Jump to content
LaunchBox Community Forums

megashub

Members
  • Posts

    68
  • Joined

  • Last visited

Posts posted by megashub

  1. This is a great plugin. Thank you for making it. Are you accepting payment for additional features, by any chance?

    Desired enhancements:

    1. Exact Pattern match search of existing Additional Apps
    2. Bulk Find/Replace (exact pattern match only).

    Example:

    • Application Command-Line Parameters of an Additional App contains a typo: -hiddne -exit -play
    • Perform a Find/Replace using pattern: dne
    • Plugin finds all Additional Apps that contain that exact pattern and presents a multi-select list for the user to select which games to modify.
    • User provides the new string to replace: den
    • Plugin makes the replacement on all selected games

    By restricting this new feature down only to exact matching, I'm hoping the complexity is much lower (no fuzzy searching, no wildcards, etc etc). Thoughts?

  2. 17 minutes ago, JohnnyK said:

    I am using 11.10 and running the CriticalZone v2 BlueBox Theme.  If I remember right I picked it because it was one of the most popular themes so I thought it would be the most ironed out.

    I have not updated to 11.11 ... I didnt see anything in there regarding any fixes to video playback.  And since everything else runs fine I was going to sit it out until others posted positive results.  Last thing I want is more shit to go sideways.

    As for which videos are causing the issue.. I have no idea.   I have over 600 games in there and I am never staring at the screen when it happens.  I did think about setting up a video camera or something, but thats kind of a pain in the ass.  Not to mention I dont know how I would do that exactly.

    I don't know if this issue is specifically what was addressed, but there was this note in the 11.11 changelog, so it may be worth evaluating it

    Quote

    Fixed: Rare error related to Big Box videos (introduced in 11.10)

     

    • Like 1
  3. Calling all 360 stick users! Please go vote on this enhancement request, originally opened in 2017 and still pending in the backlog. Requests with the most momentum behind them will generally get higher priority, especially if the level of effort required to do the work is low, which I believe this one will be.

    Basically, it's @James35 's %romname% variable concept above + the use of this plugin to bulk update your rom collection within LB.

  4. 1 hour ago, megashub said:

    If we can identify that it's definitely the theme that is the issue

    Looking pretty likely that this is a Theme issue. I'm 1 hour in using the Default theme, with the exact same elements in use, and no problems at all. I was seeing issues as early as 30 minutes when using AllNightLong. 

    Update: confirmed. 2 hours in now, and absolutely no performance problems.

  5. 3 minutes ago, Jason Carr said:

    I will take a look at the theme to see if I can figure it out. If we can identify that it's definitely the theme that is the issue, then you can let me know what view(s) and I'll take a look. Did you download the theme from inside of the Big Box Themes Manager? Or did you get it from the forums?

    It was one of the 21 offered for download in the BB Themes Manager, so I grabbed it that way. It's AllNightLong. I'll go find the view after this test is over, but it's a vertical wheel of clear logos with game details and gameplay video. I was using it with an auto-populated Favorites playlist, if that matters. 

  6. 1 hour ago, megashub said:

    Okay. Moved to new test: Switched to a vertical wheel (text titles only), but with game details (static art and gameplay video). Stay tuned.

    This test has been running for nearly 90 minutes. No problems at all. I'm going to try to find a view that uses both Clear Logos and videos in the default theme next.  

    Edit: Found one. Starting a new test. If this one comes back clean there's no difference functionally between this view and the one that causes the machine to CTD, other than the Theme in use. 

    If it does turn out to be Theme-specific, what's a way Theme makers can find the offending code?

  7. 1 minute ago, Jason Carr said:

    Yes, definitely. That's good news though. At least we're starting to narrow it down. I would test one of the vertical wheel views next. A wheel view with a video in there somewhere as well.

    Okay. Moved to new test: Switched to a vertical wheel (text titles only), but with game details (static art and gameplay video). Stay tuned.

  8. 29 minutes ago, megashub said:

    I have the default theme on a test run now. The default view (or at least the view it gave me after I switched back to it), is just a horizontal wheel of static cover art that it spins (no video or text beyond the title in the top right corner). That's been running for 15 minutes now, with no visible blips or lags developing yet. For comparison, the other theme kicks off the gameplay video for each title when the spin stops. Perhaps that's an important factor? Will keep monitoring it.  

