Manjula, How about HTTP 1.0 case ? I think we should still support HTTP 1.0 and it does not support chunking IIRC. Also what happens in SMTP case (once/if we have it in place) ? I think MTOM implementation should be somewhat (not completely) transport independent and if the underlying transport supports whatever features (chunking, compression ...etc) we should use it. But we shouldn't make it mandatory to use features that are specific to some transports.
Just my 2c Thanks -sahan On 7/3/08, Manjula Peiris <[EMAIL PROTECTED]> wrote: > Hi, > > I was able to fix MTOM caching stuff for both sender and receiving ends. > In sender side instead of loading the whole attachment it will send the > attachment + data as chunked. Only very small portion of the attachment > will be loaded in to the memory at any given time. In order to achieve > this I did some axiom level changes and transport level changes.Because > the way of constructing what to send when it is chunked and not chunked > is far different I propose we send MTOM only with chunking. So when > sending MTOM Axis2/C will automatically switch on chunking. Remember in > the receiving side it will accept both chunked and dechunked data. WDYT? > > Thanks, > -Manjula. > > -- > Manjula Peiris: http://manjula-peiris.blogspot.com/ > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
