Received: by taz.hyperreal.com (8.6.12/8.6.5) id OAA19562; Mon, 1 Jul 1996 14:09:30 -0700 Received: from arachnet.algroup.co.uk by taz.hyperreal.com (8.6.12/8.6.5) with SMTP id OAA19556; Mon, 1 Jul 1996 14:09:23 -0700 Received: from heap.ben.algroup.co.uk by arachnet.algroup.co.uk id aa14037; 1 Jul 96 22:08 BST Received: from gonzo.ben.algroup.co.uk by heap.ben.algroup.co.uk id aa15913; 1 Jul 96 21:29 BST Subject: Re: keepalive and windoze To: new-httpd@hyperreal.com Date: Mon, 1 Jul 1996 21:24:58 +0100 (BST) From: Ben Laurie In-Reply-To: from "Alexei Kosut" at Jul 1, 96 01:54:58 pm X-Mailer: ELM [version 2.4 PL24 PGP2] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2212 Message-ID: <9607012124.aa15174@gonzo.ben.algroup.co.uk> Sender: owner-new-httpd@apache.org Precedence: bulk Reply-To: new-httpd@hyperreal.com Alexei Kosut wrote: > > On Mon, 1 Jul 1996, Chuck Murcko wrote: > > > > On a server running b3-dev I'm being hassled to disable keepalive to > > > keep the windoze users from seeing dumb things... this is of course a > > > poor solution. > > > > > There's a problem. It involves keepalive. The clients under Windoze don't > > seem to catch the connection close of one or more items transferred. > > Other servers don't do this. > > Other servers *don't* do this? I asked about this a bit ago, and no one > responded, so I had assumed it was something all servers did. I assume you > specifically checked these servers and the pages you were accessing to > make sure they were enabling keepalive? And that the bug is > well-reproducable enough with Apache that you can be sure it doesn't show > up on other servers? (Ben tells us letting the Keepalive time out, then > reloading the page will reproduce this) > Other servers _do_ do this. When it came up last time I spent considerable time with a network analyzer verifying that Apache (at least on my box) was doing _exactly_ what the TCP/IP spec says it should. Furthermore, last time round, Netscape fixed it. This time round they say it is fixed in 3.0b5. What more can I say? > > It'd be embarrasing to release with this one still in there. > > Agreed. If it really does not show up with other servers that do > keep-alive, I will -1 any Apache release that has KeepAlive enabled by > default until we fix it (since that means it's fixable, even though it's > a Windows bug). With Windows being the most popular OS in the > browser market, I don't think its appropriate for the most popular server > (35% according to the July Netcraft - 15 point lead over NCSA) to not work > right with it. Agreed. Cheers, Ben. > > Doesn't that make sense? > > -- Alexei Kosut The Apache HTTP Server > http://www.nueva.pvt.k12.ca.us/~akosut/ http://www.apache.org/ > > > -- Ben Laurie Phone: +44 (181) 994 6435 Freelance Consultant and Fax: +44 (181) 994 6472 Technical Director Email: ben@algroup.co.uk A.L. Digital Ltd, URL: http://www.algroup.co.uk London, England.