<Swing Dev> [PATCH] 6179357: Generics: JList
Pavel Porvatov
Pavel.Porvatov at Sun.COM
Thu May 14 13:39:32 UTC 2009
Hi Florian,
Some time ago we discussed two different ways to add the new method
"getSelectedValuesList()". The first one is the current implementation
in your fix:
1. public List<E> getSelectedValuesList()
Benefits:
a. easy to use
b. Collection framework is very flexible
The second way is:
2. public E[] getSelectedValues(E[] a)
Benefits:
a. The same pattern used in the java.util.Collection#toArray(T[] a)
method
b. a little bit easer to port an old code
Also a lot of people noticed that there is no need to deprecate the
javax.swing.JList#getSelectedValues method because it still can be useful.
Any ideas?
Regards, Pavel
> Hi Pavel, hi Alexander,
>
> I'm back from holiday.
>
> What is the status of this patch? Any news?
>
> -Florian
>
> Alexander Potochkin schrieb:
>> Hello Florian
>>
>>> Great! :-)
>>>
>>> In the case of other issues please note that I'm on holiday until the
>>> end of next week.
>>
>> Have a nice holiday!
>>
>> (I am reviewing the fix right now)
>>
>> Thanks
>> alexp
>>
>>>
>>> -Florian
>>>
>>> Pavel Porvatov schrieb:
>>>> Hi Florian,
>>>>
>>>>> Hi Pavel,
>>>>>
>>>>> I agree that we should avoid to mix several different fixes in one
>>>>> fix, but since in this fix we change
>>>>>
>>>>> AbstractListModel to AbstractListModel<E>
>>>>> and
>>>>> JList(ListModel dataModel) to JList(ListModel<E> dataModel)
>>>>>
>>>>> I think we should also change usages of AbstractListModel such as
>>>>> "this (new AbstractListModel()...)" to "this (new
>>>>> AbstractListModel<E>()...)" to avoid warnings.
>>>> Ok, then it will not be a problem. Let's include this change in your
>>>> fix. Therefore all my comments are gone and I'm going to start our
>>>> internal process to commit your fix.
>>>>
>>>> Thanks, Pavel.
>>>>
>>>>>
>>>>> If you don't agree:
>>>>> There are several places where I changed the usage of now
>>>>> generified classes to the generic variant. Which ones should I
>>>>> change back? Only this case?
>>>>
>>>
>>
>
More information about the swing-dev
mailing list