    45 minutes in now and not a single blip during the spin on this test. That's a pretty stark contrast from the previous tests. Only thing I changed was the theme, which shifted the view to this covers-only wheel. The test is still running and I'll keep monitoring, but I think the next test should be a view using the Default theme that includes the videos, if one exists. Yes?

  9. 1 minute ago, Jason Carr said:

    These issues are usually solved here on the forums. It's still possible that your issue @megashub has nothing to do with the Nuc at all, so I would recommend the following to troubleshoot:

    • Try closing all background apps
    • Try disabling any third-party anti-virus or security software
    • Try updating any and all drivers (especially video drivers)
    • Make sure you're up to date on Windows updates (but don't jump to Microsoft's beta releases)

    The only background apps I have running to my knowledge are Steam and Epic.

    I have no AV/Security software installed. It's just Stock Win10 x64.

    Drivers: done. All drivers are current version (it's a very new installation).

    Updates: Same. Current stable release.

     

    I have the default theme on a test run now. The default view (or at least the view it gave me after I switched back to it), is just a horizontal wheel of static cover art that it spins (no video or text beyond the title in the top right corner). That's been running for 15 minutes now, with no visible blips or lags developing yet. For comparison, the other theme kicks off the gameplay video for each title when the spin stops. Perhaps that's an important factor? Will keep monitoring it.  

    • Like 1
  10. 2 minutes ago, Jason Carr said:

    Yes, they do look to be high performance machines.

    Yes, it is likely to be a software problem, but we haven't been able to narrow it down. Attract Mode is most definitely fully unloading the media and letting go of those resources, yes. We've put massive efforts into performance optimization and testing there over the years, and I regularly run multi-week long tests on various machines with zero issues for weeks on end. That's why this thread has thrown me for a loop so much. There seems to be something on these Nuc systems that is changing things up, and I don't know what that is. I also don't know if all Nuc systems are affected, or if it just happens to be an issue with some configuration on these systems.

    Ahh got it! That's good to know, and now I totally understand why this is such an odd-man-out issue. Happy to help give you as much useful testing data as I can. :)

  11. 1 hour ago, Jason Carr said:

    It is possible that Big Box is pushing the CPUs in the Nucs to their limits (at least on one particular core), and causing heat issues, which slows down the CPUs. Perhaps that is causing it to lag behind? I have figured out that if attract mode lags behind too much, Big Box will eventually crash with a stack overflow, so that could be what's going on here.

    I'm no expert on the Nucs, but my guess is that they're running mobile/laptop CPUs? I know that usually those CPUs have settings in the BIOS to control performance, but I'm not sure if that's only on laptops. I do know, however, that even on brand new Intel laptops, if I don't adjust the performance settings, I end up with terrible performance out of the box (at least when running on battery power). No idea if that's relevant to the Nucs, but I thought I'd mention it.

    image.thumb.png.01409f7ab987163e3ce8ebdf57de59ea.png

    Here's a detailed breakdown. These are high performance machines geared specifically toward AAA-capable gaming and are configured as such OOTB (at least in my case running an 8th gen Hades Canyon NUC): https://blog.itcreations.com/intel-hades-canyon-nuc-8-review/

    I think this is why it's a software-side problem. Is Attract Mode fully unloading the media for a given game and letting go of those resources before loading up the next one? Because if it's holding on to cache or isn't fully killing processes every time, it's possible it's just grinding to a halt over time. But I can't speculate well because I don't understand how Attract Mode works under the hood. 

  12. 2 minutes ago, Jason Carr said:

    @megashub Thanks for the details. It is interesting that you and @MrSco are both running Nucs (even if the generations are different) and are both running into issues (though I'm not sure how similar they are). I haven't tested on a Nuc, but I have tested with integrated Intel graphics on my laptop. But it sounds like it might not be related to the graphics, since @megashub is using Radeon graphics, so I don't know.

    How much RAM do you have @megashub? Can you try running the Big Box default theme for a bit to see if it happens to be a theme issue?

    Ram: 32GB DDR4, default theme, sure. I'll run a fresh test with default theme and report back.

    • Like 1
  13. 2nd attempt to capture details about the progression of this problem:

    12:55 start

    progressed exactly the same as first attempt

    2:01 was the last time the clock updated.  

