Wrapping up about KI18n and UIC

[Stephen Kelly]

Stephen Kelly wrote:

>> It'd need to be released quite a bit before us to
>> be something we can consider as a dependency. At that point I'm
>> considering having 2.8.12 as dependency for the release (so that it got
>> time to spread, sounds less likely with CMake 3).
> 
> I don't understand. Why is CMake 3 not likely to spread?

My point here was that suggesting with a wink and a nudge that CMake 3.0.0 
is highly likely to have lots of incompatibilities (as I guessed you were 
doing?) and therefore not spread is not appropriate. 

You have enough credibility that people would believe it and spread it, but 
it is not true. That's not how backward compatibility works in CMake.

Thanks,

Steve.