Received: from uucp-gw-1.pa.dec.com by s.ms.uky.edu id aa06737;
          15 Jan 92 22:57 GMT
Received: by uucp-gw-1.pa.dec.com; id AA24427; Wed, 15 Jan 92 14:50:37 -0800
Received: by limbo; Wed, 15 Jan 92 14:40:49 pst
Date: Wed, 15 Jan 92 14:40:49 pst
Message-Id: <9201152240.AA06485@limbo.intuitive.com>
From: Info-Source <info@limbo.intuitive.com>
To: abbott@ms.uky.edu
Subject: Re: Your Request For Information
X-Mailer: The Info-Source System


(send Digital.Games/Digests/Digest.34)

 
                        Digital Games Review, Number 34
 
                           Monday, November 4th 1991
 
SPECIAL ISSUE: 
  		      Mac: Microsoft Flight Simulator 4.0

----------------------------------------------------------------------
 
From: rdd@cactus.org (Robert Dorsett)
Date: Wed, 23 Oct 91 10:52:22 CDT
Subject: Mac: Microsoft Flight Simulator 4.0

Summary:

A very nice port of the IBM PC version of Flight Simulator 4.0, to the
Mac, producing the best VFR simulator for the machine.  Essentially the 
same basic game premise as FS 1.0 for the Mac, but with:
        - Color.  
        - A conformant Macintosh user interface.  
        - Much more realistic aircraft.  
        - Much superior graphics.  
        - Much more versatile weather.  
        - Dynamic scenery.  
        - Configurable aircraft (plus a new aircraft, a Schweizer S-32).
        - NO copy protection.

The major conceptual loss is that of the rudimentary WWI Ace dogfight
game. Two new entertainment options (crop duster and air races) have
been substituted.

The up sides: the color, the more realistic aircraft.  The pleasure of
being, at least for a while, one very large qualitative step ahead of
the PC version of the game, which has tended to be much better
supported than the Mac version.

The main down side: all but unusable on monochrome, older, slower
machines.  Your mileage will vary.  It seems to run fine on IIcx's and
better, but I've heard complaints from people running on even II's and
IIX's.  

It was suggested I assign a numerical "grade" to the application.  Kind of
difficult.  We're talking about a game whose premise hasn't changed much in
over twelve years, and whose scenery hasn't changed in ten years.  Instead, 
it keeps getting shinier and shinier.  Either you like flight simulators or
you don't: if you still found FS 1.0 interesting, but tired of the interface,
this one's a lot of fun.  If you were bored by FS 1.0, this one will probably
bore you, too.  With that in mind, assuming that one has NEVER HEARD OF
Flight Simulator, and who has only played, say, Dark Castle on the Mac:
        Graphics        9.9
        Realism         6.5
        Playability     9
        Sound           6
Overall Rating: 9, on a scale of 1 to 10.
        
                   

Systems Used:

The following is based on play on a Plus (6.0.7), an SE/30 (System 7 w/
color monitor), and a IIsi (System 7, color).  The most time was spent on
the color machines.

Background:

Microsoft marketed Macintosh Flight Simulator 1.0 in 1985, as a
symbolic extension of its professed commitment to the Macintosh product
line (prior to this, it only dealt with the MS-DOS version).  The Mac
version was a success, but was a lousy product: it wasn't usable as a
"normal" Macintosh program, had its own, non-standard user interface,
(which was quite unpleasant to use), had very unrealistic aircraft,
graphics which were painful (mainly due to an irritating, mandatory
shading scheme), and quickly started to die on the newer machines.
Never mind large monitors, different processors, or color: it became
all but unbearable for those who merely had hard disks.  By late 1986, 
users had started to clamor for a Flight Simulator upgrade.

Microsoft Flight Simulator 3.0 was announced for the PC nearly three
years ago.  It revamped much of the earlier FS2 product, providing much
more sophisticated, crisper graphics, which took advantage of EGA/CGA
modes.  It also provided a better user interface, and somewhat better 
aircraft characteristics.  It adopted a "pseudo-window" interface, and 
offered many new features: dynamic scenery, weather generation,
approach strobes, VASI lights, a nice-looking instrument lighting
scheme, better aircraft fidelity, the potential to easily customize
aircraft appearance and performance, etc. It had at least a few
problems; Flight Simulator 4.0 was released last year, to fix those
problems, and to take advantage of the groundwork established in 3.0.
The Microsoft Aircraft and Scenery Designer was a separate product,
released in support of the new generation, and permitted the user to
design his own scenery and to customize aircraft performance.

