学习.net还是MFC， 微软一专家给出了一个答案
I just posted an article on MSDN regarding the present and future of MFC. While there isn't anything earth shattering to report, we have received a number of customer inquiries about MFC, so we felt it was necessary to spell out to the VC++ community that there is still work being done on MFC. Granted, the work being done isn't major, but it's enough to keep MFC viable and build bridges to .NET and beyond.
While MFC is getting a bit long in the tooth compared to, say, WinForms, it remains viable because it still does a number of things that other application frameworks do not. In fact, it's kind of the only game in town (and by "town" I mean Redmond) when it comes to application frameworks. As long as this remains the case, I think folks can expect to see the VC++ team continue to service MFC. I think we're more likely to see targeted investments here (e.g. enabling WinForms hosting) as opposed to new big chunks of MFC infrastructure, as I think it's pretty obvious that MFC architecture is from an earlier era -- we've learned a lot since then, and it would be a better investment to incorporate what we've learned into something new instead of trying to signficantly retrofit MFC.
It's a bit frustrating to be in this in-between area, where we can see the great productivity of frameworks like WinFX on one side and the deep and wide feature set of MFC on the other. It's nice that bridges are being built between the two, but what we really need is a marriage of these worlds, where we get the productivity and the features. There are some research projects underway here to come up with a next generation application framework, but all pretty early-stage stuff, so we'll be in the in-between area for a while longer.