How original is the Mac?

[RobertBe]

Windows 3.1 and DOS weren't the best, but at least that combination could truly multitask.

I'll have to differ on that one. True multi-tasking (like that on the Amiga) for Microsoft didn't come until Windows 95, NT, etc.. Cooperative multi-tasking was in Windows 3.1 (and it didn't multi-task that well, in part because it was not pre-emptive multi-tasking like that on the Amiga).

Truly,
Robert Bernardo
Fresno Commodore User Group
http://videocam.net.au/fcug
The Other Group of Amigoids
http://www.calweb.com/~rabel1/
Southern California Commodore & Amiga Network
http://www.sccaners.org

[PaulQ]

I'll have to differ on that one. True multi-tasking (like that on the Amiga) for Microsoft didn't come until Windows 95, NT, etc.. Cooperative multi-tasking was in Windows 3.1 (and it didn't multi-task that well, in part because it was not pre-emptive multi-tasking like that on the Amiga).

You can split hairs on the differences between pre-emptive and co-operative multitasking, but the fact remains that the Macintosh computer and its respective OS of this time had no multitasking whatsoever. As inferior as co-operative multitasking was to pre-emptive multitasking, it still beat out having no multitasking available. Also, people often neglect the fact that OS/2 had pre-emptive multitasking back in '92.

[Boray]

Also, people often neglect the fact that OS/2 had pre-emptive multitasking back in '92.

AmigaDos had preemptive multitasking in 1985.

By the way, Windows Vista/Windows 7 was based on Windows NT which was based on OS/2.

[sjgray]

Also, people often neglect the fact that OS/2 had pre-emptive multitasking back in '92.

AmigaDos had preemptive multitasking in 1985.

By the way, Windows Vista/Windows 7 was based on Windows NT which was based on OS/2.

According to wikipedia:
http://en.wikipedia.org/wiki/Preemption_(computing)
Sinclair QDOS had it in 1984...

The wiki page is interesting in that it appears a reference to the "Commodore 64" was added, seemingly at random, as an example of a non-multitasking "operating system"... LOL

And, no, I'm not a Sinclair user.

Steve

[Jeff-20]

pfft... under the right conditions, the VIC could preempt too.

[Bacon]

the Macs still crashed (but with a cute picture of a bomb whenever they did).

I don't believe this is correct. Didn't the Mac crash with a picture of the Mac with an unhappy face? Because the Atari ST line crashed with bombs across the screen, later revisions crashed with mushroom clouds.

There was a bomb. When I worked for newspapers in the late 90s I used Macs almost exclusively and they crashed more or less daily so I got to see that bomb frequently.

[PaulQ]

It wasn't even so much that the Mac wasn't innovative, it was actually a backwards step. For example, we already had mice to control cursors and pointers, but then Apple made it worse by reducing it to one button and an awkward shape. We already had a GUI and full colour graphics for our computers, but then Apple made it worse by eliminating all colour and even the shades of gray, forcing everything onto a tiny black and white screen. I'm sorry; black and white in 1984? In the era of MTV? Not in my house! Using the Mac, I found it slow and unresponsive, and many of the "Killer Apps" had to have been stripped down to run on a Mac. The keyboard that came with the Mac made me appreciate the keyboard on the Vic 20 and Commodore 64.

It was a classic example of form over function. I'm amazed they were able to sell enough of these to keep their company going. That said, it's no wonder Mac never took over the market. You can only fool some of the people some of the time.

[Richard James]

pfft... under the right conditions, the VIC could preempt too.

I remember writing a toy application that did something like this. It was written in BASIC. It had two windows (Ok they were boxes) on the screen and in the right one it was printing primes. In the left box I think it was some sort of editor.

Except I didn't do the timing right and the prime number program took up all the CPU time.

[dragos]

Nowadays when Macintosh runs on Intel X86, they really can't claim to use a more powerful CPU than the competitors' PCs, since a Dual-Core is a Dual-Core no matter which operating system you load onto it.