The Macintosh version of Flight Simulator 4.0 breaks totally with the
old Mac 1.0 version, providing nearly all of the PC version's features, 
but also adding much crisper graphics, a limited aircraft customization 
mode, and a new, vastly superior, user interface.  All with decent speed, 
on the current generation of faster Macintoshes, and with an apparently 
unlimited life-span, capable of utilizing faster, newer architectures 
as they are released.

User Interface.

The user interface is straight Mac.  However, coming from FS 1.0, it's
a pure relief to see that they're not doing any of the low-level,
machine-dependent tricks Flight Simulator has become famous for.  It is
clearly written with an eye on the future.  This version should not die
nearly as quickly as the old one did; we should still be able to run it
fine, in ten years, when they next turn their attention to the Mac
again. :-)

Fidelity.

The aircraft fidelity in FS 4.0 is much better than the earlier Mac
version.  FS 1.0 essentially provided an "eye in space," with very
rudimentary flight characteristics.  Performance was a direct function
of pitch, rather than angle of attack, for instance: as a result, one
could not attain a positive pitch while maintaining a rate of descent--
thus making it impossible to land the airplanes properly.  FS 4 fixes
these problems with a new flight model.  Not only does flaring work,
but other, minor characteristics--such as actually having a realistic
rotation before lift-off, or accelerated stalls--are supported.

There are limitations to this: the ground taxi model is still lacking,
and there are still no signs of phugoids (minor oscillations in pitch
and yaw), which better simulators strive to present.  One does not
crash if one goes off the runway--a feature of FS2, even, if I recall
correctly.  Flight at the edge of the operational envelope is still way
too perky.  But at least there does appear to be an envelope, now:
there are definitely aircraft performance changes at high altitude, at
least.

The Learjet mode still needs a lot of work.  One has little feeling of
flying a jet.  Sure, it goes fast, but it's still essentially a
rocket in space: we have no feel for the limitations of a jet or its
engines, do not experience high- or low-speed buffet, and suffer 
no ill effects from seriously abusing the aircraft, at either high or 
low altitudes, or low or high speeds.  If one flies a Learjet like an 
F-16, one can reliably expect to quickly pull the wings off; this does 
not happen.  No sign of them even being bent. :-)  The one concession 
is that if one exceeds VM0, the elevator control position indicator 
jiggles, apparently to show flutter (there is talk of shock waves in 
the manual).  This doesn't appear to be a fatal situation, and goes 
away as soon as one slows down, with no penalty.

The incredible speed discrepancies of the 1.0 version are gone.  In the
earlier version, in Lear mode, we got to fly approaches at
approximately three times the indicated approach speed on the TAS-meter
(i.e., show 150 knots TAS, actually fly in at 450).  This required REALLY 
long finals (5+ miles), to get set up properly.  The new one can be 
reliably maneuvered and landed starting within 1 mile of the approach 
end of the runway.

Instrumentation.

The instrumentation can automatically adjust to the size of the
screen:  there is a "small" (512 x 342) and "large" (768 x 480) mode.  
The program recognizes multiple monitors, and saves window positions.

The brain-dead !#@!$! "True Airspeed" indicator (as opposed to
Indicated Airspeed) for the Learjet is still with us.  The explanation
for this in the manual is sophistic nonsense.  I guess they think the
pre-teens want to fly fast, and don't want to add a groundspeed
indicator, so they screwed up the instrument cluster.  I think this is
a cheap escape.  For the purposes of THIS simulation, TAS is only 
significant in navigation, i.e.,  when computing groundspeeds; one flies 
the airplane through indicated airspeed.

