Truth is that autopilots in real life are usually custom built around an aircraft so we are doing things in FS that are far more difficult. Trying to build a generic autopilot to fit several aircraft is a much tougher challenge. As you rightly say, testing the combinations for just one aircraft can be daunting on its own - which is why I have been glad my own work is confined to specific aircraft. bones -----Original Message----- From: jhb_airlines-bounce@xxxxxxxxxxxxx [mailto:jhb_airlines-bounce@xxxxxxxxxxxxx] On Behalf Of FrankTurley@xxxxxxx Sent: 12 November 2006 20:00 To: jhb_airlines@xxxxxxxxxxxxx Subject: [jhb_airlines] Re: Another FSX Foible In a message dated 12/11/2006 18:24:35 GMT Standard Time, bones@xxxxxxx writes: With your autopilot I suspect that you will find lots going on in the background that are not obvious and maybe a test readout gauge may be worth designing. Does the ATT switch disconnect the ALT? Does CRS disconnect HDG? Or do you have to code this in? The latter would be better as you would know exactly what is going on but, unfortunately, MS have built logic into their code that you sometimes do not want - and it is frustrating that you can't circumvent this. Hi Bones, Phase 1 was building a series of indicators to show the status of the various autopilot functions, which I found needed to be tested in a number of FSX models as there was a good deal of variation in functionality between aircraft types. Phase 2 was building a series of switches to switch single functions on and off, which is where I discovered there was interaction going on, some caused by the .CFG settings and some from within FS itself. This got quite complex and tired my poor brain out, so I'm afraid my vanilla autopilot project came to a halt. When I feel up to it I'll try to do some comprehensive testing, but its an amazingly huge task, just testing the various combinations for one aircraft wil take ages, and there are a number of aircraft to try out. I guess a set of rules would eventually emerge, but whether they called all be plumbed into a single gauge is open to doubt at this stage. I guess autopilot functionality is extremely variable in real aircraft too! I'm currently doing a radio stack, still learning about XML, but so far I've not encountered anything that doesn't work as expected. The radio functions do appear consistent across the aircraft range, and the .CFG definitions appear to be observed by the system. Much easier on my poor brain! Frank T.