On Apr 21, 2010, at 6:53 PM, Chris Hegarty wrote:
> Max,
>
> Good catch to find this bug!
>
> Some comments:
> 1) I don't get why salt now becomes transient. I don't see that it has
> any effect on how the object is cloned and class is not Serializable.
Reversed. I just had a habit to mark all non-core data as transient, be it in
serialization or clone. Not necessary anyway.
>
> 2) You should be able to remove L128 in the new file. The cloned object
> will have same value for nameType, and since it is a primitive there
> shouldn't be an issue.
>
> 3) You should be able to replace the arraycopy with nameStrings.clone().
> The array elements are immutable Strings, right?
Correct!
Webrev updated at --
http://cr.openjdk.java.net/~weijun/6856069/webrev.01/
Thanks
Max
P.S. Oh I hate the latest Google Chrome browser removes http:// in the address
bar: have to manually add here.
>
> -Chris.
>
> On 21/04/2010 04:56, Weijun Wang wrote:
>> Hi
>>
>> Anyone can review this code change?
>>
>> http://cr.openjdk.java.net/~weijun/6856069/webrev.00/
>>
>> Thanks
>> Max
>>
>> Begin forwarded message:
>>
>>> *Change Request ID*: 6856069
>>> *Synopsis*: PrincipalName.clone() does not invoke super.clone()
>>>
>>> === *Description*
>>> ============================================================
>>> PrincipalName's clone() method does not invoke super.clone(), and it has a
>>> child class ServiceName. This means the clone of a ServiceName object is
>>> not of type ServiceName.
>>>
>>> See "Effective Java" Item 10.
>>>
>>> *** (#1 of 1): 2009-06-30 07:34:10 GMT+00:00 weijun.w...@sun.com
>>