The artificial horizon is much better than the 1.0 version, but I still
don't get much of a pitch feedback from it.  In that respect, perhaps
it is fairly close to how real horizon gyros in smaller GA aircraft
work.  However, positive and negative displacements don't seem to
result in the same angular displacement in the out-the-window view.  One
can be ten pixels nose-high, and still have the horizon in view, yet be
two pixels nose-down, and have the screen full of earth.  In color
mode, it's also hard to distinguish where the airplane symbol is, if the
airplane is nose- down.

The artificial horizon also doesn't have a "roll indicator"; thus, one 
doesn't have an accurate indication of bank.  Even 1.0 had a roll 
indicator.  A weird one, mind you, but it worked... :-)

The vertical airspeed indicator for the Learjet shows more realistic
graduations, from 0 to 8000 fpm.  The FS 1.0 version used the same VSI
as the Cessna.  However, FS 4 shows the same graduation intervals
between 0 and 1000 fpm of climb as it does between 1000 and 2000, and
so on.  Real-life jet VSIs have a staggered, non-linear presentation,
so as to more accurately track landing low rates of descent (i.e., the
angular distance between 0 and 500 fpm is the same as that between 500
and 1000, 1000 and 2000, and so forth; this allows more precise
correlations of vertical speed to airspeed, when flying ILS approaches).

Both the VSI and the altimeter seem to have pressure lags, which is a
nice touch.  At least, I HOPE it's intentional. :-)

The Learjet has a 4-minute turn indicator, for some reason.  Back when
jets had turn indicators installed, they were all 2-minute indicators.
4-minute indicators aren't necessary at terminal maneuvering speeds: 
a maximum of 200 knots under 6000' in the United States (AIM para 372, 
FAR 70). 200 knots requires about 25 degrees of bank for a 
standard-rate turn, or about 1.1 g's.  No strain on the aircraft or 
its occupants.  This is all very trivial, UNLESS one is trying to do 
procedure turns or holds, or stay within the sector boundaries while 
doing a procedure turn.  You only start to need a 4-minute indicator 
in fighters.

I'd like to see a toggle between a turn coordinator and a needle & ball
format.

At long last, they appear to have gotten the ADF right.  It points in
the right direction (few--if any--of the versions of Flight Simulator,
on any machine, ever, have had a properly working ADF; it seems to be 
the all-time major bone of contention among instrument-rated pilots
who play with the game).

Still the "one throttle does all" for the Learjet.  No sign of
differential thrust if one "turns off" one engine or the other.

The instrumentation layout on small screens is almost identical to FS
1.0's.  It is quite readable.  The layout on larger screens is larger
(1.5" instrument faces).

The autopilot is more usable than the FS 1.0 version (with the same
features), but still comes pre-set to a 500 fpm climb or descent on
altitude captures.  This might work with the Cessna, but is unrealistic
(intolerable?) with the Learjet.  The autopilot is still set through a 
dialog.

Graphics.

The graphics of FS 4.0--in color--are splendid: we're treated to
dynamic scenery, approach strobes, VASI lights, "light-points" to fill
out the countryside, clouds, runway lights, taxiway lights, very nice
dawn/dusk views, and good views of the aircraft.  In many ways, the 
graphics (especially with with the use of light-points in night mode) 
are comparable to the display formats of the "real-life" flight 
simulators of the early 1970's.

The graphics also look good in grayscale.

The graphics in monochrome mode roughly boil down to the hellish
presentation of the 1.0 version, which was tolerable only if one flew
with the simulator clock set to the night or twilight hours.  This
approach also works with FS 4.0--especially with shading turned off
(there's a "daylight shading" mode, even with shading technically
turned off, which is aesthetically unappealing; there's no way to
completely kill the shading and get left with wireframe graphics).

The performance of the graphics is unexpectedly good: refresh rate
is a function of the complexity of the scenery being displayed: one 
can get about one frame per second with *everything* turned on, flying
through downtown San Francisco, but it gets much faster if one kills
the more frivolous items (stars, ground and air traffic, dynamic
scenery, landscaping, etc) and flies in a less "populated" area (the
default of Meigs Field works fine). 

Scenery.

