Hi,

What mail-trail are you referring to?

No, you won't have to bother about selector.select() calls when using MINA. In fact, there's no way you could get hold of the selector(s) used. Using MINA you focus on the implementation of your protocol using filters and codecs. MINA will take care of the lower level NIO stuff.

/Niklas

Nirmalya Sengupta wrote:
Hello all,

Perhaps this is too frivolous a question, but I will appreciate a
clarification nevertheless.

While going thru the mail-trail on the subject [], I got a feeling
that while building
an application using MINA, one needs to indeed take care of Selectable.select()
call prgrammatically. I am restructuring an application (typical many
client-one server
scenario) using MINA, and the existing version of the application made
use of its
own thread library and took care to call select() at the right points
in time. When I
came to MINA, I was under the impression, that since MINA is completely
API-based and the infrastructure provides demultiplexing using
select() under the
hood, I will not have to deal with it at all. I simply have to create
CoDec factories,
provide CoDecs themselves and provide code to the touchpoints of the
IoHandlerAdapter implementation.

But the mail-trail mentioned above leads me to think that perhaps I was wrong. I
still have to call select() at appropriate time, and hence I still
have to take care
of (re)setting select keys.

Could you please clarify? Many thanks in advance.

Nirmalya

--
Nirmalya Sengupta
Software Technologist
"If you have built castles in the air, your work need not be lost.
That is where they should be. Now put the foundation under them."
--Henry David Thoreau

Reply via email to