reviewing a most fascinating paper about "failed web technologies", and thinking that we should all spend much more time with tech history.

Is there a common theme? I can think of several that have failed, but often for different reasons. It would be good to know what will fail before a bunch of people spend their time on it.

Too complex (DSSSL), implementation not tied to means of deployment (XHTML 2.0), no real user demand (XLink), no real developer demand (XSLT-FO), too comprehensive (WS-* specs), too slow and bad interop (CORBA), too proprietary (MS Behaviors), too early (VRML), insecure (ActiveX)

How about CC/PP (w3.org/TR/CCPP-struct-vocab/) and DDR (w3.org/TR/DDR-Simple-API/) completely overtaken by Responsive Web design? i.e. a smarter/simpler solution. And I can't not own up to my own failure (POWDER) - lots of reasons for that.

so many interesting and informative failures in this thread! makes me want to have a workshop on "failed stuff" at the next @TheWebConf.