W3C home > Mailing lists > Public > www-amaya@w3.org > July to September 1997

Re: Problems using PUT

From: astro <astro@hades.eden.com>
Date: Mon, 14 Jul 1997 12:56:53 -0500 (CDT)
To: Jim Gettys <jg@pa.dec.com>
cc: www-amaya@w3.org
Message-ID: <Pine.LNX.3.93.970714124401.22786A-100000@hades.eden.com>
Thanks for your response:

As I indicated in my original post, I am indeed using Apache 1.2 (released
version) and it appears to deal correctly with HTTP/1.1 requests. Further,
the 'PUT' operation functions correctly using Netscape communicator

I do believe this is probably the timing problem that I've seen talked
about on this list between amaya and an Apache server running on the same
machine, so again, I'm asking if anyone can verify this, or has advice on
how to remedy the situation?

Since my first writing, I obtained and installed a different PUT script
(written in C, I believe it came straight from CERN and was linked to from
the w3 site. I realize that this should still not be considered secure,
but I'm just trying to implement this on a test basis for now.

At any rate, the C put program exhibits the exact same behavior as the
PERL script I was using - works remotely with communicator but I get
corrupted/aborted saves locally using amaya.

As always, any further information appreciated,


| Bryan White / astro@eden.com
| http://hades.eden.com
On Mon, 14 Jul 1997, Jim Gettys wrote:

> Put is fundamentally unreliable using HTTP/1.0 due to a race condition
> between the close of the connection and the reply message.  This depends
> on exact timing in the network and of the application.
> HTTP/1.1 fixes this.  What version of what server are you using to
> use PUT?  Note that current versions of both Jigsaw and Apache speak HTTP/1.1.
> 			- Jim Gettys
Received on Monday, 14 July 1997 19:53:50 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:53:18 UTC