RFR: 8169068: Add a new method: java.net.Authenticator.getDefault()

Chris Hegarty chris.hegarty at oracle.com
Thu Nov 3 18:30:03 UTC 2016


On 03/11/16 17:54, Daniel Fuchs wrote:
> Hi Chris,
>
> Thanks a lot for the feedback.
>
> Here is a new webrev incorporating your comments:
>
> http://cr.openjdk.java.net/~dfuchs/webrev_8169068/webrev.01

Thanks Daniel, this looks good.

-Chris.


> -- daniel
>
> On 03/11/16 14:40, Chris Hegarty wrote:
>> Daniel,
>>
>> On 03/11/16 10:20, Daniel Fuchs wrote:
>>> Hi,
>>>
>>> Please find below a patch for:
>>>
>>> RFR: 8169068: Add a new method: java.net.Authenticator.getDefault()
>>> https://bugs.openjdk.java.net/browse/JDK-8169068
>>>
>>> The method implementation itself is trivial.
>>> The API documentation is derived from that of Authenticator::setDefault
>>>
>>> The method checks for the existing
>>>   NetPermission("requestPasswordAuthentication")
>>> rather than introducing a new "getDefaultAuthenticator" permission,
>>> because getting the authenticator actually grants you the possibilty
>>> to request passwords - so it's better that the permission check
>>> makes it explicit.
>>>
>>> http://cr.openjdk.java.net/~dfuchs/webrev_8169068/webrev.00/
>>
>> I am happy to see this being proposed. Just a few minor comments:
>>
>>  1) "Get the ... " -> "GetS the ..."  (personal preference)
>>
>>  2) The wording with respect to the permission check is little odd,
>>     but it is consistent with elsewhere in this class, so that's
>>     fine.
>>
>>  3) trailing ';' on L34 of the test ;-)
>>
>>  4) 'count' seems to be unused in the test. If you need it maybe make
>>     it volatile rather than locking on the instance in getCount().
>>
>>  5) It might be a little less obscure to just print the value of the
>>     defaultAuth rather then using a reachability fence?
>>
>> -Chris.
>


More information about the net-dev mailing list