) is a means of extending XHTML such that each and every document can use the set of tags that are relevant to that document, <stress>whether or not those tags have been pre-defined by the W3C</>, with a simple, easily understood, syntax for mapping the new tags to a combination of existing tags and attributes. Custom DTDs are fine for those who are making a significant investment in a re-usable tag-set; something /much/ simpler is, I suggest, what the average document author needs. Philip Taylor, RHBNC
Received on Tuesday, 13 August 2002 07:30:32 UTC