I think segmentation is the better term. It is used elsewhere, but in many places (ATM SAR immediately comes to my mind) and so IMHO only has overtones of TCP if you limit your knowledge to TCP and IP.
And, of greater or lesser import depending on your religious convictions, I recall that segmentation was the term chosen by OSI to describe this mechanism regardless of which layer it appeared in. Tom Petch -----Original Message----- From: Rainer Gerhards <[EMAIL PROTECTED]> To: Anton Okmianski <[EMAIL PROTECTED]>; [EMAIL PROTECTED] <[EMAIL PROTECTED]> Date: 13 February 2004 13:46 Subject: RE: Fragmentation terminology >Anton, > >very reasonable ;) > >I think "multi-part" is a good term. OK, it triggers "MIME" in my head, >but I think every term used to split up a message into multipe >parts/fragments/segemnts/whatever is already used in some other context. >As we don't do MIME in syslog, "multi-part" should cause the least >confusion. I am going to change this in -protocol-03 if nobody objects. > >Rainer > >> -----Original Message----- >> From: Anton Okmianski [mailto:[EMAIL PROTECTED] >> Sent: Friday, February 13, 2004 12:21 AM >> To: [EMAIL PROTECTED] >> Subject: Fragmentation terminology >> >> Rainer: >> >> If we keep the syslog fragmentation feature in -protocol, can >> we change >> its name to something that can't be confused with IP fragments. >> Otherwise, I think it can creates confusion when one talks >> about syslog >> message fragments. "Segments" I think is also a term that is already >> commonly used in TCP. >> >> Maybe "multi-part" message and "message part" instead? I >> need a way to >> refer to those pieces in -transport hopefully without having >> to explain >> what kind of fragment I am referring to. >> >> Thanks, >> Anton. >> >> >> >> >> > >
