FWIW, I agree with you to an extent. In experimenting with the eT compiler over the w/e, our lead developer effectively rewrote some stuff in NET. It was easier than tinkering with the existing code to make it compile. But this was for a fairly simple task (managing an ini file/registry and writing records to and from database.)

I have noticed in recent years that many of these tools focus on "making the easy easier" - IOW they automate tasks that weren't (or shouldn't be) so difficult to start with. Some of these tools (e.g. the view designer in VFP) have their own flaws that can make them more trouble than they are worth, creating opportunity for other companies to offer add-ons and frameworks to overcome the flaws. But IMHO what developers really need are tools to make difficult things easier. FWIW I thought L2SQL had a lot to offer wrt complex sql. Sure it screwed up sometimes but other times it produced some astonishingly efficient SQL in a fraction of the time it would take to dream that up yourself. I'm talking multi-table and self-referential stuff here- the sort of thing you'd use multiple cursors for in VFP or SP. On at least one occasion, L2SQL came up with a large efficient response that worked better than the version created at substantial cost by a SQL guru. It's a shame that MS has effectively sidelined L2SQL but I guess we have to assume that something better will follow in due course.

For now, eT's promise that most existing data munging can be compiled across to NET and may offer substantial speed improvements once compiled to x86/x64, is making quite a few difficult tasks easier. Obviously it allows immediate functional modernization of existing code that may represent years of effort. It also raises the prospect of performance improvements. Neither is guaranteed if people decide to rewrite from scratch. QED.