Date: Wed, 13 Oct 1999 11:43:13 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Michael Abbott aka frEk <20014670 AT snetch DOT cpg DOT com DOT au>
cc: DJGPP Mailing List <djgpp AT delorie DOT com>
Subject: Re: allegro gone missing?
In-Reply-To: <3802B556.4F1BECF0@snetch.cpg.com.au>
Message-ID: <Pine.SUN.3.91.991013114236.16702G@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Tue, 12 Oct 1999, Michael Abbott aka frEk wrote:

> > > I don't know about everyone else but for speed I usually use Win '95s
> > > DPMI...
> >
> > Are you saying that Windows' DPMI server is faster than CWSDPMI?  Or are
> > you saying that it is smaller.  I find it hard to believe how could
> > someone make such claims.
> 
> I know this is the wrong list for this (biased opinion :)

Sorry to disappoint you: I don't think our opinions are biased in this
case.  I have hard data to support what I wrote.

> but IN MY EXPERIENCE Windows '95 DPMI has been faster.

I would guess that your system setup with CWSDPMI is far from
optimal.  Perhaps you don't have a disk cache installed (Windows 9X
installs a cache by default, but DOS doesn't).  See section 3.9 of the
DJGPP FAQ list, for advice about optimal system setup.

My measurements indicate that a given job runs about 40% slower on
Windows than on DOS, on the same machine.  However, I don't claim that
Windows DPMI host is slower by 40%, because I suspect most of the
slow-down has nothing to do with the DPMI services, but rather with
the other system processes that run in parallel.