Date: Mon, 10 Nov 1997 10:12:39 +0200 (IST)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: Charles Sandmann <sandmann AT clio DOT rice DOT edu>
cc: Demmer AT lstm DOT ruhr-uni-bochum DOT de, djgpp-workers AT delorie DOT com
Subject: Re: malloc()
In-Reply-To: <9711092240.AA14335@clio.rice.edu>
Message-ID: <Pine.SUN.3.91.971110101219.13667E-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Precedence: bulk


On Sun, 9 Nov 1997, Charles Sandmann wrote:

> I would tend to think a marginally slower package which pages less and
> fails later would be a better default choice.  But I am very conservative
> about changing working things also...

That is exactly what I was thinking.  And that is why IMHO we need to
see the results of testing with real-life programs that matter.  How
many times did you see a benchmark that shows a huge difference
between two pieces of software, only to discover later that the effect
on real programs is marginal or nonexistent?  I see it a lot.