Hi, The 'MINA In Real Life' ApacheCon EU 2009 conference slides say that only one codec filter should be used. Is this for performance reasons or is it technically impossible to use more than one?
The reason I ask is that I am looking at the feasibility of implementing a MINA TDS (http://en.wikipedia.org/wiki/Tabular_Data_Stream) server. TDS uses PDUs with an 8 byte header including type, status, length. The maximum size of the PDU is negotiated between client & server at the start of their dialog. The payload of the PDU usually contains one or more tokens (e.g. a SQL command). A token can be of various types, can be fixed, variable or zero length and can span multiple PDUs. So this is a two level (or nested) protocol and I my first thought would be to implement 2 codecs - one to handle the lower level PDU and the other to handle the payload(token). What would be wrong with taking this approach, in the light of the advice extended in 'MINA In Real Life'? Would there be a better way of approaching this? Any and all comments appreciated. I'm a MINA newbie! Thanks. Tim
