> On Oct 23, 2020, at 3:57 PM, Igor Ignatyev <iignat...@openjdk.java.net> wrote:
>
> On Fri, 23 Oct 2020 19:25:27 GMT, Harold Seigel <hsei...@openjdk.org> wrote:
>
>> I think it depends on whether the tests will be permanently or temporarily
>> excluded from running with containers. I thought this mechanism would be to
>> permanently exclude the tests. That's why I used `@requires`.
>
> I see, if this is for permanent exclusion then yes I agree that `@requires`
> is a better choice.
>
>>> enable this option based on an environment variable so we don?t have to
>>> remember the
> cryptic command line sequence.
>> I'll look into basing the option on an environment variable.
>
> one will still need to pass an environment variable to jtreg, and hence will
> need to remember some sort of "cryptic command line sequence". a solution for
> that might be to default `jdk.containerized` to `false` in `VMProps.java` and
> when only _containerized_ runs will have to set it up.
Why? Environment variables are inherited. For developers running jtreg, all
they need to do is export the variable.
% export JAVA_CONTAINERIZED=1
% bash
% echo $JAVA_CONTAINERIZED
% 1
Bob.
>
>
> btw, I'm not sure that `jdk.containerized` is the best name for this property
> as _containerization_ is more of an environmental characteristic than that of
> jdk. how about smth like `env.containerized` or `testenv.containerized`?
>
> -------------
>
> PR: https://git.openjdk.java.net/jdk/pull/844
