RE: [ActiveDir] Very OT: Please Settle a Bet

[joe]

Rick: Excellent logical breakdown you old monkey you.   Roger: I agree with you. Win9x was definitely somewhere in the transition zone so thinking of it as 24 bit or a transitional OS makes sense to me. It went so far as to have a different thunking model for 32<->16 available for use due to how many 16 bit DLLs were still in common use. The win9x only special thunking was called flat thunking and required some special compiling but allowed a 16 bit app to call a 32 bit DLL but more importantly allowed a 32 bit app to call a 16 bit DLL. They also had generic thunkking which is the only thunking available now which is one way, 16 bit app -> 32 bit DLL.   A major programming aspect to look at was that win9x brought out the Windows 32 bit API (win32 api)) as the standard API for windows. Prior to that you had 16 bit and Win32S which if you ever coded for it could be trying and you could find yourself unloading and reloading the actual binary components. You wouldn't ever find yourself only unloading the Win32 API on Win9x. You would find yourself reloading the OS which people did a time or two.   I didn't spend much time on Win9x personally, I jumped to NT4 as soon as I could figure out how to log on and I will admit my PC sat there for a day or maybe two before I figured out how to log on (sometime in 1996 if I recall). Damn thing wouldn't let me bypass the logon screen and I couldn't stop the load process at DOS which really chapped me... I don't recall how I found out about the administrator ID (I certainly didn't read any manuals) but once I did I was like, oh of course, I type in the word administrator and a blank password. Of course, how logical. ugh. I came from the world of sysgens and DEC Mini platforms where you start up in console mode when you booted the system and can do anything and then once you tell it to go multiuser you knew the needed password for the 1,1 or 1,2 accounts. Then the system would stay up and running for months. The only people who could reboot the systems either had a key (starting around the 11/77 or the 11/34a) or knew the right switches to flip on front of the CPU because booting the machine actually involved loading addresses into the proper registers and switching the machine into RUN mode (see pic of 11/70 - http://users.rcn.com/crfriend/museum/TCMtrip/images/1170-34.jpg). The secret of the switches to flip was a trade secret handed down from sysadmin to sysadmin and you were required to memorize it versus writing it down, or at least it was where I came up through the ranks at.   Another major programming aspect was around memory management. Obviously you had more memory available to you by jumping to 32 bit pointers but there was also a jump from shared memory for all of the apps to each app having its own virtual space. This broke quite a few apps trying to go to 32 bit because they were all used to be able to talk directly to each other versus having to marshall data between the processes. Basically it wasn't a simple recompile for many apps that communicated to work on 32 versus 16 which is YAR for making the 16/32 border a little nebulous. Companies don't like to have to redesign applications, heck many companies don't like to design applications... They throw some code through a compiler and see who will pay.    Win3.0/1/1.1 could all run on the 386 but one of the big complaints about it was that it was a 16 bit OS riding a 32 bit machine. I recall when win95 came out and how MS really pushed the point of it being full 32 bit to take advantage of the power of the newest PCs and corresponding complaint from press that a majority of the stuff available was only 16 bit so you really didn't get the full benefit. I wonder how much better this will be handled in the 32->64 switchover. The big problem we have this time is competing architectures which should cause it to take longer to all shake out. As a developer I intend to stick with 32 bit for some time and rely on good thunking capability in the OS.     joe From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Roger Seielstad Sent: Saturday, February 12, 2005 11:41 PM To: ActiveDir@mail.activedir.org; 'Send - AD mailing list' Subject: RE: [ActiveDir] Very OT: Please Settle a Bet I've alway described Win95 as a 24 bit operating system myself...   Actually, the OS (i.e. the kernel) is (was) definitely 32-bit code. Rick backed into the correct answer with that damn logic thing again.   