The program ships with scenery for New York/Boston, Illinois, western
Washington, and the Los Angeles and San Francisco areas.  A README file
on the distribution disk offers seven other sectors (including Western
Europe, Hawaii, and Japan) for $59.95 for a combination of three.  This 
pretty much boils down to the old $19.95 a sector that Sublogic used to 
ask for.  The sectors are sold direct from BAO, not Microsoft.

The program will apparently recognize the old Sublogic scenery disk
format, but loses color information.

I have not yet attempted to see whether the program will accept data
from the PC Aircraft/Scenery Designer (not quite true: I did try, but
the horrible user interface quickly dampened my inclination to finish
the project--if they DO try an Aircraft/Scenery Designer for the Mac, I
hope it'll be done with a proper CADD interface).

Sound.

The propeller engine sound seems different (better) than that used in 
FS 1.0, but it could just be the Macs I ran it on.  The Lear sound is 
almost identical (but that one was quite good).  Marker beacons are now
supported.  There is also now a crash volley, which is great.  Using
sound doesn't appear to slow down the simulation much, although it will
play only one sound at a time (for example, engines go quiet if one
crosses the outer marker, to be replaced with the OM signal).

Navaids.

No sign of any sort of universal navaid structure.  Navaids are still
confined to the sector areas; therefore, still no chance of an IFR
cross-country flying between sectors (it'd be especially nice between
northern and southern California).  There appear to be several more ILS
approaches (just about every runway at ORD, for instance, and two more
at OAK), and more NDB's.

There are dialogs to set navaid values, but these are just as
cumbersome to use as the controls on the instrument panel.  To switch
the bearing selector on a CDI 180 degrees, for example, one pretty much
has to click 90 times to get it to turn (2 degree increments each
click).  I miss the "click and hold" option of FS 1.0.  Barring this, a
SMALL dialog, which would pop up, be drawn quickly, and let the user
type in the value, would be satisfactory.

VOR's all seem to have the obligatory DME associated with them, as do
the ILS's (not realistic).  VOR's also still unexpectedly produce a
glide slope bar.

Input.

FS 4 works fine with a mouse; it's easy to land both the Cessna and the
Learjet, regardless of the realism selected.  Still no sign of an
evolved control law, but the direct-response mechanism (complete with
the control indicator tree on the instrument panel) lets one control
the aircraft fairly well.  Still a tendency to encourage the user to
fly by reference to control positions, rather than by the
instrumentation.

There is a provision for both joysticks and yokes.  These generally
involve additional hardware, and possibly driver packages, PLUS the
actual joystick/yoke.

The package comes with a brochure for the Advanced Gravis Mousestick,
which is probably one of the cheapest "add-on" input devices
available.  I purchased one a year ago, in support of Yeager's
Advanced Flight Trainer.  It's a neat little gadget, with the necessary 
drivers in firmware, but is not a satisfactory input device for a 
flight simulator of any type, let alone this one.  The essential
problem is that one cannot comfortably apply back pressure to the
stick: the heel of one's hand gets in the way.  The stick is too
short, too narrow, and the base is way too wide.  One CAN control it
fairly well with one's fingertips, but a mouse is real good for that,
too.  Save your money.

For the true masochist, an elaborate keyboard control structure is
supported.

Manual.

The manual is odd.  On the one hand, it has a strong visual appeal:
various elements of the layout and graphics are well done.  On the
other hand, it has a very low signal to noise ratio.  The actual text
is very muddy: for example, no attempt at all is made to hilite the
discussion of individual menu commands (no bold or italicised text, or
quotes) as they appear in the text.  It reads very much like yet
another marginal translation of a PC manual, which is odd, since the PC
version's manual isn't much like this one.

A major gripe I have is the cursory treatment given to the Learjet mode.
"Flying by the numbers" is very important with jets.  One should at least 
be provided with ball-park figures of what constitutes a suitable approach 
speed, flap setting, etc. to track a 3 degree glide slope.  The manual 
ignores all this (apart from a list of stall speeds in an appendix), pre-
senting instead a hashed-down version of the Cessna chapter, thereby 
suggesting that it be flown by the seat of the pants.  Considering how 
much trouble many people have had with figuring out the Learjet, I had 
hoped to see more attention given to this (from my experiments, flying an 
approach at 140-150 knots in still air with 40-42% power, gear down, 
flaps full, should give you a 750 fpm rate of descent, which will track 
the ILS glide path).

