From: leathm AT solwarra DOT gbrmpa DOT gov DOT au (Leath Muller)
Message-Id: <199709090339.NAA19032@solwarra.gbrmpa.gov.au>
Subject: Re: fread slowstart
To: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Date: Tue, 9 Sep 1997 13:39:43 +1000 (EST)
Cc: dj AT delorie DOT com, djgpp-workers AT delorie DOT com
In-Reply-To: <9709090320.AA16818@clio.rice.edu> from "Charles Sandmann" at Sep 8, 97 10:20:16 pm
Content-Type: text
Precedence: bulk

> > Currently, the amount starts at 512 and doubles after each read (i.e.
> > 512, 1k, 2k, 4k, 8k, etc).  Should it quadruple instead?
> > (512, 2k, 8k, 32k)  That would get to the larger transfers faster, but
> > be less optimal for some programs that fseek a lot.
 
> I would probably quadruple, or even 8X (512, 4K, 32K).

I normally don't comment too much - but I would vote for even a larger starting
size and then quadruple... ie: start at say 4k, then 16, then 32 etc. Why
not do this?
 
> > Also, I've changed the default transfer buffer size to 32k and stack
> > size to 512k in stub.asm.  
> > Should we default to a max size (63.5K) transfer buffer now ?
 
> I think it's excessive.  It eats into the DOS memory left for other 
> programs, and I didn't see much improvement in data transfer speeds for
> transfers over 32K on any system I tested.
 
I would vote for a default of 32k...

Leathal.