Why CTFE is context-sensitive?

[Ola Fosheim Grøstad" <ola.fosheim.grostad+dlang at gmail.com>]

On Tuesday, 28 January 2014 at 17:01:24 UTC, Ola Fosheim Grøstad
wrote:
> On Tuesday, 28 January 2014 at 16:29:24 UTC, Timon Gehr wrote:
>> _Now_ you mean finite. :o)
>
> I just tried to make things easy…
>
>> (In any case, there is even an infinite number of programs 
>> that are not behaviourally equal but can be automatically 
>> proved to terminate by one and the same algorithm.)
>
> Oh… don

Woops. The program terminated without a proof! :-]

I am sooo not going to start on a rant on program transformations 
and the relevance to the real world. Let's just agree that real 
world compilers are pragmatic and that you can write tractable 
programs if you want (like choosing a functional programming 
style).