Rinat Abdullin

Crazy times, crazy times. I’m highly interested to see who ends up over there next. Will it be another of the IoC folk? Someone from the unit testing realm? Someone from the mocking space? (many seem to already be wondering if Oren is on his way to MSFT, for example.)

Should be interesting to see how this all plays out over the coming years.

I guess, that there will be some improvement of the .NET for all the developers in general. But the scissors will not get sharper.

In the computational mathematics there is that “There are no free lunches” theorem. Simply put, it means that the specialized solution is always better in solving specific problems (that it is optimized for). And the generic solution is just good on the average.

Same could be applied to the development. MEF is the generic extensibility implementation. Autofac/Castle - specific IoC containers good for the specific scenarios (same with the mocking, testing etc).

So in general it would be easier for us to develop. But sharp and highly specific scenarios would still require sharp scissors.

What is painfull,

is that people are not working to extend their original frameworks, but now working at someoneelse’s idea. Now we have CTPs…
My personal problem is: should i invest my time to use CTP, should i use older but stabler stuff like Windsor, StructureMap, etc?
Or should i just write my own little IOC container taylored by my specific needs.