pointless?!? we've been writing about it for more than a hour: it
induces users on implementing race conditions!
{code}
if ( cacheService.containsKey( "key" ) )
{
// in the meanwhile someone else drops "key"...
Object retrieved = cacheService.retrieve( "key" ); // got null,
expected a value
}
{code}
http://people.apache.org/~simonetripodi/
http://simonetripodi.livejournal.com/
http://twitter.com/simonetripodi
http://www.99soft.org/
On Sun, Feb 26, 2012 at 10:08 PM, Michael André Pearce
<[email protected]> wrote:
> What is pointless is not exposing these methods in the CacheService, when it
> holds the map of keys and can so speedily expose these, without nasty
> hacking/work arounds.
>
> If the Cache implementation which is the user exposed cache you dont want to
> expose this i can understand.
>
>
> On 26 Feb 2012, at 21:00, Simone Tripodi wrote:
>
>> JSR107 will be supported not in the core API, but rather in an
>> external module - as a side note: if you are interested, you are more
>> than welcome on contributing it - that doesn't mean DM has to expose
>> that method.
>>
>> http://people.apache.org/~simonetripodi/
>> http://simonetripodi.livejournal.com/
>> http://twitter.com/simonetripodi
>> http://www.99soft.org/
>>
>>
>>
>> On Sun, Feb 26, 2012 at 9:58 PM, Michael André Pearce
>> <[email protected]> wrote:
>>> JSR107 specifies that you expose the method to user also.
>>>
>>> A user should always be aware that checkKey is a snapshot in time and if
>>> you need to it is for the developer to lock or single thread, and this is
>>> understood by many developers who implement cahce's which expose this.
>>>
>>>
>>>
>>> On 26 Feb 2012, at 20:50, Simone Tripodi wrote:
>>>
>>>> ou still need a shortcut? JSR107 integration module can
>>>> comfortably use that way to check keys.
>>>>
>>>> The use for a key check is potentially dangerous, as it may expire
>>>> between checking for the key, and whatever you want to do with the
>>>> stored object, and produc
>>>
>
