From edi at agharta.de  Sun Oct  8 21:11:35 2006
From: edi at agharta.de (Edi Weitz)
Date: Sun, 08 Oct 2006 23:11:35 +0200
Subject: [mod-lisp-devel] Chunked encoding
Message-ID: <uvemu4ins.fsf@agharta.de>

Is there a way mod_lisp2 could copy with body content sent by the
client if chunked transfer encoding is used?  (Apache 2 can.)  I
thought about this a bit and my understanding is that this is not
possible due to the way mod_lisp works - you wouldn't know when to
stop reading from the stream.  But maybe I'm just missing something
obvious.

(Rationale: I'm currently porting Hunchentoot to some non-LispWorks
Lisps and I'm going through the code while doing that.  Hunchentoot,
if run as a stand-alone web server, can deal with chunked transfer
encoding in both directions, but I guess I'll simply have to disable
the incoming direction if Hunchentoot is sitting behind mod_lisp.)

Thanks,
Edi.


From marc.battyani at fractalconcept.com  Mon Oct  9 09:20:25 2006
From: marc.battyani at fractalconcept.com (Marc Battyani)
Date: Mon, 9 Oct 2006 11:20:25 +0200
Subject: [mod-lisp-devel] Chunked encoding
References: <uvemu4ins.fsf@agharta.de>
Message-ID: <0f9401c6eb84$27b399d0$1402a8c0@marcx2>

From: "Edi Weitz" <edi at agharta.de> wrote:

> Is there a way mod_lisp2 could copy with body content sent by the
> client if chunked transfer encoding is used?  (Apache 2 can.)  I
> thought about this a bit and my understanding is that this is not
> possible due to the way mod_lisp works - you wouldn't know when to
> stop reading from the stream.  But maybe I'm just missing something
> obvious.

Well you are right, for now mod_lisp reads all the data sent by the browser 
by asking Apache to decode the incoming chunks when chunked encoding is 
used. This could be modified for very big downloads by modifying somewhat 
the mod_lisp protocol. Maybe by sending an header like "chunked-encoding t" 
and then sending all the chunks preceded by their size with a size of 0 to 
signal when all the chunks have been transmitted.

Marc



From edi at agharta.de  Mon Oct 16 12:33:10 2006
From: edi at agharta.de (Edi Weitz)
Date: Mon, 16 Oct 2006 14:33:10 +0200
Subject: [mod-lisp-devel] Chunked encoding
In-Reply-To: <0f9401c6eb84$27b399d0$1402a8c0@marcx2> (Marc Battyani's
	message of "Mon, 9 Oct 2006 11:20:25 +0200")
References: <uvemu4ins.fsf@agharta.de> <0f9401c6eb84$27b399d0$1402a8c0@marcx2>
Message-ID: <uwt70cuex.fsf@agharta.de>

On Mon, 9 Oct 2006 11:20:25 +0200, "Marc Battyani" <marc.battyani at fractalconcept.com> wrote:

> Well you are right, for now mod_lisp reads all the data sent by the
> browser by asking Apache to decode the incoming chunks when chunked
> encoding is used. This could be modified for very big downloads by
> modifying somewhat the mod_lisp protocol. Maybe by sending an header
> like "chunked-encoding t" and then sending all the chunks preceded
> by their size with a size of 0 to signal when all the chunks have
> been transmitted.

Oops, I missed this reply somehow.

Thanks for the info.  I have no immediate need for this right now, but
I'll keep it in mind.

Thanks,
Edi.