<AWT Dev> Review Request for 6879044

[Oleg Sukhodolsky]

On Thu, Sep 17, 2009 at 12:56 PM, Alan Bateman <Alan.Bateman at sun.com> wrote:
> Oleg Sukhodolsky wrote:
>>
>> On Tue, Sep 15, 2009 at 11:13 PM, Mandy Chung <Mandy.Chung at sun.com> wrote:
>>
>>>
>>> :
>>>>
>>>> complicate porting JRE implementation.
>>>>
>>>
>>> How does it complicate the porting?
>>>
>>
>> I'm not sure that IBM's or some other's version of JDK is allowed to
>> contain such
>> classes, thus it may be harder to port our RI to their implementation .
>>
>
> I don't see a problem here. These are implementation classes (you'll see
> that AWT already makes use of lot of implementation classes from sun.awt,
> sun.security, sun.java2d, and more). Furthermore, these changes aren't
> introducing any platform dependent or native code that increases porting
> efforts. If there are ports that already remove these loggers then the
> effort, once Mandy's changes are in, isn't any different.

Ok, they just add one more issue to solve for porters :)

>> Why we have to remove all usages of logging in our code instead of
>> changing logging package to be
>> more startup friendly?
>>
>
> I haven't seen any proposals to eliminate the logging but rather the
> suggestion is that this logging should be re-examined because there are way
> too many loggers created at startup. For example, one of the suggestions
> that Mandy has put in 6880089 [1] is that there be a logger per core
> component rather than class.

By "remove all usages of logging" I meant "remove all usages of
standard logging package".
And, again, I do not understand why we can not change the logging
package instead (perhaps we can not,
but no body has explained why).  BTW if we can change the logging
package to resolve this problem for our code,
such changes will also help to other java developers.

Also since it look like all suggested changes are about startup
performance it would be nice to see results of
performance testing which would prove that suggested changes will
help, and show what kind of applications
will benefit from the changes.  Without such results we can not be
sure that the changes will help and
they can be treated as "premature optimization"

Oleg.

>
> -Alan.
>
> [1] http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6880089
>
>
>