Well, not entirely accurate. If you can point out a similarly price dual Nehalem desktop windows machine, I will concede,. However, having just built a core I7 machine for windows 7 after searching high and low for a realistic dual nehalem board, I Can tell you you most likely wont. (servers do not count as they are SERVERS)

Therefore, if they so chose, they could (and truthfully so) claim they use more powerful processors.

I am aware they use core2quads in the lower end macs, so don't respond with the spec to a mac mini. I am speaking, of course, on the mac pro.

[carlsson]

Perhaps HP, Dell and all the others consider that CPU you mention too expensive for their customers, while Apple users have no problem paying premium money? After all that is what they've been used to do for 20 years or so.

But you'll agree if that dual Nehalem PC shows up, hardware wise the CPU inside the Mac no longer is more advanced than the identical one inside the PC?

[dragos]

I would concede your CPU point. The omission of BIOS and use of EFI is still a valid point. There are others...

[Pedro Lambrini]

So, could someone help an old ignoramous (my experiences witht he Mac ended with the G4 stuff)? If the Mac now uses Intel architecture is the only real difference between a Mac and a Windoze PC the OS or is there more to it?

[carlsson]

Dragos makes a mention of the extensible firmware interface (EFI) which is a layer between hardware and operating system. It replaces the good old BIOS, and EFI implementations can be much larger and more complex.

EFI removes the old 1980's PC limitations modern BIOSes tend to build upon. I suppose it means it can make more efficient use of the hardware, but according to article there is some criticism about that. It seems today mostly Itanium systems, Macs and rack/blade servers have EFI. Starting with Windows Server 2008, you can boot a PC without BIOS.

So it seems a X86 Macintosh can either boot a BIOS operating system like old school Windows XP/Vista or an EFI compatible system like OSX or Linux are. On a dual booting machine, in theory you should see quite some difference in performance between running a hardware intensive task in Windows or OSX if this EFI thing makes a big difference.

[rhurst]

Thanks, carlsson, that's a good summary. BIOS, EFI, et al primarily function as the machine initialization to bootstrap code -- the variances and enhancements each offer all depend on the motherboard's targeted solution. Beyond that, BIOS/EFI's OS integration (if any) is determined by which OS is loaded. DOS and earlier Windows makes more use of BIOS, and EFI for Itanium2 booting HP/UX or for Xeon booting OSX has tighter integration, whereas Linux and newer Windows essentially take over the machine after bootstrap -- which has the benefits of being more portable (hermit crabs) across many mobos, but it adds even more startup costs over the other integrated solutions.

My experiences with EFI over BIOS is that it is slower to get to the boot process, and it is definitely larger and more complex to operate. BIOS and EFI-based startups on commodity hardware can take up to several minutes, and you have to strip down a lot of default settings to make their execution tolerable. EFI size and complexity offers the ability to launch a richer set of utilities and diagnostics using either a console or GUI -- which is useful because it is integrated with the machine and you don't have to fuss over slow booting and handling of external "live" media for the same. That said, I am not convinced EFI mobos assist the commodity hardware / OS mix (Windows, Linux) targeted users in any significant way; I can see its fit within enterprise servers and designer computers (Apple, Sony) which can make it work well and leverage its presence, because they are the OEMs of both hardware and OS -- which they can then sell that integration for a premium. It's not evil, it's business. The alternative is enterprises can staff in-house expertise to forego those costs by the manufacturer's VAR and stick with "open" solutions and manage themselves.

The coreboot project is another animal (rabbit) to replace the good ol' BIOS, too, and the benefit there is to simplify the machine startup process, load a Linux (or Windows) kernel into initialized memory and I/O controllers that are required for fetching (cdrom, disk, network) the payload. The significance of this new startup process is that it takes only seconds from machine power-on to an initialized Linux console -- an experience not unlike turning on any ROM-based Commodore machine, with the extended option of booting into a workbench (GEOS, Amiga). And then, you might have the option for the workbench to restore very quickly to your last session from its hibernate functions.

[aneurysm]

My favorite thing about the Trashbin is that it is on the desktop. Who keeps their garbage can on their desk?!?!