🚀 go-pugleaf

In article <xzfnml+.msmith@netcom.com> msmith@netcom.com (Martin P. Smith) writes:
>In article <1992Aug29.125049.3797@hellgate.utah.edu> brian%jaguar.cs.utah.edu@cs.utah.edu (Brian Sturgill) writes:
...
[LOTS of text deleted, I hope this was the only part of the message
I'm supposed to respond to.]
...
>>
>>Mainly what it inherits for OS/2 is a portion of its API, though NT's
>>version is considerably cleaner.
>
>Can you provide examples of where the NT API or is this another example
>of BRIAN.EXE ( NT version at work ) where just because it is NT it 
>MUST be cleaner and MUST be better although no facts are provided  to
>support the claim. 

Have you ever even seen the Win32 API?  If you have it should be quite
apparent.  To avoid confusion, I'm not refering to PM vs the corresponding
Windows drawing sorts of functions... PM is somewhat cleaner (though
the Windows API was cleaned up some in Win32s too).

What I'm refering to are system services.  For example the console API
is much cleaner under NT than under OS/2... OS/2 still uses thunked
16-bit calls and does not even bother to define it in its documentation...
I had to go back to an old 1.2 set of manuals to find out what the calls do.

There is a defined API to get process information (i.e. what PSTAT tells
you.)

Many of the API calls now take fewer arguments... OS/2 has a huge number
of instances where you have to specify whether you want struct type 3 or
struct type 4 returned... and the only difference between the two is an
extra field.

All system calls accept UNICODE.  This makes internationalization much easier.

Serial ports are much more convenient to program.

There are many more examples... I suggest you go look for yourself -- it
only costs $69.  Or you can get the Win32 spec for free from ftp.uu.net
in vendor/microsoft... make sure you pick up the Windows 3.0 version
of the Windows 3.1 help viewer program as OS/2 doesn't support Windows 3.1
help files.

>
>NT is not based on the MACH kernel but MS took a lot of ideas ( by 
>their own admission ) from the mach kernel. 

Yes they did get _IDEAS_ from MACH, but the left out MACH's inefficiencies.
I covered this in a recent post (but everybody seems to ignore the
the parts of my posts where I put out information... they just seem
to want to pick on my rantings... I am beginning to believe there's a part
of this group that doesn't want to see information about NT.)
If you can't find my recent post, I can dig it up for you... send me e-mail.

>
>I don't know about other folks but the blatant use of BRIAN.EXE is getting
>very tiresome. 

Oh joyous event!... I've been lifted to the ranks of those with '.EXE's
named in their "honor".  I should ask Pete if it he wants to work out
a bundling deal to increase sales (not that they aren't brisk already :-)).
(Especially the NT versions -- everybody's buying those.)

>
>Martin Smith
>msmith@netcom.com
...

Brian
-- 
C. Brian Sturgill                       *** OS/2 2.0 is for YOU! ***
University of Utah                  Microsoft needs some competition, but
Center for Software Science        -I- want to be one of the many using NT.
brian@cs.utah.edu; CIS: 70363,1373           :-) :-) :-)