home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!TOROLAB6.VNET.IBM.COM
- From: ADUNSMUI@TOROLAB6.VNET.IBM.COM (Al Dunsmuir)
- Message-ID: <19920828.135123.113@almaden.ibm.com>
- Date: Fri, 28 Aug 92 16:18:21 EDT
- Newsgroups: comp.os.os2.misc
- Subject: Re: OS/2 Win3.1 support *better* than MS!
- Organization: IBM Canada Laboratory Ltd, Toronto, CANADA
- Disclaimer: This posting represents the poster's views, not those of IBM
- News-Software: UReply 3.0
- References: <1992Aug28.044505.2361@news.Hawaii.Edu>
- <pathak-280892123631@virtual.mitre.org>
- Lines: 82
-
- In <pathak-280892123631@virtual.mitre.org> Heeren Pathak writes:
- >In article <1992Aug28.044505.2361@news.Hawaii.Edu>,
- >tholen@galileo.ifa.hawaii.edu (Dave Tholen) wrote:
- >
- >> You know, every once in a while, I'd really like to find out exactly why it
- >> is that one company is able to do something that another company can't. This
- >> is a case in point. If IBM is able to make OS/2 run Windows 3.0 applications
- >> that Windows 3.1 can't run, why or how are they able to do it? Does IBM
- >> know something that Microsoft doesn't? Or does Microsoft know something
- >> that IBM doesn't? Will OS/2 users run into hidden gotchas that Microsoft
- >> already knows about, which might be why they designed Windows 3.1 to be
- >> incompatible with some Windows 3.0 applications? Or has IBM found a clever
- >> solution to a problem that Microsoft couldn't solve?
- >>
- >
- >From what I understand, MS was forced to make Windows 3.1 incompatible with
- >some Windows 3.0 apps to protect Windows from a misbehaved application.
- >Since the Windows kernel is located in the same memory space as a Windows
- >app, the app can cause the entire environment to crash. MS put firewalls
- >around the kernel to trap a "bad" application. This firewall doesn't make
- >Windows crash proof, it just makes it a bit more crash resistant.
- >
- >OS/2 doesn't need this "extra" protection. It all ready gives each process
- >a separate, protected address space which is enforced by the hardware.
- >Thus, IBM can keep the code which lets Windows 3.1 run faster while
- >throwing out the firewall code.
- >
- >> Another case in point is memory boards for PS/2s. IBM says that the system
- >> maximum memory for a Model 70-A21 is 16 Mbytes, yet there is a large number
- >> of third party vendors who will sell Microchannel memory boards that support
- >> more than 16 Mbytes on this model. How do they do it? Do they know something
- >> IBM doesn't know? Or does IBM know something that the third party vendors
- >> don't know, like some sort of hidden gotcha when dealing with more than
- >> 16 Mbytes on such a machine? I've asked a local IBMer why these third party
- >> folks are able to provide more than 16 Mbytes, whereas IBM says 16 Mbytes is
- >> the maximum, but I didn't get an answer. Why is it so hard to get a straight
- >> answer to a simple question?
- >
- >This is probably more a marketing thing. A vendors product line is
- >generally designed so a certain class of users buy a certain type of
- >machine. By limiting the Model 70 to 16M, IBM is trying to encourage
- >people who need more than 16M to buy a different machine.
- >
- >-------------------------------------------------------------------------
- >Heeren Pathak | Millions long for immortality who do
- >pathak@mitre.org | not know what to do with themselves
- >Mitre Corporation | on a rainy Sunday afternoon.
- >(617) 271-7465 | -- Susan Ertz
- >-------------------------------------------------------------------------
- >Disclaimer: Mine not Mitre's.
- >
-
- It's a wee bit simpler than that...
-
- When the marketting announcement for an IBM PS/2 (or PC) is published,
- it contains the specifications for that unit. IBM will only publish specs
- for those features which have been fully tested _at_the_time_of_announce.
- If IBM didn't make a memory card which supported more than X MBytes when
- the Model 70-A21 was announced, the announcement can show the maximum
- configuration which was tested.
-
- As various feature cards (such as memory) are produced, their respective
- announcments will show the machines in which they are supported. Systems
- like the PC/AT "grew" from supporting fairly small amounts of memory to
- supporting a full 16M as time went on. Sometimes system board, BIOS or
- reference disk changes were required to make it all happen. Combinations
- of feature cards and systems not in the announcments _may_ work to various
- degrees: 100%, mostly, or not at all.
-
- As other postings have noted, many of the older Model 70's and 80's use
- a DMA channel for memory refresh, and since the DMA channel was limited
- to 16 MBytes any additional storage must be refreshed by circuitry on the
- memory card. The BIOS (or ABIOS for OS/2) routines probably would also
- have to be updated to support that extra memory too. At some point, the
- upgrade process becomes too comples/costly and you either accept the
- limitations, or buy a new motherboard/machine.
-
- Al Dunsmuir Internet: adunsmui@torolab6.vnet.ibm.com
- CODE Development II + TEAM OS/2 Voice: (416) 448-3315
- IBM Canada Laboratory Ltd, Toronto, CANADA
-
- OS/2 V2: Why settle for anything less?
-