    2:20 all video and images stopped displaying almost entirely. Every third or fourth game, the video appears and plays, sometimes only for a second or two. Only the text title information is reliable and keeps changing with every spin.

    2:47 CTD

     

     

  14. 2 hours ago, Jason Carr said:

    That's not something I would worry about personally, but I'm not an expert enough on that subject to guarantee that it won't cause issues with any emulators. I do know it won't be an issue with LaunchBox or Big Box.

    • 10:19 start
    • 10:23 slight jitters noticed toward the end of the spin
    • 10:33 jitters still slightly noticeable, a few more, but still very slight. Only apparent on the longer-duration spins where max spin speed is achieved. 
      • Noticing on the long-duration spins that when the wheel is spinning at its fastest, only a small subset of logos are being rotated and displayed. Though the game selection itself does still seem to involve the whole game list. At the beginning when Attract Mode first started, the high speed logo spin appeared to include a larger quantity of logos. 
    • 11:07 during the spin, logos are starting to disappear for a second or so and then reappear and the spin has begun to get disjointed. It’s like the visualization of the spin is starting to lag.
      • Seems to happen less severely or even occasionally not at all on short-duration/slower spins.
      • I noticed throughout the test that the audio of the video that was just playing stays playing during the spin to the next title. Is that by design? I think the previous video remains up and playing in the background during the spin.
    • 11:13 slowly getting worse. It’s not just the wheel that lags, but the text titles changing lag at the same points as the wheel. 
      • Despite the now-obvious lag while spinning, the video playbacks themselves remain smooth and lag-free.
    • 11:16 the previously selected game is starting to flash back up on the screen for a second or two intermittently while the wheel is spinning to the next game.
    • 11:17 Sometimes it takes so long for the lag to catch up that the spin kicks off again before it settles on a game. This may be the death spiral that ultimately crashes the app.
    • 11:20 LB’s clock is no longer staying current with the actual time. Stuck at 11:17. The wheel while spinning is almost not visible at all. When it does stop on a game, the wheel visually displays in chunks before getting spun again a second later.
    • 11:21 There’s almost no graphical content left on the screen the majority of the time. The lag is cumulatively so bad now that it’s just the theme background and the text elements that stay on screen. The images and video content flash on and off, and mostly stay off.
    • 11:27 Interrupted Attract Mode to check Event viewer. LB:BB’s clock has refreshed to current time. While Attact Mode was not active, LB:BB Process was consuming 18% of CPU, 582MB of memory. Disk and network are both at 0%. Nothing else running on the machine comes anywhere close. When Attract Mode began again, it’s performance was still badly degraded. It did not clear up as a result of being interrupted.
      • Maybe none of the video players are actually being unloaded all the way, and they gradually build up until the system runs out of resources? 
    • 11:36 LB:BB has stalled and is entirely unresponsive. Awaiting CTD. 
    • 12:23 Still hung and not responding. Hasn't CTD'd yet. Probably would have if I hadn't interrupted it to check the resource consumption.
  15. 6 minutes ago, MrSco said:

    Ya, your NUC is a beast with the radeon graphics! bigbox should be child's play running off m.2 ssd drives.

    That was definitely an intentional decision. The last thing I wanted to worry about on this build was hardware horsepower. I'm sure it would struggle to play the topmost tier of games, but those aren't in scope for this build anyway. I was pleasantly surprised it played Tekken 7 and KI flawlessly.  Nah, the problems I've run into have all been, predictably, software quality-related (low quality rom dumps, pain-in-the-ass emulator configs, controller configs, etc). So chasing this Attract Mode annoyance is minor in comparison. Still, if there's a memory leak at play here (which is my current belief), if I can help the dev find and fix it, great. :)

  16. 2 minutes ago, MrSco said:

    megashub, you running bigbox from external drive? Or on those m.2 ssd?

    Using those m.2 ssds via the Internal m.2 slots. 

    2 minutes ago, MrSco said:

    your NUC is 2gens newer then mine so should be more then capable of playing even some of the latest AAA games 

    Yep, and it is. This is a multicade build with some Windows/Steam games, console emulators, classic mame, v pinball (fx3, etc), and sinden lightguns (whenever they arrive), and front ports for USB controllers. I can play Tekken 7 via Steam, the Killer Instinct steam version, etc all just fine thanks to the integrated Radeon RX Vega M GL GPU.

