1353 replies to this topic

[boiydiego]

is there work beeing done on the full color dmd for lcd screens ?

[ronaldvg]

ronaldvg, this is probably the ddraw=0 problem I mention above... it's coverered in depth earlier in the thread.

That works. Stupid thing is that I knew about this setting but was not needed in 2820, so I did not realise that it would be necessary now. I do not use the DMD at all because I have a RGB LED DMD. Now VPX does not crash anymore but the B2S backglasses don't show ;-) Not the biggest problem but nonetheless strange.

 

Edit: when I change back to 2820, the B2S backglasses still do not show, and also in PinballX they do not show. Is that normal ?

 

Edit 2: So the B2S backglasses do not work anymore, period. Tried to register the B2S server again but it's a no go. For now I just changed PinballX to not use B2S and not hide the backglass. Works but not ideal of course.

 

Edit 3: When I tried a new table, the B2S worked. That made me think about which files get changed during the start of the first time that a table starts and that led to the solution. I just deleted every table entry in the B2STableSettings.xml and after that everything works again.

Edited by ronaldvg, 29 October 2016 - 12:48 PM.

[MrKnister]

hey guys, a stupid question which may be covered somewhere, so forgive my ignorance .

 

whenever i start a table in exclusive fullscreen mode with the latest vpx beta, i see that the fps get locked

to 59,9 fps instead of 60 fps which i defined in the video settings within vpx under fps limiter / vsync with 

a "1". also, the "1" settings seems to not work for me in the fullscreen mode. to get the tables locked to

60fps ill have to change the "1" to "60" but this gives me only the mentioned 59,9 frames.

 

since my tables only run really smooth with the vsync option i wonder what im missing here.

 

thx in advance for any tips

 

besides this little issue the latest changes and progress made within vpx and mame are really awsome !!

you guys are machines !!! highly appreciated !!!

Edited by MrKnister, 30 October 2016 - 01:48 PM.

[toxie]

as the fps are measured from system time and then truncated this could be simply a floating point precision issue, so this can be considered normal. 

[shadowshd]

Hi,

 

First of all thanks for your hard work 

 

Excuse me if it has already been answered, but I noticed an issue with 10.2 core.vbs I have on some VP9 /PM5 tables as well as with VPX 10.1 tables (like Cactus Canyon).

 

Very often, when the ball enters a hole where it should be ejected, it won't be ejected.

 

Sorry if I don't use very technical words to explain clearly so I'll give some examples :

- In Cactus Canyon (VPX version), when the ball enters the gold mine or the saloon it stays here and isn't ejected. The game doesn't enter in a "seeking missing ball" sequence.

- In Popeye (VP9.9.2 version), the ball often gets stuck in the skill shot spinner and the game won't enter the seeking missing ball sequence.

 

The trick I use with tables that don't need VPX 10.2 is to rename core.vbs to core102.vbs and put back 10.1 core.vbs in my script folder.

 

Thanks again

 

++

[MrKnister]

Thx for the reply toxie . But ill have to say, that i feel some slight unfluid Ball flow when in exclusive full Screen with the 59,9 frames compared to the locked 60s in windowed Mode. So i "fear" your implemented meassuring methods are exact . A way to Lock to exact 60 via some Kind of Trick would be nice to try

[Umpa]

Thx for the reply toxie . But ill have to say, that i feel some slight unfluid Ball flow when in exclusive full Screen with the 59,9 frames compared to the locked 60s in windowed Mode. So i "fear" your implemented meassuring methods are exact . A way to Lock to exact 60 via some Kind of Trick would be nice to try

Not sure if you have multiple monitors but I read somewhere make sure they are all at 60hz

[MrKnister]

Thx for the hint indeed i have a two Screen setup with a real dmd in the Cab. And yeah too Bad (at least for me ) i Set the refresh rate for both screens in Windows to Fixed 60 Hz. While using windowed Mode it gets recognized proper grml

Edited by MrKnister, 30 October 2016 - 08:51 PM.

[toxie]

but vsync enabled will always wait for vsync, no matter what. so no matter what the counter says, it will be vsync'ed (e.g. matching your monitors refresh rate).

[sliderpoint]

I'm having a couple issues with lights hopefully you guys can take a look.  

 

First the Duration method doesn't seem to be working correctly with setting a blinking state.  using code from the commandreference.txt:

Flasher1.Duration 2, 1000, 0

 

I set the light (flasher1) blink pattern to 10 and blink interval to 80.  When I set the light state to just be 2 it blinks normally.  If I have it set to Off and then use the duration method, the light will have very erratic behavior.  Sometimes just on solid or is pulsing crazy fast (almost looks like it's on solid). Sometimes it might blink once.

 

Second is with a LightSequence.  They work correctly, but I'm noticing that the time(update interval) behaves differently on different machines. They are slower sweeps on my laptop vs same table on my cab. Any idea what could cause these to run at different speeds on different computers?

 

Thanks.

-Mike

[DJRobX]

If I might trouble you for another small change for true fullscreen.   On my cab the mouse cursor does not go away.   The old "Trick" of moving the mouse off-screen doesn't seem to work in exclusive fullscreen.  :/

 

Around the same spot as the new LockSetForeground command, if we make it:

 

if (m_fFullScreen)

   {

  while (ShowCursor(false) >= 0);

  // blocks processes from taking focus away from our exclusive fullscreen app

  ::LockSetForegroundWindow(LSFW_LOCK);

 

   }

 

 the cursor shuts off correctly.   Of course, we probably then need to also turn it back on (probably in WM_DESTROY?).

[toxie]

phew.. plus its kinda risky to have that potentially loop forever.. but one could have a break if x iterations were run?

