Page 3 of 5
Posted: Fri Feb 27, 2015 12:46 pm
by MagisterXF94
Shall we do a beta test anytime soon?
Posted: Wed Apr 01, 2015 9:43 am
by Dome
jbanks sent me this:
Dome,
I have discovered a flaw in a controller sensitivity fix that was introduced into R8. In WinTrek.CPP the value of "m_pjoystickImage = CreateJoystickImage(1.0f / 2048.0f)" on line 2900 reads 1024 instead of the 2048 that it should be.
When Djole copied this value over from Imago's github repository, he copied it before Imago put in the proper value of 2048. So, Djole has 1024 in R8 and Imago has 2048 on his version.
I have tested this value difference and found that Imago has it correct. Using the 2048 value makes the joystick feel proper in both linear and quadratic settings in Allegiance. Furthermore, with the proper 2048 value, the deadzone is more round than square giving a better aiming capability when tracking diagonal targets.
If you would like a patch for TRAC, I can forward the fix in a patch.
_____________________
(Another PM from jbanks)
I provide a patch to correct the joystick sensitivity error. Please get this pushed into the next beta test release.
Thanks
TRAC
____________________
Is this something we can use? Djole?
Posted: Wed Apr 01, 2015 1:34 pm
by jbansk
Thanks Dome,
I see what happened now. On the 27th, it was at 2048 (which made the stick behave properly), then two days later he moved it to 1024 (which is too sensitive and makes the deadzone feel square.
The version I have from Imago's github was prior to it being changed to 1024 so it wasn't a copy / paste error as I first thought. (sorry Djole). The value of 2048 is noticeably correct when you change this value to anything else. I tried it with different values then kept switching back to 2048 to see the difference. After awhile, you become very sensitive to the changes and can see that the 2048 makes the stick work in all axis with the proper smoothness and most importantly is that the deadzone no longer feels square.
After the last beta test, Djole informed us that he wasn't working on matters that he wasn't directly concerned with and if anyone wanted something fixed, that they would have to do it themselves. (I think that's what pissed Wild and some others off). They didn't feel that they should beta test and report bugs when those bugs were deliberately being ignored.
To that end, can you set me up so that I can commit changes if Djole or another dev is not going to work on areas outside their own project? If we can get this implemented before the next beta release / test....we can get confirmation from those who actually fly using joysticks and finally put an end to this 5 year bug.
Thanks
Posted: Wed Apr 01, 2015 2:42 pm
by Djole88
No jbansk, I am ignoring you because I've already explained to you what the 'JoystickImage' does.
Mouse is now much slower (lower sens) in r8 than r7, I've had people complain. The value that you want changed influences only mouse and not the physical stick in any way.
I have only adopted the 1/1024 because the underlying image res also changed.
If you do somehow prove me wrong I'll be happy to implement it, till then...
Posted: Wed Apr 01, 2015 2:47 pm
by jbansk
....nevermind, you've reminded me of why I gave up trying so long ago.
best of luck ya'll
Posted: Fri Apr 03, 2015 3:53 pm
by Terran
jbansk wrote:QUOTE (jbansk @ Apr 1 2015, 10:47 AM) ....nevermind, you've reminded me of why I gave up trying so long ago.
best of luck ya'll
What's the matter? If djole doesn't understand the fix its only fair for him to ask for clarification. If this fix really does work let's work at getting pushed through, not giving up.
Posted: Fri Apr 03, 2015 4:49 pm
by Imago
Posted: Fri Apr 03, 2015 5:03 pm
by raumvogel
Posted: Thu Apr 09, 2015 7:16 pm
by pkk
New build, based on changeset 1162, on is on ACSS AU...
Posted: Thu Apr 09, 2015 8:23 pm
by raumvogel
I tried to launch allegiance.It said I needed to download 2 updates. It downloaded 1, then Avast blocked the other saying it had "Drep". Now the launcher stops working as soon as I allow the update.