[OpenJDK 2D-Dev] Turn on Xrender by default in JDK 8

Jim Graham james.graham at oracle.com
Tue Aug 16 19:24:21 UTC 2011

We should normalize all of these via a helper function.  All combinations of

     false, f, False, F, true, t, True, T

should work for all boolean properties and the upper case versions 
should all be verbose.  Arguably we could easily allow any number of 
letters (so that "fal" works) by just using the "startsWith()" method 
instead of a direct equals() comparison.

Also, if they spell the property name correctly, but we don't recognize 
the value - shouldn't we print out a complaint?


On 8/12/2011 3:08 PM, Phil Race wrote:
> Yes, "False" is a no-op. You need "false" .. I did almost add "False"
> support,
> but I was thinking of it for printing a message that the pipeline is
> disabled,
> however so far as I could see, no pipeline prints anything for "False",
> even
> if it understands it, which may only have been the case on Windows, and
> all mis-spellings are silently ignored. Of course there's only so much we
> can do here. If the user mis-spells the system property for example, but
> I agree it would be nice to have some way to know that your request got
> through.
> I have a separate bug to clean up the pipeline verbose reporting, and
> update user docs. I'll revisit this then.
> -phil.
> On 8/12/11 2:25 PM, Jim Graham wrote:
>> You probably should have updated the verbose half of the tests while
>> you were inverting the logic. It's only a debugging tool, but right
>> now there is no way to verbosely turn it off like there used to be a
>> way to verbosely turn it on. Now if someone uses "False" or "F"
>> thinking that this will verbosely turn it off, they will have no
>> effect...
>> ...jim
>> On 8/10/11 12:03 PM, Phil Race wrote:
>>> Clemens,
>>> I think we should turn on Xrender by default in these early days in
>>> JDK 8
>>> http://cr.openjdk.java.net/~prr/7077423/
>>> The change itself is simple .. any concerns ?
>>> -phil.

More information about the 2d-dev mailing list