On 2021-09-02 11:21, Francois Gouget wrote:
On Wed, 1 Sep 2021, Saulius Krasuckas wrote: [...]
If you remove it, a guy working on some old app (that misbehaves using Wine) will possibly try to add it back. Because one will not know the API crashes on newer Windows.
It's not clear that there ever was a Windows version where the API did not crash when given a NULL input pointer.
That's why I call for saving the documented state (for possible research in future).
Note that the NULL pointer being tested is an _input_ pointer (though that's not entirely obvious from the call). There is little the API can do without that single input.
I agree. But for example when I tried to extend the LZOpenFile[AW] tests, I went for testing all combinations of wrong arguments I could imagine (still unfinished). That's because this way I could know more about the internal flow/branching of the black-boxed code.
I mean, if some questions arises about when that input pointer is handled (before other argument validation or maybe after that, or maybe at the end), the crash (and maybe additional crashes) would give some insights about the original algorithm.
Especially if the function works in some old native installation (like I made up before). Which we cannot test directly anymore of course, since Win2000 is already retired from the testing infrastructure.
S.