🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: comp.sys.ibm.pc.games.strategic
Article #2099

Re: Win95: Been there, done that.

#2099
From: hunt... Charles Randolph Hunter
Date: Sun, 04 Jun 1995 00:00
78 lines
3698 bytes
Markus Becker <bec...@zess.uni-siegen.de> writes:

>an...@ifm.liu.se (Anthony de Vries) wrote:
>>16 bit DOS we all now. It still has the darned 640k limit.
>The DOS does have the limit, Win95 doesn't

>>Only a little part of win95 is 32-bit.  It's a bit like the adding of 
>the 32-bit disk access in old windows. Also only native win95 can pre 
>emptive multitask.

>see below for my comment...

>>But thats no good when there's a "old" DOS program above it. Then it 
>>just won't get any time, because that uses cooperative multitasking.

>Old DOS-Programs use NO KIND OF COOPERATION, and even under Win 3.11
>ther was TRUE pre-emptive multitasking (concerning several running DOS-
>boxes at a time in 386enh-mode)

Agreed, although Windows didn't do this very well, it did pre-emptively
multitask in DOS textmodes, and VGA 320X200 modes [ linear and planer ]
but nothing else, and it was damn slow! 8-)


>>Thats so much as, well if you give some time, than others can use it. As 
>no DOS programmer cares how well other programs will run behind it's own 
>program, normally DOS

>Try, e.g. Descent in one Dos-Box and WC3 in another.
>Do you think those games were designed for MULTITASKING?
>(ok, ok, it won't run as smooth, but...)

>>NB If you're only running native win95 programs, then it will be 
>>preemptive multitasking.

>or if you're running DOS-Programs.
>And even "old" win 3.1 Programs have a smoother multitasking
This is not true. Only smoother if they provide their cooperation, Windows96
might be able to swap tasks faster etc, but this will not yeild much
of a performance gain for 16bit Win3 apps.
>>
>>Compare this to OS/2 which is completely 32-bit, which has HPFS (it doubled the transfer rate of my harddisk!), and will pre-emptive 
>^^^^^^^ i doubt that

Actually, I have found this to be true for many harddrives. Remember
FAT was invented for floppy disks, not harddrives. IBM and Microsoft
 codeveloped HPFS for NT, if I'm not mistaken. They are diskformats 
*intended*  for Harddrives.  Also the IBM drivers provide,internal block 
mode support 32-bit disk access, and streamlined drivers for IDE and SCSI 
devices, instead of generalized BIOS drivers. I tested out HPFS on an old 
MFM 40MB hdd and I doubled my file transfer rate. Operations on the disk 
are faster, but the disk itself stays the same. Some areas like "dir" are 
faster because of internal implmentation of the directory structure.

>>multitask ANY program it runs.
>>So even that old DOS program.

>so does win95 (as well as win 3.1(1))


>>compared to win95. Just don't buy ANY win95 apps, and then company's 
>>will make them for other OS's. 

>Do you think this is the way things work?
>naive, I would say

Not really naive, ever hear of a boycott? 

>>
>>> BTW - just how insuperior is Win95 gonna be to OS/2??  
>>> 
>>VERY 

>What an exact measure!!

Hmm... well maybe a little more accurate  would be  "a few years worth"

>Markus
Technical Details
Message-ID: <huntercr.802319759@lab11.cs.purdue.edu>
Path: mbox-mo1.pugleaf.net!unknown!not-for-mail
References: <3q8ngd$rej@cville-srv.wam.umd.edu> <3qcl9u$7k6@newsy.ifm.liu.se> <3qk1f3$i59@si-nic.hrz.uni-siegen.de>

Show Full Headers
Complete Article Headers
From: hunt...@cs.purdue.edu (Charles Randolph Hunter)
Subject: Re: Win95: Been there, done that.
Date: 1995/06/04
Message-ID: <huntercr.802319759@lab11.cs.purdue.edu>
X-Deja-AN: 103741217
references: <3q8ngd$rej@cville-srv.wam.umd.edu> <3qcl9u$7k6@newsy.ifm.liu.se> <3qk1f3$i59@si-nic.hrz.uni-siegen.de>
organization: Department of Computer Sciences, Purdue University
newsgroups: comp.sys.ibm.pc.games.action,comp.sys.ibm.pc.games.adventure,comp.sys.ibm.pc.games.rpg,comp.sys.ibm.pc.games.strategic,comp.sys.ibm.pc.games.flight-sim,comp.sys.ibm.pc.games.misc,comp.os.os2.games
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit