Wirt wrote:
>
[snip intro]
> In one, an "open systems" approach is
> adopted. In the other, a well-engineered, well-integrated, optimized approach
> is used.
>
Then why does it seem so hard to provide good performance on the PowerPC Mac?
Most apps I've looked at run much faster under Win95 on a Pentium than
on a Mac with a PowerPC - with the same ram and Mhz processors.
If it is well-engineered, wouldn't it be easy to port to the PowerPC?
Was it hand optimized because Apple used to be able to afford it, but
can't now, and all that hard work is missing from the PowerPC versions?
 
> Ultimately, the second approach is the only one that is going to succeed. If
> computers recapitulate the history of cars, the only people who will
> eventually buy a car as an "open system" (where you have to buy the seat
> covers, fenders, and fuel injectors from whomever manufactures those parts --
> and put them together yourself, trying to resolve "system" conflicts and
> incompletely adhered-to "standards") will be teenagers and neophyte
> mechanical engineers.
[snip end]
>
 
I'd like my computers to run a bit better and protect me better than any
recent generation of a car.  MPE might be like a NASCAR stock car from a top
team - it takes a beating, but keeps on running and, most important, doesn't
damage the contents - yet it gets good performance, too.  Yet, nobody much
could afford this in an everyday car. Maybe a Volvo is a better analogy (don't
want to upset, too much, HP's view MPE's market)?
 
Cars are an example of just what I don't want to have happen with computers.
For the most part, cars are very similar and might approximate "openness" the
same way Unix is open - they are all different in details, but mostly,
bascially, the same.  It is the exceptions to this generalization that may
apply to MPE and to the Mac, if they apply at all. More recently, there has
been more innovation in car design, but I think that is beside the point.
 
Microsoft may become more and more like GM, in that it plays by the rule that
it can afford to irritate and have flaws, if a *lot* of people buy the
product anyway because it is good enough and it does the job needing doing.
Most Microsoft upgrades make it easier to do the work I'm trying to do -
through the
price of the features is reliability.  On the Mac, I have to pay more (for
software, and it appears, I'll have to pay more for hardware once again,
soon), to get fewer upgrades, fewer bug fixes, and lower performance - to get
what? A machine that crashes less often than a Windows PC? Even with crashes,
if I get more work done on the PC than the Mac, I'll take the PC anyday and if
it saves money - that's even better!
 
I use a new Powerbook 5300.  It came with System 7.5.  I got the Apple
Internet Connection Kit to do remote PPP because it included support for the
lastest tcp/ip implementation from Apple (Open Transport).  But, it turned
out, I need some fix level of System 7.5 to use it, so I'm using MacTCP
instead.  Then, I downloaded the most recent version of Netscape Navigator,
and it came with a warning that to use Open Transport I needed an even later
fix/update level.  I can download it from Apple - it is only 21 megabytes.
Even with the MacTCP, PPP only works on some of our dailup lines, and not
others, even though all of our lines use the same modems, the same software,
and the same hardware (Shiva's LanRovers).  Finally, I have to restart the
Mac everytime I want to try something different (like changing the phone
number) with the PPP dailup connection! What a waste of time.
 
On the other hand, Windows 95 installed on my old Zenith 386 (with 8
megabytes), runs faster than Windows 3.x did, and handles my modem and dailup
PPP out of the box.  It works on all the dailup lines without fail.  I can
reconfigure most things - like changing the number to dail, without
rebooting.  The only patch I've needed is for the Novell client to support
NDS - which is networking I haven't even gotten to try with the Mac.
 
There are a lot of things to like about the Mac, just like MPE, but the value
of the difference can easily be outweighed by the costs, for many normal business
environments, if it is not carefully controlled and managed well.
 
Richard
--
-- - - - Speaking for myself and not necessarily anybody else - - - - - -
Richard Gambrell        | Internet: [log in to unmask]
Mgr. Tech. Services     | POT:      504-483-7454     FAX: 504-482-1561
Xavier University of LA | Smail:    7325 Palmetto, New Orleans, LA 70125