Comment by dcminter
3 days ago
Or my favourite...
Short w = 42;
Short x = 42;
out.println(w == x); // true
Short y = 1042;
Short z = 1042;
out.println(y == z); // false
3 days ago
Or my favourite...
Short w = 42;
Short x = 42;
out.println(w == x); // true
Short y = 1042;
Short z = 1042;
out.println(y == z); // false
Once, after we had an application go live, we started getting reports after a few hours that new users were unable to log in.
It turns out, somewhere in the auth path, a dev had used `==` to verify a user's ID, which worked for Longs under (I believe) 128, so any users with an ID bigger than that were unable to log in due to the comparison failing.
I’ll bite. Why does this not work as you’d expect?
I'm comparing object references (pointers) not the value boxed by the object (what the pointer points to).
For performance reasons boxed Short objects are interned when they represent values in the range -127 to +128 so for 42 the pointers will point to the same interned object after 42 is autoboxed to a Short. Whereas 1042 is outside this interning range and the autoboxing creates two distinct objects with different pointers.
It's very simple but (a) non-obvious if you don't know about it and (b) rather wordy when I spell it out like this :)
In general in Java you want obj.equals(other) when dealing with objects and == only with primitives, but autoboxing/unboxing can cause confusion about which one is dealing with.
In other other words, the surprise ought to be that w == x is true, not that y == z is false!
That's another gotcha-- interning of strings and boxed primitives.
Are there linters for this sort of thing? I don't write Java much any more.
> Are there linters for this sort of thing?
Yes and they're pretty good so it's rarely an issue in practice. Using == on object references will indeed usually get you yelled at by the linter.
I’ll say from experience that even if your IDE highlights that by default people won’t pay any attention to it and the bug will get in.
I’ll be happy when it’s fixed.
If JEP 401 is ever delivered (Value Classes and Objects), then this sort of problem should go away.
Do Java problems go away? I thought the selling point was that your huge un-rewritable enterprise software will crash tomorrow like it crashed yesterday.
Here's a talk from Netflix (hopefully sufficient enterprise for the discussion) that goes over how a JDK version upgrade to generational ZGC improved a bunch of their request timeouts: https://youtu.be/XpunFFS-n8I?si=XG6zYYZy50sfNE4j
This is nuts lol