Sun ignored the rise of commodity hardware. They were relying on the Internet and Wall Street bubble in the late 90's. When those both popped, a Sun took a big hit.
Anyway, why does answering this help with Borland?
Sun also had an overpriced compiler, to go with the overpriced Unix running on overpriced hardware. Then it saw these replaced with free, free, and cheaper products, respectively.
Sun were, in many ways, the worst of the commercial Unixes, they just had a very loud fan base insisting they were the One True Unix. Consider how long it took to get a journaling filesystem, never mind volume management!
> Anyway, why does answering this help with Borland?
It would be interesting to explore whether or not there are any commonalities in terms of what happened to Sun, Novell, Borland, etc. And if there are, it might be interesting, in turn, to ask if any of those lessons would be useful to contemporary software companies (especially startups, given the audience here at HN).
I suspect that there are some common factors that could be found, but I don't have a good feel for what they are off-hand, aside from falling back on cliches or tautologies.
The common factor would have to be that they were squeezed by both Microsoft and Linux. Sun sold "Enterprise" UNIX until Linux-on-Intel was good enough. Novell sold networking until Windows+AD was good enough (round about the launch of Vista). Borland sold IDEs until free IDEs were good enough, and also traditional client-server dev went out of fashion along with a language migration.
