On Tue, 3 Dec 2024 13:43:25 GMT, Alan Bateman <al...@openjdk.org> wrote:
>> A few additional changes should be made to the API specs for these classes .
>> These changes are documenting behavior of other APIs that has already been
>> approved as part of JEP 486, so I don't think a CSR is necessary, but
>> opinions on that are welcome.
>>
>> The `DomainCombiner` and `SubjectDomainCombiner` classes contain wording
>> about "inherited" and"privileged" `ProtectionDomain`s which is no longer
>> implemented, and so has been removed. The `PrivilegedAction` classes also
>> contained wording about "privileged" computations which is no longer
>> accurate. I also added an API note to these classes, similar to the ones in
>> the `Permission` subclasses. Finally, I removed a couple of instances of
>> "privileged work" from the `Subject` class which no longer applies.
>
> src/java.base/share/classes/javax/security/auth/Subject.java line 456:
>
>> 454:
>> 455: /**
>> 456: * Perform work as a particular {@code Subject}.
>
> The new callAs is specified as "Executes a Callable with subject as the
> current subject". ScopedValue uses "Runs an operation with each scoped value
> in this mapping bound to its value in the current thread" and "Calls a
> value-returning operation with each scoped value in this mapping bound to its
> value in the current thread".
>
> Looking at the doAs and doAsPrivileged methods makes me wonder if "Perform
> work" should be replaced rather than just dropping the word "privileged".
> That is, maybe these methods should are re-worded to specify that they run an
> action rather than "perform work". I realise this is more than what this PR
> is about, I'm just observing that "Perform work" is strange to see in the API
> docs.
Well the "perform work" words have been there for a long time. Perhaps
"Performs the specified action as a particular subject" which would be more
consistent with `AccessController.doPrivileged`. But, I'm also wary of making
too many changes.
-------------
PR Review Comment: https://git.openjdk.org/jdk/pull/22500#discussion_r1867783030
