3 ms·
> simultaneously both that "the standard is not wrong, it can not possibly be wrong" and that "the standard should be improved". I'm not the one saying that a
by JohnFen 3y ago
> simultaneously both that "the standard is not wrong, it can not possibly be wrong" and that "the standard should be improved".
I'm not the one saying that a standard cannot possibly be wrong. However, it is entirely coherent to say that a standard is not wrong and yet there is still room for it to be improved. "Not ideal" isn't the same as "wrong".
> And how do you improve standards anyway, without breaking compatibility?
Lots of ways, depending on what you think should be improved. In this case, you can improve it by adding a new set of functions that behave as you wish rather than altering the old functions in an incompatible way.
If it's not possible to improve a standard without maintaining backward compatibility, then the solution is to introduce a new standard entirely. One of the main points of even having a standard is that you can rely on it, and things you make that adhere to it will continue to work into the future.