IFR Simulator from Programmers Software is a hi-res simulation of flying an aircraft on instruments. It has been around in various guises for some years, but this is the latest and most sophisticated version.

It's aimed (according to the manual) at those who want to understand the workings of aircraft instruments and navigation, not at training real pilots.

Quite why anyone would choose to learn about such a rarefied subject without at least wanting to become a pilot, though, I can't imagine.

Don't think that this is an alternative to the classic flight-simulator, A2-3D1 from Sublogic, though. It isn't the same thing at all.

IFR Simulator is intended to train you in the complexities of blind-flying. For this purpose the 'basic six' instruments must be available. These are altimeter, airspeed indicator, artificial horizon, turn and bank indicators and rate of climb or descent indicator.

Also included is a set of radio-navigation instruments, namely two OMNIs and an automatic direction finder. These are the real reason for the program, and it is intended to train you in their correct use.

Training is done by 'flying' a series of problems, which range from simple to very difficult. The task is to orientate yourself by using the OMNI and ADF, then to fly to a given position. If this is done successfully it is assumed that you could have made a successful landing from that point.

Whenever your altitude drops below a minimum, the session is over, and a plot of your flight-path is shown, since you have effectively landed. Naturally, doing so anywhere but on the runway is a crash, but the program doesn't take any notice if you do.

The documentation would be good if it were better organised. It has all the information required, but it's in such a mess that it's not a lot of use.

The manual is OK if you know a bit about instrument or radio-flying, but rather confusing otherwise.

Getting into the pilot's seat is a matter of menu-driven teleportation. You boot the disk and are presented with a menu of problems. Selecting one of them immediately transports you to the starting position.

The problem is shown on the bottom of the screen in the typical terse, cryptic style of the airways. For example: "LR NDB RWY 4. Direct to NDB 344 Out Procedure trn. 164 inbound to runway FAF is bottom OMNI"

From there on in, you're on your own. Happy landings ... mind you, it isn't easy.

The actual flight characteristics of the plane are noticeably different from the Sublogic simulator. It seems much less jumpy but rather less stable, and if left unattended in level flight, has a habit of wandering without any noticeable tendency to correct.

The speed of response if quite adequate, because most of the real work is done in machine code.

The use of the Apple's facilities is excellent. The instruments, both conventional and electronic, are beautiful, and their response is uncannily accurate. But my main criticism is against the 'out-the-window' view. This is a picture of an airport runway, stretching away into the distance, with several large buildings on either side.

When the plane turns, the view scrolls sideways. It's beautifully done, but utterly pointless since it doesn't respond as a real horizon.

The main trick when you're flying this thing is not to worry about the apparent lack of response from the OMNIs, as you barrel along waiting for something to happen.

As long as you're flying towards the beacon, not away from it, the OMNI will suddenly swing over as you cross the beam. Only then do you get some idea of where you are.

The joystick or paddles are used to fly the plane. P adds 3 per cent more power each time it is pressed. L subtracts the same amount.

The only other valid Keypress is ESCAPE, which gives an X-Y plot of your course so far.

The authors don't approve of doing this, so for some time after looking at the plot, the compass is dead. Pendantry.

To build new problems, you need to use the Seed program, which contains the actual flying problem. Installing a new flying exercise consists of making some minor changes to the Basic code. This, the authors suggest, is not something you would want to do yourself, so they advise you to send them the details of your approach-problem together with $100, and they'll do it for you.

I bet they will - but I can't figure out why they didn't put all the relevant details into a file, and add a little file-manager to take care of it.

After making a thorough mess of the first few problems, I began to get the hang of it, and apart from the irritating display, found it quite fun.

It isn't strictly a game... it's a real simulation, and as such is instructive. Certainly there is nothing like it on the market.

Overall, I would say that my feelings were of modified rapture. It's a good solid program which is interesting and different.