activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Timothy Bish <tab...@twcny.rr.com>
Subject Re: char signedness
Date Fri, 08 Jun 2007 14:43:04 GMT
There really isn't any interoperability issues, if you want a signed 
byte on the C++ side just cast and you are done. 

Motl wrote:
> I do agree 'unsigned' is much more suitable for 'byte'. But the problem here
> is that we can't have Java-to-C++ interoperability if we leave different
> signedness for 'byte'. If it's not critical,  okay.
>
>
> Hiram Chirino wrote:
>   
>> I vote keep it unsigned.  Signed bytes in Java were a mistake IMO.
>> Almost every time I have to work with bytes in Java I have to do the
>> "& 0xFF" tricks to turn it unsigned.
>>
>> Regards,
>> Hiram
>>
>> On 5/25/07, Motl <motl@orcsoftware.com> wrote:
>>     
>>> At the moment, 'unsigned char' is used for BytesMessage representation in
>>> C++,  whereas signed char is used in Java, that means the same message
>>> will
>>> be encoded differently in Java and C++! I agree that 'unsigned char' (or,
>>> precisely,  uint8_t) is more suitable for 'byte' type in C/C++, but if we
>>> follow JMS standard, we should use 'signed char' instead. The
>>> representation
>>> of signed chars isn't a part of ANSI C Standard, but most compilers use
>>> 2's
>>> complement, and so Java does.
>>> Hereby, I propose to replace unsigned char with signed one.
>>>
>>> --
>>> View this message in context:
>>> http://www.nabble.com/char-signedness-tf3814717s2354.html#a10798606
>>> Sent from the ActiveMQ - Dev mailing list archive at Nabble.com.
>>>
>>>
>>>       
>> -- 
>> Regards,
>> Hiram
>>
>> Blog: http://hiramchirino.com
>>
>>
>>     
>
>   


Mime
View raw message