Update implementation

[dguo]

This update takes care of the situation where the storage is supported
by the browser, but it's not actually available.

See a more in-depth explanation and updated implementation at https://developer.mozilla.org/en-US/docs/Web/API/Web_Storage_API/Using_the_Web_Storage_API#Feature-detecting_localStorage

A reason not to merge this would be that someone might only want to know if the feature is technically supported and doesn't care if it's actually usable. It's hard for me to imagine such a case, but it's possible.

[Download]

I wrote that actual MDN section.

A reason not to merge this would be that someone might only want to know if the feature is technically supported and doesn't care if it's actually usable.

The current implementation proves it is usable and returns false in all other cases.

The updated implementation would return true in some circumstances even after setting and removing a key actually failed. So the statement that it more clearly indicates whether storage is actually usable is false. In fact it might return true even though writing a key failed.

The first lines of the code are getting the storage object from the window and calling setItem on it. These lines might actually already fail with exceptions before even attempting the write, because Firefox will throw exceptions in some cases when you access the storage object (without even calling a method on it yet). IE will in some case give you a null object which would cause the method call to fail as well. So you can get other types of exceptions in the catch block than just quota exceeded exceptions. So I simply do not agree with these changes. I think they are wrong.

[Download]

Thanks anyway, I do appreciate you are trying to contribute. But people on MDN keep changing a perfectly fine code example and in many cases are actively breaking it in the process so I sort of gave up on trying to maintain the MDN docs.

[dguo]

Roger that. Thanks for the detailed explanation.