Path: utzoo!attcan!uunet!lll-winken!ames!pacbell!att!whuts!homxb!hound!smikes
From: smikes@hound.UUCP (S.MIKES)
Newsgroups: comp.windows.x
Subject: Re: informal survey
Summary: I think that it was a good business decision and that they
	 finally recognized that to attract users from the DOS environment
	 they had to give them something they would feel comfortable with
Message-ID: <2858@hound.UUCP>
Date: 12 Jan 89 04:27:07 GMT
References: <8901072026.AA28912@EXPO.LCS.MIT.EDU>
Organization: AT&T Bell Laboratories, Holmdel
Lines: 94


Most DOS users are (shall we say) less than sophisticated, in fact,
some are just downright ignorant. (I said SOME not ALL, after all
I am also a DOS user -- a very sophisticated one.)

Anyway, I have seen (and evaluated) OpenLook and a few other user
interfaces that are attempting to become standards.  While OpenLook
and some of these others have a few nice features that would be useful
in something as simple as Presentation Manager, they also seem to lack
the ease of use that PM seems to provide.

The average user on a UNIX/X Window platform is bound to be far more 
sophisticated one than their DOS counterpart.  It is for this reason
that simplicity is so important.  As it happens, I am a member of a 
group that specializes in User Interface design and development.
Most of my colleagues will agree on this last point.

As for the decision of OSF to adopt the PM standard, it was probably a good
idea from the point of view that in order to effectively compete with Sun
and AT&T's OpenLook in a timely manner, they almost had no other choice but
to use the PM standard.  Presentation Manager, based upon Microsoft Windows,
is fairly well defined and is definitly widely accepted in the "real" 
business community.  (I can cite large numbers of businesses in Manhatten
alone, where there are full scale MS Windows development projects 
under way.)

The decision in the PC world to use PM (based upon MS Windows) also
demonstrates that they have their act together.  The chief concern of
the vendors is to provide a graphical user interface soon enough so
as not to miss the marketing opportunity.  Keep in mind that the PC
industry is only 5 or 6 years old; and really only 3 or 4 years old
in the corporate world.  This is in sharp contrast to the UNIX world,
where they have had almost 20 years to come up with an adequate
front end user interface; and also where there is STILL no clear UI
standard.

I believe that as the PC hardware evolves to provide more powerful
graphics capabilities and reasonable processing power, as do the
bulk of the UNIX workstations currently available, Presentation
Manager will also evolve to provide more sophisticated look, feel,
and functionality [assuming Apple Computer doesn't sue everyone
out of business!]

I don't believe that the X Window based UNIX user interfaces will 
ever be driven by what is going on in the PC world, however, I
dare say that the UNIX application world could stand to "borrow"
a few ideas from some of the more elegant DOS applications.  That
would be much more likely with a powerful windowing system like X.

Lastly, having developed major applications in X, MS Windows and in
Presentation Manager (for OS/2) I have experienced the shortcommings
of each.  X is overwhelming to new developers, so it is likely that
it will be some time before it really takes off.  MS Windows is 
riddled with bugs, is poorly documented, poorly supported, runs only
on PC's and places severe limitations upon developers.  OS/2's most
major shortcomming is that it is also a new entity; serious interest is
only just beginning to emerge.  There are no exhotic toolkits available
for MS Windows or PM other than those provided by Microsoft or IBM, which
are only the basic developers kits.  (SQL Windows and Actor, for MS Windows,
are not true toolkits; they don't generate the source code necessary
to develop Windows applications, they generate object modules that 
are linked into Windows executables.)

Hewlett-Packard is about to release a new product used to develop X
applications that is very likely to become quite popular, codenamed
"builder".  I don't want to steal HP's thunder before their formal
announcement, but this tool will generate the actual C code necessary
to produce the screens, widgets, gadgets or whatever.  It is kind of
like MS Windows' Dialog Editor program, only far more sophisticated.

What all this leads up to is, that if any vendors out there ever
provide a usable X environment for PC's (one that doesn't totally
hog up all the available resources and gives realistically usable
performance), then it is quite likely that we will see Presentation
Manager (and MS Windows if it's still around) ported over to X.

I am aware that somewhere within AT&T (I can't say where) there is 
already work being done on devising a method to provide MS Windows
compatability (through emulation) using X.

The main disadvantage that OpenLook and other attempts at UI stand-
ardization have compared to MS Windows/Presentation Manager is that
they will take time to develop; after initial development, all of 
the available "standards" for X based UI's must then battle it out
to see which one becomes top dog.  In the mean time, PM and Windows
continue to evolve, further entrenching themselves as standards
in the PC world.  For this reason, perhaps it is not such a bad
business decision for the folks at OSF and UNIX International to
have endorsed the PM look/feel.  After all, there are a LOT of
DOS systems out there; and if you wanted to attract some of those
customers to buy and use your products you must provide them with
technology they can understand and are not afraid of.

Steve Mikes