A good job was done with the program's user interface; it's a shame the
same effort wasn't put into the documentation.  At 230 pages, you've got 
to wonder what they had in mind...

Bugs/Problems

Some of the following are conceptual problems, which have ALWAYS been
with the product (from FS2 onwards), but which never seems to get fixed.  
It seems like the basic airplane never gets changed; instead, max. effort
is placed squeezing ever more performance out of the graphics package.

I've experienced two fatal bombs, in approximately 15 hours of play.
These bombs appear to be fairly infrequent, but completely wiped out my
system, including Macsbug.  They appear to be related to the dialogs.  
Fortunately, saving situations is easy enough now that it's easy to restore
one's position after rebooting. 

When in "Yoke" mode, flying with the mouse, occasionally clicks (such
as to adjust throttle) may be intercepted by System 7 and result in
whatever's behind FS 4 being pulled to the front.  It's a pain.  I
suspect that the mouse location may be getting lost after some menu
commands are invoked, or a dialog is pulled up.

The controls can get stuck in strange positions, especially after "Set
Location" altitude changes.  For example: level flight, full power,
requires total mouse-down to maintain level flight.  The slightest pull
upwards results in a gross nose-up, and a quick stall.  I vaguely
remember this problem with 1.0, too.

The brakes don't always work.  Occasionally, brakes get set and WON'T 
go off.  When they do work, even the Learjet can be stopped within 2000' 
or so--and it's not the abrupt "instantaneous" stop effect of FS 1.0, 
either.

The keyboard-driven brakes are not equivalent to the mouse-driven 
brakes, although the manual seems to apply they're the same critter.  
The mouse seems much more reliable.

Occasional weird behavior in dialogs.  Non-blinking insertion points,
for instance.

Users of small monitors (Plus, Classic, SE) could find the use of full-
fledged window title bars somewhat annoying: this can conceivably kill
up to 60 pixels of potential screen space.

The Save/Save As commands don't work as one might expect.  The "Save
As" command has an option to save an aircraft with the startup
situation.  Thus, if one flies the sailplane all the time, one can open
a saved game, and not lose the aircraft information.  However, if one
subsequently does a "Save" (command-S) on the currently-running game,
it does NOT save the aircraft.  Fine, you say: just do a Save As all
the time.  But the Save As *sometimes* defaults to the "My Situation"
prompt, and loses the name of your saved situation!  The ";" keyboard
save command always produces Save As, with the correct file name.

My suggestion: remove the toggle, and ALWAYS save the aircraft with the
game.  I'd wager that users want to start from a PARTICULAR location in
a PARTICULAR aircraft more than they'd just want to save a location and
start up in the default aircraft (Cessna).  The user can always reset
the aircraft to something else later.

The weird FS coordinate system is still with us (an obsolete anti-piracy
measure, perhaps?).  We can, however, go directly to airports by name,
now.

The trim is still very poorly designed.  I am certain there is no
reliable way to set it, and the presentation in the manual is all but
irrelevant to that task.

The range on the DME's is still hardwired, and unrealistic.  At 30,000',
one should be able to receive a VORTAC (H) while still 300-odd miles
away.  At ground level, it shouldn't be received more than 20-odd miles
away, terrain permitting.  The ones in FS 4 are hardwired to 80 miles
max. range, regardless of location.  The navaid structure also seems to
be linked to the graphical sector dimensions, which is a mistake: one
can cross a sector boundary and suddenly receive a VOR only 10 miles
away.  

There are more ILS approaches, but the model doesn't seem to have
improved much.  One can track an ILS fine, until the last 2 miles or so
to the runway, when the needles start jumping around.  Don't set your
cloud base under 600', or you'll have to do a lot of scooting around to
land.  Some localizers (e.g., OAK 11) seem to be significantly canted
away from the centerline of the runway.  Such critters do exist, but
not at any of the FS airports.  Properly configured, one should be able
to track an ILS from the outer marker to decision height without any
major course adjustments (in still air, at least).

