I agree with Ashish.  This provides the most flexibility.

On 5/5/07, Maarten Bosteels <[EMAIL PROTECTED]> wrote:
On 5/5/07, Ashish Sharma <[EMAIL PROTECTED]> wrote:
>
> i think have a constant size header which will contain payload length
> for rest of message.


I agree, this is very easy to implement.


On 5/5/07, mat <[EMAIL PROTECTED]> wrote:
> > Can anyone help me out with my protocol design? Is it better to design
> > protocol as a fixed length or non fixed length? What I meant is fixed
> lenght
> > message(for instance 512bytes) for each message? Non fixed length means
> the
> > message size could be big. Which method is better? If fixed length, i
> faced
> > a problem. How to reassembly the message since message is cut into the
> > piece?
> >
>



--
..Cheers
Mark

Reply via email to