Xref:	news2.mv.net comp.os.msdos.djgpp:5686
From:	s0052mol AT sun10 DOT vsz DOT bme DOT hu (Molnar Laszlo)
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: Bug in stubify.c
Date:	5 Jul 1996 12:20:26 GMT
Organization:	Technical University of Budapest
Lines:	15
Distribution:	world
Message-ID:	<4rj1aa$93s@goliat.eik.bme.hu>
References:	<1 DOT 5 DOT 4 DOT 16 DOT 19960703185459 DOT 266fdb4e AT mailhost DOT cyberhighway DOT net> <4rfmnh$dr4 AT tempo DOT univ-lyon1 DOT fr>
NNTP-Posting-Host:	sun10.vsz.bme.hu
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

Dominique Micollet (dmicolet AT u-bourgogne DOT fr) wrote:
: Today, I think that using a ramdisk which DJGPP is not a good idea because
: the consumed memory is lost for DJGPP : the program will swap more frequently.
: Am I right on this last point ?
: 
Not really. I use a program that is a combined disk cache AND ramdisk. It 
shares the memory dinamically between the cache and the ramdisk. If I 
allocate 5 Megs of XMS for this program I have a virtually 5 Megs 
large ramdisk and a virtually 5 Megs large cache. The rest from my 16 Megs 
can be used by DJGPP.

Bye, ML

Ps: This program is called COMBI. Search for combi118.zip (or similar) if 
interested.  

- Raw text -