On Jun 6, 2006, at 11:22 AM, harvey wrote:

> SoftQuad's Panorama, Synex and one other (can't remember the name now)
> were based on the same technologies.

Yes. Magnus & Co. wrote some very nice code (in terms of structure/ 
object breakdown).

> It was because of the experience with Panorama that prompted the
> requirement for DTD to be dropped in XML.

I think that was one data point. DynaText was another, as was 'SGML  
in the wild' where in a lot of cases (from my experience anyway), the  
DTD was only necessary at certain points in the process... documents  
transitioned in and out of a valid form depending on the phase in  
processing (Rick really drove that home to me in some discussions we  
had... and the idea showed up in Topologi in spades too).

> The really cool thing about these documents was the use of HyTime.

HyTime was cool, but overly complex for many applications. Even the  
simplest decoupling of the markup from the interaction semantics (as  
all the early systems had) would have resulted in a much different  
web. Abominations such as DHTML would never have appeared.