Technology with a
great future behind it.
Push technology is just Usenet all over again, but poorly designed to allow only a few data sources and not have a proper distributed system to transfer the content. Of course, when "push" was invented, there wasn't one extant system on the Internet which could push content.
There were 2. One was called email, and could "push" per-subscriber content, using SMTP; the other was called Usenet, and could "push" the same content to multiple subscribers in a highly distributed manner, using NNTP. That the people designing (and talking about) push were unaware of these 2 ancient "technologies" is amazing; that anybody poured money into it is unbelievable.
But still, 3 companies were strong competitors in this market (they were competing to get in, not out, as one would expect): PointCast, Marimba, and BackWeb. Of the 3, Marimba still has some viable product: distributed software installation.
The concept of turning the Internet into a TV failed. It's not clear if it failed because people didn't want InterneTV (or whatever you'd want to call it), or because of the technical dimness of the proponents, and the highly closed protocols they used. Unfortunately, I tend to believe the second -- after all, today you can "personalize" CNN.COM, NYTimes.com, and even Slashdot, and have them push (sorry, send) you updates and daily news...