- From: Anish Karmarkar <Anish.Karmarkar@oracle.com>
- Date: Thu, 11 Sep 2003 15:20:31 -0700
- To: noah_mendelsohn@us.ibm.com
- Cc: Mark Nottingham <mark.nottingham@bea.com>, "John J. Barton" <John_Barton@hpl.hp.com>, "Xml-Dist-App@W3. Org" <xml-dist-app@w3.org>
noah_mendelsohn@us.ibm.com wrote: > ANish Karmarkar writes: > > >>I am using the term 'streaming' in a difference sense than you are. >>I was using it from the point of view of applications APIs (i.e. >>streaming APIs) and not as a unending broadcast stream. > > > But I think these are related. If your "broadcast" stream is of modest > size relative to the link speed, then streaming APIs are not particularly > necessary. To make the point, consider and extreme case: you wouldn't > want lots of streaming APIs for a 10 byte message "hey, I got 2 more > bytes." That's slower than a single call that says "I got all 10". The > deep reason for streaming APIs, IMO, is to deal with very large messages. > Agreed. The message size is important here. I was trying to elaborate what I meant by "steaming" to distinguish with what John was envisioning, which was, streams of unending broadcast data which does not go over TCP/IP. -Anish --
Received on Thursday, 11 September 2003 18:24:40 UTC