Re: Initialization and ProtectionDomain/String

[Brian Jones]

Mark Wielaard <address@hidden> writes:

> Hi,
> 
> On Tue, 2002-09-24 at 18:00, C. Scott Ananian wrote:
> > The recent change to java.lang.String (aug 13, v1.50) add a call to
> > Character.zeroBasedStringValue() which (inadvertently?) makes
> > java.lang.Character part of the "CORE CLASS" initializers (because String
> > is a core class, and the static initializer of String invokes a method in
> > Character, causing the initialization of Character).  Character should
> > be either marked as a CORE CLASS or the shared method should be moved
> > to java.lang.String.
> 
> Thanks for catching that. I had tested it with Kissme that didn't seem
> to have a problem with the initializer order. But I have committed your
> patch since it makes the implementation behave as documented in
> Runtime.java.

Did you move zeroBasedStringValue() too?

Brian
-- 
Brian Jones <address@hidden>