W3C home > Mailing lists > Public > www-lib@w3.org > July to September 1998

Re: leave HTEventList_loop back to application?

From: Henrik Frystyk Nielsen <frystyk@w3.org>
Date: Mon, 03 Aug 1998 18:52:50 -0400
Message-Id: <>
To: Markus Steindl <markus.steindl@akumiitti.fi>, "'www-lib@w3.org'" <www-lib@w3.org>
At 10:09 8/3/98 -0400, Markus Steindl wrote:

>I would like to do a single POST request (into a chunk) and after its
completion continue with the application that initiated the request..
>A POST request apparently doesn't work using a preemptive client, so I
have to use a NoCache-client, define a callback function with
HTNet_addAfter() and start the HTEventList_loop(). (I got this by modifying
the source code of the command line tool, i.e. keeping just the few parts I
>In the callback function I convert the chunk into a string and would be
ready to go on with the application, but I never get back into it. I tried
to stop the loop by calling HTEventList_stopLoop() in the callback
function, but it didn't have any noticable effect.

Because of the differences between HTTP/1.0 and HTTP/1.1 it is really
difficult to make it work using blocking sockets. I basically need the
select call to change state in the HTTP statemachine.

I have written a small test program that does what you need (I think) - it
POSTs a form with some arguments and saves the result in a chunk.

It's called postform and I have added the source in


If you do a CVS update as described in


then you should get it along with an updated Makefile.am. automake will
automatically regenerate the make files.

Run it as

	./postform <uri> "a=1" "b=2" "c=3"

encode space as "+".

Henrik Frystyk Nielsen,
World Wide Web Consortium
Received on Monday, 3 August 1998 18:52:32 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:33:48 UTC