I (Charles B Cranston) wrote:
> Michael Sierchio wrote:

It cannot be emphasized more clearly: TCP is a byte stream protocol.
This is quite true.

... There is NO WAY in TCP to indicate in
an out-of-band way that there is a 'record break'.
This is not quite true.  You can certainly send OOB data
via TCP.  Urgent data are read out of byte order, if you
like.  This obscures your previous point, which is that TCP IS
A BYTE STREAM PROTOCOL.  Ahem.

Yes, but the very point is that the OOB data is read OUT OF BYTE ORDER so it is impossible (difficult?) to use it as some kind of out-of-band end-of-record delimiter.

The point is to stop thinking in a record-oriented fashion and
start thinking in a byte-oriented fashion.

Maybe you can kluge something up, but at the time I learned the
TCP/IP stuff (which was admittedly some years ago) not all TCP
implementations even IMPLEMENTED out of band data, so we were
taught not to assume that it would be available.

--
Charles B (Ben) Cranston
mailto: [EMAIL PROTECTED]
http://www.wam.umd.edu/~zben

______________________________________________________________________
OpenSSL Project                                 http://www.openssl.org
User Support Mailing List                    [EMAIL PROTECTED]
Automated List Manager                           [EMAIL PROTECTED]

Reply via email to