There are now three ILS's at SFO, but none seem to work at all.  The
southern approach to OAK also seems to have an unacceptably short range.
The ILS model now seems to indicate a reception, even if the "signal"
isn't reliable.

Interestingly, although the ILS has some really weird characteristics,
the VASI bars seem to work fine throughout the approach (i.e., when one
is close enough to see them :-)).

It'd be nice to have a "ground roll" component to the landing analysis
graph.  Obviously it's tracking ground roll, and figuring out a scale, 
so why not just totalize it?

It would be nice to see ALL the user-configurable options put in one
place.  I.e., move the "realism" toggles to the Preferences dialog, and
turn the Preferences dialog into another "options" menu.

The EFIS mode is interesting, but I have one basic question: why is its
altitude control oriented around AGL, whereas the autopilot works on MSL?
It'd be nice to have the EFIS and A/P running in synch.

IMHO, the Lear model *should* be polished a bit more.  I don't know
how most people play the game, but from *my* perspective, if I'm going
to play with the instrument approaches, or fly around in the World,
I'm going to want to do it in a faster airplane--not at 120 knots in a
Cessna 182.  It's a question of wasting SOME time, or taking three or four 
times that time...  I would wager that a lot of people would take a similar 
position; therefore, the Lear mode would seem to be a high-visibility 
feature, and should be made at least as usable -- from the instrumentation 
perspective, if not performance limitations -- as the Cessna mode.

Caveat:

My aviation orientation is toward airliner simulation, so my views may
be somewhat skewed.

Disclaimer:

I have NO business or personal relationship with BAO or Microsoft,
whatsoever, either directly or indirectly.

Glossary:

ADF       Automatic Direction Finder.  
AGL       Above Ground Level
AIM       Airman's Information Manual.  
BAO       Bruce Artwick Organization.
CDI       Course Deviation Indicator.  
DME       Distance Measuring Equipment.  
FAR       Federal Aviation Regulations.
FPM       Feet per Minute.  
FS 1.0    Flight  Simulator 1.0 (Microsoft, Mac version).  
FS2       Flight Simulator 2 (Sublogic/Microsoft).  
FS3       Flight Simulator 3.0 (Microsoft).  
FS4       Flight Simulator 4.0 (Microsoft). 
IAS       Indicated Airspeed.  
IFR       Instrument Flight Rules.  
ILS       Instrument Landing System.  
Knots     Nautical Miles/Hour.  1' of longitude per hour.  
MM0       Maximum Mach--operating.  
MSL       Mean Sea Level
Navaid    Navigational Aid.  
NDB       Non-Directional Beacon.
OAK       Oakland.  
ORD       Chicago O'Hare.
PC        IBM Personal Computer & Compatibles.  
TAS       True Airspeed.  
VASI      Vertical Approach Slope Indicator.
VFR       Visual Flight Rules.  
VM0       Velocity--Maximum Operating.
VOR       VHF Omni Range (navaid).  
VSI       Vertical Speed Indicator.  
WWI       World War I.

-------------------------------------
 
To join this group or have your reviews appear in the next issue, please
send electronic mail to Dave Taylor at the following address;
 
		digital-games@Intuitive.Com
 
We welcome reviews of games for any computer system, including the IBM
PC, Apple Macintosh, Atari, Nintendo Entertainment System, Atari Lynx, 
Sega Genesis, Nintendo GameBoy, NEC TurboGrafix, Amiga and so on.

Back issues are archived as digests, as well as broken into individual
reviews.  To receive them via electronic mail, send a note to the
address "info@limbo.intuitive.com" (or decwrl!limbo!info) with the
body containing the lines "help" and "listall Digital.Games".  To 
obtain them via anonymous FTP, connect to "eklektik.cs.pitt.edu"
[130.49.2.135] and look in 'videogames/digest'.

The views expressed in Digital Games Review are those of the individual 
authors only, and do not represent the views or opinions of Intuitive 
Systems or any of our affiliates worldwide.  The copyright of each 
article is that of the original author, and the copyright of the digest 
itself, as well as the name "Digital Games Review" are stated here.
DGR is also registered with the Library of Congress: ISSN 1059-5457.

*********************
End of Digital Games Review
*********************