However. explorer.exe (i.e. the GUI) was most definitely a 16-bit app, because at the time they hadn't figured out all the 32 bit optimizations for graphics - they had done all the 3.x work in 16 bit. IMO - this is one of the reasons 9x has always been relatively unstable - the mixture of 16 and 32 bit code.   Roger -------- Roger Seielstad E-mail Geek & MS-MVP   From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Rick Kingslan Sent: Saturday, February 12, 2005 12:18 PM To: ActiveDir@mail.activedir.org; 'Send - AD mailing list' Subject: RE: [ActiveDir] Very OT: Please Settle a Bet Charles,   I follow your line of thinking and would tend to agree except for my first foray into Networked OS’s – Netware.  Netware is CLEARLY an OS – is CLEARLY 32-bit, but requires DOS to boot the kernel, which then continues to load the required pieces of Netware on the Netware kernel.    So, in that – Netware is not a frontend for DOS – it simply uses the load routines of DOS to get going, then switches the processor to privileged mode to operate with all of the features of the processor in 32-bit mode.   The question that should be asked is this, which should solve the current puzzle and bet:   Can Windows 95 be run on a 80286 processor?  If not – and must be run on a 80386 and greater – it’s 32-bit and using privileged mode and the features that it affords.   The answer to the above question is no – it must be run on a 386 or greater processor because it requires 32-bit addressing.  It emulates 16-bit for those legacy apps the needed it.  DOS was used, as in Netware, as a launching platform for the ‘kernel’ (though not in anyway as complex).  The downside to Win95 was the obvious leverage on some DOS functions, and complete lack of any security and a very lackluster separation of program to program corruption.   If you want more info – see here.  http://www.webdevelopersjournal.com/archive/win95.html   I remember Greg from the ‘Chicago’ (code name for Win95) beta days, and thought he wrote an article or two.   Hope this helps.   Rick Kingslan  MCSE, MCSA, MCT, CISSP Microsoft MVP: Windows Server / Directory Services Windows Server / Rights Management Windows Security (Affiliate) Associate Expert Expert Zone - www.microsoft.com/windowsxp/expertzone WebLog - www.msmvps.com/willhack4food     From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Carerros, Charles Sent: Friday, February 11, 2005 4:18 PM To: 'ActiveDir@mail.activedir.org'; Send - AD mailing list Subject: RE: [ActiveDir] Very OT: Please Settle a Bet   My vote is that Win 95 required DOS and therefore was a frontend DOS application and not a true OS.  A good example, watch a Win 95 box boot, it always starts out with DOS and then DOS runs the interface, WIN 95.   Gnome isn't and OS its simply a shell, DOS is the same thing. -----Original Message----- From: Dean Wells [mailto:[EMAIL PROTECTED] Sent: Friday, February 11, 2005 4:01 PM To: Send - AD mailing list Subject: RE: [ActiveDir] Very OT: Please Settle a Bet 32 bit cooperatively multitasked if memory serves ...but it might not ;) -- Dean Wells MSEtechnology * Email: [EMAIL PROTECTED] http://msetechnology.com     From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dan DeStefano Sent: Friday, February 11, 2005 4:54 PM To: ActiveDir@mail.activedir.org Subject: [ActiveDir] Very OT: Please Settle a Bet Could anyone settle a bet for me? I would like to know if Windows 95 was a 16 or 32-bit OS. One of us is saying that it was natively 32-bit, but ran 16-bit apps in a VM, while the other one is saying the reverse: it was a 16-bit OS that was capable of running 32-bit apps in a VM.   Also, one person is saying that W95 required DOS (like Win3.1.1) and the other is saying that, while built on DOS, DOS was not required and the OS went above and beyond its DOS roots.   If anyone can settle these issues and offer proof like links to Web pages and such, we would be grateful.   _________________________   Daniel DeStefano PC Support Specialist   IAG Research 345 Park Avenue South, 12th Floor New York, NY 10010 T. 212.871.5262 F. 212.871.5300   www.iagr.net Measuring Ad Effectiveness on Television   The information contained in this communication is confidential, may be privileged and is intended for the exclusive use of the above named addressee(s). If you are not the intended recipient(s), you are expressly prohibited from copying, distributing, disseminating, or in any other way using any of the information contained within this communication. If you have received this communication in error, please contact the sender by telephone 212.871.5262 or by response via e-mail.