1622 replies to this topic

[randr]

Well I will say the pindmd2 works better with the patch all around! Thanks mjr.
Call me stupid but I know have hyperspin, pbx, nailbusters popper and now pby all working and honestly why the hell not! Each have a few bugs so it’s nice to test each one and keep up with what’s changed. Now to catch up on table releases I’m 9 months behind in cab updates lol

[mjr]

I'd be VERY curious if this is also related to the problem that "a couple of people" had mentioned in relation to PuP-Packs not closing out when VPX is closed out. 

 

You know, you might be right.  At first glance I thought: no, that can't be it, it kind of has the opposite effect, of crashing the host process instead of leaving it running.  But in this case it could actually make sense - VP is the process that's crashing, which in turn leaves pinup hanging.  Pinup must be monitoring VP for exit in such a way that it doesn't recognize "VP just crashed" the same as "VP just exited".  So yeah, someone who's been having the exit-hang problem should try out the patch and see if it fixes it without the need for the TASKKILL's.

 

(For what it's worth, pinup should probably *really* fix it by monitoring VP's process handle instead of whatever approach it's taking now.  Its current approach clearly misses VP's exit under some conditions.  The proper approach is this: get a handle to the process with OpenProcess(), and then use WaitForSingleObject()  on the handle to monitor for exit.  That should be 100% reliable.  That would be better than depending on VP to exit gracefully in general, since this particular DLL issue is only one of many ways that VP could crash.)

 

You wouldn't mind making this patch for 1.7.1 too ? This one has the same problem as freezy' 1.7.2 - no color for vni/pal altcolor. I can see that it is probably a perfect fix for Randr though.

 

About forcing messages :

How do you turn off swf warning in PBY ? I don't have flash installed. PBX seems to have that integrated.

 

Same question almost for DOF :

If I choose to enable dof while browsing, there is a warning message. It used to work flawlessly in your earlier.

 

I understand that the error from missing flash is because of my swf resources in PBX. What if I don't care about them for the moment. They are hidden normally when I run PBX (instructions). I could probably just rename the PBX folder. But, that would not really make PBY compatible with PBX would it ?

 

1.7.1 - yeah, I could add it there.  You'd want it on my branch with my earlier 1.7.1 patches as well, rather than the original freezy 1.7.1, right?

 

SWF - I'll take a look.  I agree, you shouldn't have to put up with a warning over and over if you just don't want to install Flash.  I'll look into adding a "don't show this again" option or something like htat.

 

What's the DOF warning message you're seeing, and what's triggering it?  I can put it in some kind of "don't show this again" there too.

Edited by mjr, 10 March 2019 - 08:19 PM.

[Thalamus]

Yes please add to your 1.7.1 - that is what I'm currently using and for me it seems to be the best alternative out there.

 

Sounds great for swf.

 

Dof. All my mistake. I've changed from the 32bit to 64bit last time around and the warning makes sense. It is triggered when I ask for dof effect to be on. But, yes, still not running that version. Sorry for the confusion. Will have a go at it soon.

[mjr]

Yes please add to your 1.7.1 - that is what I'm currently using and for me it seems to be the best alternative out there.

 

Okay, just posted a new cumulative 1.7.1 patch that includes this fix.  Give it a try and let me know if it works for you.

 

https://github.com/m...7.1+mjr20190310

[Thalamus]

Thank you so much., highly appreciated. I'll give it some good run tomorrow. We live in quite different timezones ;-)

[Outhere]

Does this version also have the problem?

https://github.com/f...nsions/pull/152

[mjr]

Does this version also have the problem?

https://github.com/f...nsions/pull/152

 

All versions other than my patched versions have the problem, as far as I can see - it's an oversight in the core PinDMD code (for all device versions) that's been there from the start.

 

I just sent freezy a pull request with the fix on the mainline, so hopefully it'll get into all of the newer versions eventually if it looks right to him.

Edited by mjr, 10 March 2019 - 10:34 PM.

[Thalamus]

MJR. I just had to give this a quick run before heading to bed. All tests I did looks to be working great. Just knowing that you've found something that is fixed makes me excited because I've had my suspicion that my dmddevice has caused issues at times. I just hope that I now can trust them to be gone.

 

Too early to conclude of course. But, never the less. A HUGE thank you !!

[dannygey]

Hello everyone,

I have a strange problem with pinbally that suddenly came up.
When i start pinbally and push a flipperbutton ( or shift key ) to go to the next game in the list it start runnning verry fast trough the games. It's almost like the button is always pressed or hangs but this is not the case in a game the buttons work fine.
I’m not sure but i suspect the problem started when i connected a wireless keyboard.
The wireless keyboard is now disconnected but the problem stays.
Does someone have any idea what the problem is ?

[mjr]

I have a strange problem with pinbally that suddenly came up.
When i start pinbally and push a flipperbutton ( or shift key ) to go to the next game in the list it start runnning verry fast trough the games. It's almost like the button is always pressed or hangs but this is not the case in a game the buttons work fine.
I’m not sure but i suspect the problem started when i connected a wireless keyboard.
The wireless keyboard is now disconnected but the problem stays.
Does someone have any idea what the problem is ?

 

The super-fast scrolling mode is triggered when the next/previous keys are auto-repeating, which happens when you hold them down for a few moments.  So it sounds like your keyboard is going into auto-repeat mode very quickly.

 

The auto-repeat delay is configurable at the Windows control panel level - go to Control Panels > Keyboard > Speed and use the Repeat Delay slider to adjust the delay.