 

but i never understood why the showcursor(false) was not done all the time in the first place? maybe because someone was too lazy to en/disable it in all the respective places?

[toxie]

Excuse me if it has already been answered, but I noticed an issue with 10.2 core.vbs I have on some VP9 /PM5 tables as well as with VPX 10.1 tables (like Cactus Canyon).

 

Very often, when the ball enters a hole where it should be ejected, it won't be ejected.

 

Sorry if I don't use very technical words to explain clearly so I'll give some examples :

- In Cactus Canyon (VPX version), when the ball enters the gold mine or the saloon it stays here and isn't ejected. The game doesn't enter in a "seeking missing ball" sequence.

- In Popeye (VP9.9.2 version), the ball often gets stuck in the skill shot spinner and the game won't enter the seeking missing ball sequence.

 

The trick I use with tables that don't need VPX 10.2 is to rename core.vbs to core102.vbs and put back 10.1 core.vbs in my script folder.

 

 

@DJRobX: Could this be a sideeffect of your solenoid changes? Cause i don't see anything else that could've caused this.

[DJRobX]

 

 

 

@DJRobX: Could this be a sideeffect of your solenoid changes? Cause i don't see anything else that could've caused this.

 

 

I reviewed a diff of the core.vbs changes when I saw that post.  If coming from VP9 and UseVPMModSol = 0, there should be absolutely no difference, the code paths are the same.

 

If VP10 then you get the changed timer interval.    If VPMModsol is true, then of course you open a bunch of possibilities, but that's why ModSol is an option.   There are other changes but they don't affect solenoids or switches so I don't see why they would cause this. 

 

I will test Cactus Canyon later to see if I can replicate.

 

 

 

phew.. plus its kinda risky to have that potentially loop forever.. but one could have a break if x iterations were run?

 

but i never understood why the showcursor(false) was not done all the time in the first place? maybe because someone was too lazy to en/disable it in all the respective places?

 

 

That while loop is what I found when googling that subject.   It's apparently a poorly designed Microsoft API that works like a stack.    Unless Microsoft breaks their own API (possible!) it shouldn't cause a problem.  If you want to code it more defensively I certainly wouldn't argue with you one bit.     I suspect people like the ability to move the mouse up to the backglass to alter its settings and the VPM window too.    But coding it to only do that in true FS mode is a good compromise.   It just really sucks to see a mouse pointer smack dab in the middle and have to grab the mouse to move it out of the way.   

Edited by DJRobX, 31 October 2016 - 10:24 PM.

[DJRobX]

I checked both of those tables (Cactus Canyon VPX 1.1 by ninuzzu and Popeye VP9) and I am unable to replicate the problem.  

 

What I find most interesting in his report is that he says neither table goes into ball search mode.  

 

Shadowshd, 

 

When this happens, are the flippers still functional? 

Edited by DJRobX, 31 October 2016 - 11:16 PM.

[Sir Cheddar]

I've seen this behavior - Gold mine doesn't eject ball - on my system a while ago but didn't look further into it. But I've got it on CCC, not CC.

Didn't get any ball search either but I can't remember if I could flip or not - I guess I could or I never would have waited for an ball search to occur.

 

I played a few games on CC and CCC today but couldn't reproduce it anymore.

Pretty confusing error if it prevents ball search on both VPM and P-ROC...

Edited by Sir Cheddar, 01 November 2016 - 10:57 AM.

[wrd1972]

Forgive me if this has already been requested.

 

The visual physics of gates is driving me nuts on my latest table. On my BOP table, the two upper plate gates bounce like a ball when they close. Its as if they are counterbalanced and they float like they are in the wind. As I am sure we all know. Gates, should drop like a rock with very little rebound.

 

Are there any plans to fix this gate closing issue? Can a mass setting be applied to the gates?

[shadowshd]

Hi DJRobX,

 

Thanks for your help.

 

When I update VPX 10.2 beta on my pincab, I always update VBS scripts (copy and replace) in my Scripts folder as well.

 

If I use the core.vbs from VPX 10.2, with Cactus Canyon Continued P-ROC when the ball enters the gold mine or the saloon it won't be ejected like it should.

The game just gets stuck here, flippers are still functional, no error and no crash, and no ball search mode.

 

I've just played CCC now. I launched the ball two times in the saloon and it was ejected, but at the third time it wasn't.

 

I've just tried also with the VPM version of Cactus Canyon and haven't have the issue.

 

A few days ago I've played Popeye and when I got the multiball mode, two balls got stuck "under" the spinning disk and when I lost the third ball the game just got stuck here, with no error no crash and no ball search mode.

 

A member of a french pincab forum I know has the issue too, at least with Cactus Canyon Continued.

I thought it was a P-ROC issue but what is weird is that I encountered it with Popeye.

 

I know sometimes with Popeye that the ball can get stuck somewhere, but then the game enters the ball search mode.

Not with core.vbs 10.2.

 

So the "fix" (if we can call this a fix) for now is to put back the 10.1 core.vbs, thanks to Arngrim's advice.

 

I have noticed this issue on last september 15 - if I don't say silly things 10.2 rev was 2791 at this time.

 

Thanks

Edited by shadowshd, 01 November 2016 - 11:07 AM.

[fuzzel]

rev2829 is up:

 

- apply changes done in the material debugger to the table after user quits the player
- workaround precision issues with FPS counter and add ambient occlusion visualization for F11 mode
 

[DJRobX]

I have no idea how the CCC/PROC stuff works.  I've played a real CCC before; it'd be fun to have that on my cab, though, so I'll try setting it up.   Hopefully I get the same problem so I can fix it!