But that was true for vfp as well - IMHO vfp3 was just as much a testing balloon as .Net 1.x -
interesting technology but not really good/ready to base critical apps on.

I'm fairly sure I was involved in the first VFP app (which by definition was also the first non-MS Windows 32-bit app) released after VFP was officially released in August 1995. ;-) Remote data handling was far better than the DDE/connectivity kit. The UI offered many features "out of the box" that had to be concocted manually in FPW2.x. What about proper toolbars to match the look and feel of other contemporary windows apps. Clever frontends made possible by multiple Datasessions. The grid. Dynamicbackcolor and other dynamic features in the grid. Tooltips. Mouse and other UI events. Sure VFP3 had bugs but it carried lots of cool extra features that translated directly into stuff customers wanted.

The *real* trouble is that hardware speedup hasn't kept up with software bloat after PIV2.5 Ghz or Amd XP 2.0: in the last 6 to 8 years we definatly did not see the speedup to be formerly expected as a by-law to Moore. ALWAYS blame the HW <bg>

I never minded upgrading hardware as long as I saw real advantages coming through- not just fatter software but extra features and benefits that make a business difference.