users@websocket-spec.java.net

[jsr356-users] [jsr356-experts] Re: For Review: v002 API and example code

From: Remy Maucherat <rmaucher_at_redhat.com>
Date: Mon, 25 Jun 2012 17:30:38 +0200

On Fri, 2012-06-22 at 18:13 -0700, Scott Ferguson wrote:
> Without streams, it's not a serious proposal.
>
> Chunking/framing is a low-level implementation/protocol detail related
> to implementation buffering; it's not visible at the API-level. The
> fact that the draft introduces an iterator to simulate a stream should
> by itself be an indication that something is wrong.

I have to agree not putting in possible asynchronous stream capability
in favor of that blocking stream "simulation" is not a good idea.
Essentially, it would enforce a message size limitation on input (or
else the application will not scale as well).

-- 
Remy Maucherat <rmaucher_at_redhat.com>
Red Hat Inc