4027 replies to this topic

[Gravy]

rev 363

 

 unstable in edit mode.

the layers disappear without any reason.

had to take the table out and reload it.

 

glowing dots appear on wireframe ramps.

it's normal ?  my environment is that of flupper1

 

the metallic material has a plastic aspect?  I didn't touch anything

it's due to the environment ?

 

This is likely related to a similar issue JPSalas reported https://www.vpforums...=44135&p=507269

[kiwi]

The flippers do not remain on the selected surface,

this can happen when I change the name of the object of the surface where the flippers are placed,

obviously then I select the object with the new name as the surface, but the flippers remain on the playfield.

If I change the name of the object before assigning it as a surface, the flipper will be placed at the right height.

 

However, when the problem happened to me for the first time I didn't do anything of what I said above,

I edited two flippers by changing only their length and position, something different happened here.

[studlygoorite]

One thing I do notice using 10.7.2 RC vs v354 is a much longer table loading time with RC, With RC 31 secs to load Twister and with v354 it's 10 secs.

[wiesshund]

One thing I do notice using 10.7.2 RC vs v354 is a much longer table loading time with RC, With RC 31 secs to load Twister and with v354 it's 10 secs.

 

that be weird, it is only taking 8 seconds here

[Outhere]

@wiesshund  --  How long does it take you to load Harry Potter, Takes me 23 seconds > whatever EXE I use

[studlygoorite]

And it wasn't just Twister, I noticed the delay with all tables 

[LoadedWeapon]

And it wasn't just Twister, I noticed the delay with all tables 

Make sure that you didn't grab the debug version I did that once lol

Twister with puppack and even Harry Potter only takes 5 sec to load but that is on a m.2

[Outhere]

 

And it wasn't just Twister, I noticed the delay with all tables 

Make sure that you didn't grab the debug version I did that once lol

Twister with puppack and even Harry Potter only takes 5 sec to load but that is on a m.2

 

WOW - I must have a serious bottom that thing going on somewhere

[LoadedWeapon]

And it wasn't just Twister, I noticed the delay with all tables

Make sure that you didn't grab the debug version I did that once lol
Twister with puppack and even Harry Potter only takes 5 sec to load but that is on a m.2

WOW - I must have a serious bottom that thing going on somewhere

Yea not sure why it would take so long...
https://youtube.com/...g?feature=share

Edited by LoadedWeapon, 10 October 2022 - 11:24 PM.

[wiesshund]

 

@wiesshund  --  How long does it take you to load Harry Potter, Takes me 23 seconds > whatever EXE I use

 

i dont have harry potter, that is a pup table, and i dont use pup

 

the tables i have that take the longest to load are Terry Red's FP tables, but i just click start on those and then watch lord of the rings, while they load

 

Only VPX tables i have ran into that loaded abysmally, had strangely oversized textures, like 14k playfields and 8k plastics and huge images for targets

which a quick pass through paintshop rectified in short order, no point in a texture that is 4 times larger than a person could even display

[toxie]

Coming back to overall 10.7.2RC testing: So we have some setups where some tables do not load (still waiting for error messages or crash dumps there, please) and the increased loading time (mystery :/ , are the loading times larger with the 10.7.2RC also when compared to 10.7.0 for you, studlygoorite?).

[studlygoorite]

Coming back to overall 10.7.2RC testing: So we have some setups where some tables do not load (still waiting for error messages or crash dumps there, please) and the increased loading time (mystery :/ , are the loading times larger with the 10.7.2RC also when compared to 10.7.0 for you, studlygoorite?).

Just threw in the original 10.7.0 (just the .exe) and my loading times are extended (same as the RC build). Funny I don't remember this issue when 10.7.0 first came out, maybe another file other than just the .exe has some changes that affect my loading times? (I am using all other files from the RC build) I threw in the .exe from v354 and my loading times are normal again. 

 

Just loaded up 10.8 (v368) and my loading times are normal

[naboodiver]

Just installed the RC.    Runs great on my machine (15 year old i5 with 32g memory).     The long load times and some tables aborting present in 10.7 release are now gone.    Great work guys.    

[lilalien]

Installed the recent RC. No problems to report on my end, speed of loading appears to be similar to an earlier beta (before RC state) I was running.. I'll report back if I notice a table loading very slow.

[RacerRiders]

"the tables i have that take the longest to load are Terry Red's FP tables, but i just click start on those and then watch lord of the rings, while they load"

 

LOL

[toxie]

@studlygoorite Maybe you mentioned this already, but what is your config (how many cores/threads does your CPU feature?)?

[studlygoorite]

@studlygoorite Maybe you mentioned this already, but what is your config (how many cores/threads does your CPU feature?)?

Using i7-10700 (8 cores 16 threads) RTX 3080 32GB RAM with Windows 10 Home (21H2)

[toxie]

Thanks! Could you please try https://github.com/v...runs/3116927980 and tell me what the loading times are with that one? I suspect that the c++20 change may have caused the faster loading times for you (maybe).

Oh, or its simply the new pre-processing for static elements that vbousquet included. It is now fully GPU accelerated, no more CPU involved like before.

But its too risky still to be included in 10.7.2, so will only be featured in 10.8.0.

[studlygoorite]

Thanks! Could you please try https://github.com/v...runs/3116927980 and tell me what the loading times are with that one? I suspect that the c++20 change may have caused the faster loading times for you (maybe).

Oh, or its simply the new pre-processing for static elements that vbousquet included. It is now fully GPU accelerated, no more CPU involved like before.

But its too risky still to be included in 10.7.2, so will only be featured in 10.8.0.

With the v289 (and just swapping out the .exe) that you supplied above my times are slow again, 20 secs to load a table direct from the VPinballX.exe and 30 secs to load a table from PinUP Popper frontend. Swapping out to 10.8 v368 or v354 and my times are good, 8 secs to load direct from the VPinballX.exe and 11 secs to load from Popper.

Edited by studlygoorite, 11 October 2022 - 11:02 PM.

[toxie]

Okay, thanks for testing! I think i know the causes then. Its that the new improvements by vbousquet are still missing from 10.7.2.

 

 

@Everybody, please test this new 10.7.2RC, done using a more 'official' build setup by jsm:

https://github.com/v...ase-win-x86.zip

 

Let me especially know if you experience tables not loading/crashing AND if it also cures the mystery startup hang bug (like Thalamus' build did) that some of you were experiencing!