 

Your button encoder might have an auto-repeat setting of its own, depending on what you're using.

 

You might also look to see if that wireless keyboard you installed and removed left behind any extra software that might have yet another bunch of settings of its own.  Some keyboards and mice have their own customization software that will mess with the system settings as a "convenience", so that might be why the change coincided with installing that new keyboard.

[NailBuster]

You know, you might be right.  At first glance I thought: no, that can't be it, it kind of has the opposite effect, of crashing the host process instead of leaving it running.  But in this case it could actually make sense - VP is the process that's crashing, which in turn leaves pinup hanging.  Pinup must be monitoring VP for exit in such a way that it doesn't recognize "VP just crashed" the same as "VP just exited".  So yeah, someone who's been having the exit-hang problem should try out the patch and see if it fixes it without the need for the TASKKILL's.

 

(For what it's worth, pinup should probably *really* fix it by monitoring VP's process handle instead of whatever approach it's taking now.  Its current approach clearly misses VP's exit under some conditions.  The proper approach is this: get a handle to the process with OpenProcess(), and then use WaitForSingleObject()  on the handle to monitor for exit.  That should be 100% reliable.  That would be better than depending on VP to exit gracefully in general, since this particular DLL issue is only one of many ways that VP could crash.)

 

 

 

FYI.  a little background to help you and your users with PinUP Player.  PuP is a Com Server and there are com-clients/activex that can come from freezy/pup,  b2s pup plugin,  table script, doflinx, even Popper.  It can be from one or many, and it works fine as long as all activex are 'cleanly' closed by the different systems.

 

Suggesting taskkill is a bad idea on the PuP Com server parts...  it is not required.  If pby users have 'stuck' PuP-Packs its the result of a 'dirty' close, and should be fixed.  In fact, its a 'bonus' that you see the PuP-Pack still running as its a visual cue that your table is causing a 'dirty' close on your system and should be fixed.  

 

The simple troubleshoot,  is to have them run the table via vpx editor directly/launch game/and exit editor completely.  If all components close then its a 'clean' close and they need to figure out why pby isn't closing things 'cleanly'.   Sometimes its a timing thing, perhaps its people taskkilling ultradmd and causing issues... Terry and I have spent countless hours helping people cleanly close out tables/systems so that their vpin/PinUP runs stable.  There should never be a need to taskkill vpin components.

 

Best to send PinUP Player issues to this page:  http://www.nailbuste...ck_troubleshoot

Edited by NailBuster, 12 March 2019 - 02:40 PM.

[mjr]

FYI.  a little background to help you and your users with PinUP Player.  PuP is a Com Server and there are com-clients/activex that can come from freezy/pup,  b2s pup plugin,  table script, doflinx, even Popper.  It can be from one or many, and it works fine as long as all activex are 'cleanly' closed by the different systems.

 

Yes,I understand that you're a COM server and that creates some more difficulty on your end, but you still might want to look into more reliable shutdown.

 

In this particular case, the reason your users are forced to use TASKKILL was that dmd-ext was crashing your VP client process on you.  If you're going to work as a COM server, it's best for your users if you can be robust against that sort of thing.  Clearly it's preferable if the client process exits cleanly on its own, but the problem is that's not under your control.

 

Hopefully now that we've found the dmd-ext issue, VP won't crash as often.  If it were me, though, I'd still want my end of the software to be robust against faults in the client.

 

 

The simple troubleshoot,  is to have them run the table via vpx editor directly/launch game/and exit editor completely. 

 

If only Windows were so simple!  If you look at this case in detail, you'll see why that advice was misleading.  We were dealing with a sporadic VP crash (not VP's fault, but in the VP process) that had some timing dependencies.  Running under different conditions changed the timing and changed the behavior.

[dannygey]

Hello mjr,

 

I Have found the problem.

Sticky keys was activated in 'Ease of access center'

I turned on sticky keys on another computer and same thing happens.

 

Danny

[mjr]

I Have found the problem.

Sticky keys was activated in 'Ease of access center'

I turned on sticky keys on another computer and same thing happens.

 

Oh, that makes sense!  I actually have a note about that in the Windows Setup section of the build guide, since it's one of those weird little glitches that's hard to track down if you're not thinking about it - which I wasn't this time, as it didn't occur to me until you mentioned!

[dannygey]

I Have found the problem.
Sticky keys was activated in 'Ease of access center'
I turned on sticky keys on another computer and same thing happens.

 
Oh, that makes sense!  I actually have a note about that in the Windows Setup section of the build guide, since it's one of those weird little glitches that's hard to track down if you're not thinking about it - which I wasn't this time, as it didn't occur to me until you mentioned!

Cabinet is completly finished now !
I have used your software pinbally and the pinscape board with expansion board and everything works perfectly.
Thanks for the help on this build and keep up the good work.

Regards Danny

[DreamTrap]

when i run this both normal and admin mode it says it can't find directx in win7 64x but i have the last version of dx11 installed they made directx_Jun2010_redist?

Edited by DreamTrap, 16 March 2019 - 12:57 PM.

[Thalamus]

VP is actually DX9. Did you run the one-in-all installer ?

[DreamTrap]

so this won't work in win7 because that installs dx11 by default. o well

Edited by DreamTrap, 16 March 2019 - 02:42 PM.

[Thalamus]

Sure. I run Win7.

[DreamTrap]

how did you get it to run than?

Edited by DreamTrap, 16 March 2019 - 03:22 PM.