RFR : 8003228 : (props) sun.jnu.encoding should be set to UTF-8 [macosx] (original) (raw)

Brent Christian brent.christian at oracle.com
Fri Dec 21 18:25:04 UTC 2012

On 12/21/12 7:33 AM, Alan Bateman wrote:

I also think it's important that the folks on macosx-port-dev that have been complaining about this help us test it out, particularly folks that work in Asian locales. I don't know how many of them build the jdk so you might need to send a reminder once there is a jdk8 build available with this change. I don't think we should consider backporting this change to jdk7u until there is strong evidence that the change in jdk8 is good.

Definitely agree.

Thanks for creating a test. I'm not sure that test/java/util/Properties is right place as that directory is for tests of the java.util.Properties API. I guess I would move it to java/util/System as there isn't an obvious place in the sun/** tree.

I wasn't quite sure where to put it. Moved to java/lang/System (assuming you didn't actually mean java/util/System).

In ExpectedEncoding then it looks it is missing an exit in the bad usage case.

Thanks, fixed.

Also as an alternative to the failed flag then you could just throw an exception.

When practical, I like to test and report all tested conditions (instead of failing fast). I think it helps provide scope to a failure, just from looking at the test report. Bit of a personal preference, I guess. I can change it to fail fast if that's preferred.

I do think an exception stack trace is easier to spot to someone just glancing through test results. So I now have the flag and an exception

Updated webrev: http://cr.openjdk.java.net/~bchristi/8003228/webrev.02/

Thanks, -Brent

More information about the core-libs-dev mailing list