Lua problem on latest versions of FSUIPC

[Artem Crum]

Hello Pete,

Here is the example script (test.lua) which is not working as expected:

local i = 0

function test ()
   i = i + 1
   ipc.log(i)
end

event.timer(100, 'test')
[/CODE]

1. Load any flight

2. Start script by "Lua Test" joystick button assignment in FSUIPC

3. Check fsuipc4.log for constantly increasing numbers

4. Change current aircraft to any other

5. Script STOPS working after new plane is loaded (no more numbers in log)

On previous versions of FSUIPC script continues to work after plane change. On latest version I have to restart script manually each time I change aircraft while in flight.

[Pete Dowson]

On previous versions of FSUIPC script continues to work after plane change. On latest version I have to restart script manually each time I change aircraft while in flight.

Please ALWAYS give version NUMBERS. I've no idea what you mean by "previous " and "latest"! The latest one uploaded for tests by one or more users was 4.825, the previous one 4.824. The last "report" of any problem I received here quoted 4.703 as "previous" and 4.80 as "latest". I really can't work things out without proper reports. All releases are numbered, and this is the reason why, so I can keep track of things.

It's a bit late here at present for me to start looking at this. I'll check it with 4.825 in the morning.

Regards

Pete

[Artem Crum]

I'm sorry. My "previous" is 4.756, "latest" is 4.824.

Several users reported the similar problem after updating from 4.80 to 4.82. Script stucks randomly (i.e. after opening FSUIPC settings) which was never happened on 4.80 and earlier versions.

UPDATE: got the confirmation that 4.80 works fine and doesn't stuck the script on aircraft change, but 4.82 - does.

Edited April 25, 2012 by Artem Crum

[Pete Dowson]

UPDATE: got the confirmation that 4.80 works fine and doesn't stuck the script on aircraft change, but 4.82 - does.

Okay. Found it, thanks. The change was the suspension and resuming of Lua threads when the DirectInput devices are being re-scanned. There was a path through this which exited without doing the "resume" so all running Lua threads remained suspended.

Ironically the change was to prevent FSX crashes which appeared to be due to LINDA activity simultaneously with FSUIPC's rescans.

Sorry about that. Wish it would have been found earlier as there have been at least six releases since I made that change. :-(

FSUIPC 4.826 is being built right now. This also corrects the separate Lua logging, where the time value got stuck at the end and the Lua.n number was unnessessarily included (only needed when all Lua logs merge into the FSUIPC log).

I'm awaiting confirmation that 4.825/6 also handles the G3D.DLL crashes in Prepar3D. Then I'll build a complete new Install package.

Regards

Pete

[Artem Crum]

Good news, thank you!

[guenseli]

Many thanks from me, too.

Seems to work!!!