bob@imspw6.UUCP (Bob Burch) (09/25/88)
From Ted Holden at HTE: From: Doug McDonald, uxe.cso.uiuc.edu!mcdonald >This is somewhat off the topic, but I saw an interesting thing >at an IBM "product fair" at our university on Monday: IBM'ers >wearing T-shirts and blue jeans. (Most T shirts were orange and blue... >wonder why.) Anyway, one guy was wearing a suit (but no tie). He >informed me that OS/2 fully supported the native 32-bit protected >mode of the 80386, and indeed that OS/2 was really INTENDED to....... The charitable assumption is that the gentleman simply didn't know what he was talking about. Recent articles in PC Week and elsewhere indicate that any version at all of a 386 BS/2 is at least a year away. Basically, IBM FEARS the 386, knowing full well it can kill their 36's, 38's etc. They thought to convince the world the 386 was for file servers and exotic graphic workstations (partly by charging $10,000 for their 386 PS/2). Recently, they have looked around and seen the whole rest of the world selling 386 machines for $1700 - $2500 and panicked, making a high priority of a 386 version of BS/2. Presently, BS/2 running on a PS/2 (or any 386 machine) runs it in 286 emulation mode, which isn't really running it; that's walking it. Microsoft has just spent the last 6 - 8 years perfecting a relatively simple monitor program (MS/DOS). A real operating system such as UNIX or DEC VMS takes eight or ten years or more to get healthy, robust, and fast for a given architecture. UNIX, which is the ONLY portable OS worth talking about, represents the work of true geniuses at several hundred sites over about a fifteen year period; as a basis for comparison, a comparable intellectual endeavor might be the entire literature of a semi-major nation over a 200 - 500 year period. The chances of Microsoft bringing BS/2 anywhere close to challenging UNIX within the next ten years are slim and none, and Slim's riding out of town on his horse even as you're reading this article. Ted Holden HTE
mcdonald@uxe.cso.uiuc.edu (09/27/88)
>>Anyway, one guy was wearing a suit (but no tie). He >>informed me that OS/2 fully supported the native 32-bit protected >>mode of the 80386, and indeed that OS/2 was really INTENDED to....... > The charitable assumption is that the gentleman simply didn't know >what he was talking about. That appears to have been the case. He sent me a glossy (glossy MAROON!) blurb that indeed said OS/2 did not use 32 bit mode. > Microsoft has just spent the last 6 - 8 years perfecting a >relatively simple monitor program (MS/DOS). A real operating system >such as UNIX or DEC VMS takes eight or ten years or more to get >healthy, robust, and fast for a given architecture. UNIX, which is the >ONLY portable OS worth talking about, represents the work of true >geniuses at several hundred sites over about a fifteen year period; as >a basis for comparison, a comparable intellectual endeavor might be the >entire literature of a semi-major nation over a 200 - 500 year period. >The chances of Microsoft bringing BS/2 anywhere close to challenging >UNIX within the next ten years are slim and none, and Slim's riding out >of town on his horse even as you're reading this article. >Ted Holden Genius, well I wouldn't know. But Unix and OS/2 are not similar things. Unix is a MULTIUSER operating system. OS/2 is single user. Unix is missing several things that OS/2 has: in particular, device monitors, installable device drivers, and the ability to do IO (but unfortunately not to capture interrupts) in user mode. The main thing Unix has is portability. But in the IBMPC world, nobody cares a hoot about portability. Of course, all those things could be added to Unix. To me the ability to do actual IO in user mode is absolutely vital. Doug McDonald
james@bigtex.uucp (James Van Artsdalen) (09/28/88)
In article <45900156@uxe.cso.uiuc.edu>, mcdonald@uxe.cso.uiuc.edu wrote: > [...] But in the IBMPC world, nobody cares a hoot about portability. This is somewhat out of context, but perhaps should go up for being the Quote-of-the-Month! I did game programming on the PC for several years and now I do BIOS programming, and the >only< major issue is compatibility. > To me the ability to do actual IO in user mode is absolutely vital. Graphics on the screen perhaps, but I don't see user access to such things as floppy/hard drives or the CPU cache controller as being a good thing. We *will* be abandoning the WD1010 interface and all of the other hardware things at some point and it would be nice if the software still worked. -- James R. Van Artsdalen ...!uunet!utastro!bigtex!james "Live Free or Die" Home: 512-346-2444 Work: 338-8789 10926 Jollyville Rd #901 Austin TX 78759
beckman@iuvax.cs.indiana.edu (Peter Beckman) (09/28/88)
In article <8605@bigtex.uucp> james@bigtex.UUCP (James Van Artsdalen) writes: >In article <45900156@uxe.cso.uiuc.edu>, mcdonald@uxe.cso.uiuc.edu wrote: >> [...] But in the IBMPC world, nobody cares a hoot about portability. > >This is somewhat out of context, but perhaps should go up for being the >Quote-of-the-Month! I did game programming on the PC for several years >and now I do BIOS programming, and the >only< major issue is compatibility. Compatibility and Portability are two different notions. Maybe you read his sentence incorrectly.