I also added a type trait test to check whether the conversion operator is marked explicit.
this change has the advantage of being clearer imo as it is a conversion operator for conersions not a constructor.
and also it allows users to take the address of convesion operator which I don't know if someone might find that useful.
As a rule of thumb, if the type is less than or equal to the CPU
register width times two then you ought to pass it by value. This
will lead to more efficient code generation.
Similar renaming for Font::loadFromMemory and Font::loadFromStream.
The goal is to better express the need to keep the source available,
similar to Music::openFromFile for example.
It's rare that a type truly needs a .reset function. Copy/move
assignment typically accomplishes the same thing with less code
and is easier to maintain since it doesn't require updating your
.reset() function as new data members are added.
To reset a type is conceptually the same thing as simply assigning
from a newly constructed instance of the same type.
This change was made in 359fe90 due to recommendations from tooling.
On its face this change makes sense since it removes a copy that
isn't always necessary. In practice it caused ergonomic issues due
to now being forced to make a copy of the render states when needed.
The performance gains of eliding this copy are unsubstantiated. We
have not done any profiling to measure its impact. For lack of such
measurements I'd rather err on the side of improved user experience.
If future benchmarks prove this copy is rather expensive then we
can reconsider removing it with that evidence in mind.
sf::RenderWindow still inherits a virtual destructor from a base
class so there's no need to explicitly declare a virtual destructor.
I added a test to ensure this property was not broken.
