right Jörn, good catch :-)

Tommaso

2011/9/30 Jörn Kottmann <[email protected]>

> hashCode will limit the ID to 32 bit, that will for example not work
> with files names, URN based IDs, etc.
>
> Jörn
>
>
> On 9/30/11 4:02 PM, Tommaso Teofili wrote:
>
>> Just putting down some quick thoughts.
>> I think it would be something very useful because, especially in those
>> cases
>> where one serialize / deserialize lots of CASes, I've also had such a need
>> for a (predefined) CAS ID several times.
>> As this would be an important change (maybe also in terms of the OASIS
>> specification) we should pay attention to committing such a change.
>>  From the code point of view we could reimplement the hashCode() method
>> (this
>> could also not need an API change) of current implementation classes and
>> 'suggest' to use that as a CAS ID.
>> My 2 cents.
>> Tommaso
>>
>> 2011/9/30 Jörn Kottmann<[email protected]>
>>
>>  Hi all,
>>>
>>> is there a reason why the CAS does not define an ID field which is
>>> always there independent of the type system?
>>>
>>> Jörn
>>>
>>>
>>>
>

Reply via email to