OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [xml-dev] Here's a good question





> -----Original Message-----
> From: Tom Bradford [mailto:bradford@dbxmlgroup.com]
> Sent: Thursday, September 27, 2001 2:19 PM
> To: xml-dev@lists.xml.org
> Subject: [xml-dev] Here's a good question
> 
> 
> Why does the DOM suck so badly?  I'll leave the many reasons for the
> question to your imagination and cumulative experience.

Uhh, if I didn't know of Tom's impeccable reputation as a man of refined
manners and subtle eloquence, I might be offended.  But I guess I'll plead
ignorance and ask for a clarification:  explain what you mean by "suck".

If it "sucks" because it doesn't exploit all the power of Java
classes/collections/Strings/Lists/etc." ... well, duh, it's
language-neutral!

If it "sucks" because it's full of kludgy compromises between browsers and
editors, HTML and XML, trees and lists, clients and servers, and between the
various competitors who have drafted it ... well, duh! It was written by a
committee reprsenting all these incompatible viewpoints ... what do you
expect?  

If it "sucks" because it exposes XML 1.0's dirty linen (external parsed
entities, CDATA sections, the bizarre differences in whitespace processing
in validating and non-validating parsers, the overwhelming number of ways
namespaces find to confuse people ... you know my litany) ... I guess the
DOM sucks because life sucks :~)

If it "sucks" because things that are obvious in 2001 weren't obvious in
1997-1998, well duh! If it continues to suck because the W3C is very
reluctant to introduce backward incompatibility, that's a good point to
debate.

If it "sucks" because of all the things it doesn't have ... well, the
feeling at the time was that a DOM Recommendation that didn't come out in
time to be supported by IE5 and  (the non-exisitent, but we didn't know that
at the time) Navigator 5 would be DOA in the Real World.  There seemed to be
a narrow window of opportunity to get a Dynamic HTML API out in time to be
supported by the browsers, and an XML API in time to be supported by all the
XML tools that were being planned. For better or worse, the DOM hit that
window.

The DOM (like XML) is not a triumph of elegance; it's a triumph of "if we do
not hang together, we shall hang separately."  At least the Browser Wars
were not followed by API Wars. Better a common API that we all love to hate
than a bazillion contending APIs that carve the Web up into contending
enclaves of True Believers.

 
> Is it all Mike Champion's fault? :)

Probably.

All I can say is that the experience has been immensely humbling.  If anyone
has ideas for a better API that meets DOM's fundamental constraints, that
would be a very interesting discussion to have here.