From: Jim Jagielski <jim@jaguNET.com>
Message-Id: <199610142219.SAA06817@shado.jaguNET.com>
Subject: Re: WWW Form Bug Report: "Hard-coded 150 connection limit not
 documented " on FreeBSD (fwd)
To: new-httpd@hyperreal.com
Date: Mon, 14 Oct 1996 18:19:26 -0400 (EDT)
In-Reply-To: <Pine.SGI.3.93.961014140819.21267k-100000@fully.organic.com> from
 "Brian Behlendorf" at Oct 14, 96 02:10:23 pm
Content-Type: text
Sender: owner-new-httpd@apache.org
Precedence: bulk
Reply-To: new-httpd@hyperreal.com

Brian Behlendorf wrote:
> 
> On Mon, 14 Oct 1996, Jim Jagielski wrote:
> > Ben Laurie wrote:
> > > 
> > > Rob Hartill wrote:
> > > > 
> > > > 
> > > > This has become a frequent complaint/request.
> > > 
> > > I haven't quite understood why this limit is hard? Is there any reason not to
> > > make it soft?
> > 
> > I haven't looked too closely at it... I'm guessing that it's only on
> > the file-version of the scoreboard that it would be a concern doing
> > a "on-the-fly" malloc of the space, and even then it's a simple matter
> > of when it's done...
> 
> The only reason I can think of why this would be a "feature" is if the
> individual responsible for admining the machine (let's call them "root") and
> building the software was different from the individual maintaining the config
> files (let's call them "www-user"), and "root" didn't want to allow "www-user"
> to be able to configure anything which "root" knew would bring the machine to
> its knees.
> 

>From what I can tell, it would require a bit of work to allow on-the-fly
malloc of scoreboard space, since a lot is set and assumed in scoreboard.h

I'm 0 on changing it... I think most people just wanted to know there
_was_ a hard, internal limit.

-- 
Jim Jagielski  << jim@jaguNET.com >>   |   "If we took the bones out
  **  jaguNET Access Services  **      |    it wouldn't be crunchy"
      Email: info@jaguNET.com          |            Whizzo Crunchy Frog
++    http://www.jaguNET.com/         +++      Voice/Fax: 410-931-3157       ++