From: "John Beamish" 

I liked OS2 and used it for some time.

But ... Egan's law holds true.  The better OS software did drive out the
poorer software.  The actual code is OS2 was probably *way* better but the
OS is not a standalone product.

1.  It's the install.  Win3.1 and Win95 were better simply because they
worked.  (Admittedly they worked part of the time because the install was
already done when you bought the machine but even at their worst, the Win
installs were better than OS/2

2.  It's the support.  Win3.1 and Win95 supported way more hardware and
software.  If I want to hook up some gizmo or install some software, I do
not want the OS telling me it can't do it.

IOW ... you don't buy an OS, you buy access to other hardware and software.
Win3.1 and Win95 did that much better than OS/2.


"Robert Comer"  wrote in message
news:3e617352{at}w3.nls.net...
> Hi Egan, nice to see you around here again!
>
> I'm very much aware of OS/2's shortcomings, as well as its strengths, but
> the point was that it's plainly better software wise than Win3.1 and
Win95,
> yet it didn't displace the bad.
>
> > Not really. IBM mishandled OS/2 horribly all throughout its life.
>
> No doubt about that.
>
> >First
> > they associated it with the PS/2 (limit PCs to 286 chips, yeah right!)
> > with MicroChannel. Thus they had their clock promptly cleaned by Compaq,
> > and updated OS/2 2.0 for the 386, but by then the association had stuck.
>
> Microchannel was yet another better computer product that didn't win (plug
> and play that really worked!) -- IBM's problem with it was licensing and
> pricing.
>
> > By OS/2 3.0 that system was fairly capable, but the installation was a
> > real pain (earlier this year I finally threw out that box of floppies),
> > plus networking wasn't well integrated, thinking of it makes me shiver.
>
> I actually got into it at the 2.1 level with Microchannel hardware -- it
was
> nice to say the least -- a bit hoggish, but a properly outfitted machine
was
> rock solid stable, something I haven't reached yet with Windows and the
> hardware we have today.
>
> 3.0 the first release wasn't bad, and like you say, the install and the
> networking was a bit odd -- that was pretty much "fixed" in the 3.0
> "Connect" version.  I ran a Lotus Notes server for years on
that without a
> single crash. I ran it on a '486 DX4/100 with 32M of RAM, by far the
> smallest hardware requirements of any Notes server in our corporation (all
> others were NT based then), yet we had the record for least problems. I
had
> got another job so wasn't doing OS/2 desktop anywhere else but at home by
> that time.
>
> > OS/2 4.0 was the first competent version, IMHO.
>
> It had some install and hardware problems, but I used it at home for a
long
> time, and what I move the Notes server to when the hardware died for the
> above. (We moved the Notes server to our iSeries after that and still have
> the most solid server.)
>
> >But IBM didn't know how
> > to market it, the IBM PC company was actively hostile to it,
>
> Yep.  That was the main problem with it, not the software. 
>
> >there were
> > still some issues with Presentation Manager and its single input queue
> > problems (never fixed).
>
> I never had a problem with PM, and as for the SIQ, I had an add-on program
> that would break that kind of deadlock.  I forget the name of it now
though.
> :(
>
> >So OS/2 was finally and irretrievably doomed by
> > Microsoft's backstabbing, proprietary DLL software, an upgrade treadmill
> > that IBM wouldn't invest enough to match, and that lock on OEM preloads
> > (determined to be illegal, but too late to save OS/2). With OS/2 one can
> > say that everything that could go wrong, did go wrong, all of its life.
>
> I can't argue about anything there!  It's really to bad as I still liked
the
> way things worked -- it fit my style pretty well.
>
> > So it's arguable that OS/2, although a better fundamental design, wasn't
> > well enough executed, delivered or supported to make it better software,
> > and it was badly handicapped by its vendor plus Microsoft's shenanigans.
>
> Have you forgot the Win3.1 resource problems, the Win95 Win16Mutex lockups
> and plug and play that never plugged and played well, not to mention the
> memory leaks and just general instability of Win95 for years that never
was
> solved. Even after they revved it several times up through WinME.
>
> > I don't regard OS/2 as a valid counter-example for all of these reasons.
> > One can imagine that IBM's OS/2 debacle is one reason for it deciding to
> > support Linux at arms-length -- without trying to build some IBM distro.
> > The other reason is probably its overhead and limited success with AIX.
>
> Actually I haven't figured that one out yet, other than maybe they just
> really don't want to be responsible for the OS part so much. (though they
> really do make good OS's for the bigger hardware -- OS400 and the
mainframe
> OS's)  I'm not much of a fan of AIX, just another UNIX with hardware that
> cost to much for what it did.
>
> > PS -- I still have an OS/2 partition, but I hardly ever boot it anymore.
>
> One of my PC's here at home dual boots OS/2 V4 and Lindows. (most of the
> time in Lindows) And I run WinME on my main PC when I play games. 
>
> - Bob Comer
>
>
>
> "Egan Orion"  wrote in message
> news:3e616539$1{at}w3.nls.net...
> > Not really. IBM mishandled OS/2 horribly all throughout its life. First
> > they associated it with the PS/2 (limit PCs to 286 chips, yeah right!)
> > with MicroChannel. Thus they had their clock promptly cleaned by Compaq,
> > and updated OS/2 2.0 for the 386, but by then the association had stuck.
> >
> > By OS/2 3.0 that system was fairly capable, but the installation was a
> > real pain (earlier this year I finally threw out that box of floppies),
> > plus networking wasn't well integrated, thinking of it makes me shiver.
> >
> > OS/2 4.0 was the first competent version, IMHO. But IBM didn't know how
> > to market it, the IBM PC company was actively hostile to it, there were
> > still some issues with Presentation Manager and its single input queue
> > problems (never fixed). So OS/2 was finally and irretrievably doomed by
> > Microsoft's backstabbing, proprietary DLL software, an upgrade treadmill
> > that IBM wouldn't invest enough to match, and that lock on OEM preloads
> > (determined to be illegal, but too late to save OS/2). With OS/2 one can
> > say that everything that could go wrong, did go wrong, all of its life.
> >
> > So it's arguable that OS/2, although a better fundamental design, wasn't
> > well enough executed, delivered or supported to make it better software,
> > and it was badly handicapped by its vendor plus Microsoft's shenanigans.
> >
> > I don't regard OS/2 as a valid counter-example for all of these reasons.
> > One can imagine that IBM's OS/2 debacle is one reason for it deciding to
> > support Linux at arms-length -- without trying to build some IBM distro.
> > The other reason is probably its overhead and limited success with AIX.
> >
> > Egan
> >
> > PS -- I still have an OS/2 partition, but I hardly ever boot it anymore.
> >
> >
> > Robert Comer wrote:
> > >>I offer Windows 3.x...95 vs. OS/2 as a counterexample.
> > >
> > >
> > > Well, that kind of squashes Egan's law.
> > >
> > > - Bob Comer
> > >
> > >
> > > "Jeff Shultz" 
wrote in message
> > > news:3e615353{at}w3.nls.net...
> > >
> > >>Egan Orion wrote:
> > >>
> > >>
> > >>>An update of Gresham's Law... for software:
> > >>>
> > >>>http://www.theinquirer.net/?article=8059
> > >>
> > >>I offer Windows 3.x...95 vs. OS/2 as a counterexample.
> > >>
> > >>--
> > >>Jeff Shultz
> > >>I don't speak for anyone, and only One speaks for me.
> > >
> > >
> > >
> >
>
>

--- BBBS/NT v4.01 Flag-4