🚀 go-pugleaf

In article <9209091128.AA24447@ucbvax.Berkeley.EDU>, KENKAHN@PKSMRVM.VNET.IBM.COM ("Kenneth A. Kahn") writes:
> Organization: Staff of IBM Fellow - IBM Personal/370
> Disclaimer: This posting represents the poster's views, not those of IBM
> News-Software: UReply 3.0
> X-X-From: KENKAHN@PKSMRVM.VNET.IBM.COM (Ken Kahn)
> References: <9209080106.AA19548@ucbvax.Berkeley.EDU>
>             <715981801snx@sugs.tware.com>
> 
> 
> First off, there is no *wrong* answer; just answers that you might not like.
> By definition (granted IBM's definition) the OS/2 V2.0 platform is a 32-bit
> environment that uses PM as it's interface.  The manuals that support this
> environment describe the functions (e.g. APIs) that you need to program and
> use this environment.  Right or wrong, IBM's position (and I would think that
> of MS and others that support GUI environments) is that all progams written
> to this environment must use the supported GUI for I/O.  The fact that you
> can use 'older' interfaces like VIO and KBD is besides the point.  They are
> not part of the 'official' 32-bit PM platform and thus are not documented as
> such.
> 
> The question you should really be asking, as have a number of IBMers, is why
> the decision was made to not support 32-bit versions of the Kbd, Vio and Mou
> APIs.  I myself have a number of important programs that use these APIs that
> I must convert if I want to stay strictly 32-bit.
>
	All true, never the less IBM seems to have no qualms about using the
Vio and Kbd functions for it's own code.  Is not REXX also part of the
definition of OS/2.  Take a look at the REXX library included as sample code
in the toolkit.  You can't clear the screen, move the cursor around in text
mode or anything else(w/o writing assembly language).  If these API's are being
used for REXX, I think they should be updated to 32-bit code and made
officially part of OS/2.

					John Stewart