  17. 5 minutes ago, Jason Carr said:

    That's not something I would worry about personally, but I'm not an expert enough on that subject to guarantee that it won't cause issues with any emulators. I do know it won't be an issue with LaunchBox or Big Box.

    Alright. I unplugged everything but the USB keyboard and the wireless mouse dongle (which have always been plugged in from the start), then powered it back up. I haven't touched it, and will watch Attract Mode to see how it goes. If it still misbehaves, I'll reboot, then also shutdown any/all background apps (though there aren't many... just GPU control center, etc). The first few scrolls kicked off as I was typing this, and they have been smooth (not jittery). Let's see if it gets worse the longer it stays in Attract Mode. I'll also time the durations.

  18. 7 minutes ago, Jason Carr said:

    Okay, it could be interference from a background app, or it could be some kind of stray controller as well. Unfortunately it's likely to take some troubleshooting work to get it figured out. I would start by disconnecting any controllers and closing any background apps to see if either of those happen to be the issue.

    No problem. I'm pretty patient, especially since I'm relatively new to the world of arcade building and frontends (though not at all new to the technology involved). Will I need to worry about Windows re-addressing my controllers when I plug them back in again later? 

  19. 2 minutes ago, Jason Carr said:

    Did you happen to change themes recently? What theme are you running there? Have you tried rebooting?

    Not recently, no. That's an auto-populated Favorites playlist. Theme: AllNightLong (not customized). I'm still building my cab, so the machine is only powered up when I'm working on configuration or playing games. It's fully shutdown the rest of the time (not asleep). 

  20. 41 minutes ago, Jason Carr said:

    @megashub Were you having this issue with previous versions? What version did you have installed before updating to 11.9? I'm running some stability testing on 11.9 to make sure that we didn't introduce any new stability issues.

    That error is a stack overflow error, which could be a programming error, or it could be that the hardware couldn't keep up. Could you give me details on your hardware and your Windows version?

    I'm a recently new LB user, so I was running the stable version immediately preceding 11.9. I never let Attract Mode run for a prolonged period of time in the previous version, so I'm afraid I don't know if the problem existed then. What I can say is that the wheel spins erratically now, where before it used to spin smoothly.

    Here's my hardware: https://ark.intel.com/content/www/us/en/ark/products/126141/intel-nuc-kit-nuc8i7hnk.html

    Ram: 32GB DDR4 

    HDD: 2x 500GB Samsung 970 m.2 SSDs

  21. Adding my event log for analysis. Purpose-built for a cab, Win10 x64. No other apps or usage. Just LaunchBox+BigBox, MAME and Retroarch. I haven't installed any other video players, etc. Video drivers were already up to date from the manufacturer. Crashes to desktop within an hour or so.

     

    Log Name:      Application
    Source:        Application Error
    Date:          1/30/2021 9:33:57 PM
    Event ID:      1000
    Task Category: (100)
    Level:         Error
    Keywords:      Classic
    User:          N/A
    Computer:      ARCADE
    Description:
    Faulting application name: BigBox.exe, version: 11.9.0.0, time stamp: 0x5fb3100b
    Faulting module name: coreclr.dll, version: 4.700.20.26901, time stamp: 0x5ec48a29
    Exception code: 0xc00000fd
    Fault offset: 0x000000000004a4c7
    Faulting process id: 0x237c
    Faulting application start time: 0x01d6f7864632d69d
    Faulting application path: C:\Users\LAUNCHBOX\LaunchBox\Core\BigBox.exe
    Faulting module path: C:\Program Files\dotnet\shared\Microsoft.NETCore.App\3.1.6\coreclr.dll
    Report Id: 5f663230-da5f-4cd8-871d-d3e20a2d77b9
    Faulting package full name: 
    Faulting package-relative application ID: 
    Event Xml:
    <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
      <System>
        <Provider Name="Application Error" />
        <EventID Qualifiers="0">1000</EventID>
        <Version>0</Version>
        <Level>2</Level>
        <Task>100</Task>
        <Opcode>0</Opcode>
        <Keywords>0x80000000000000</Keywords>
        <TimeCreated SystemTime="2021-01-31T05:33:57.1474479Z" />
        <EventRecordID>3505</EventRecordID>
        <Correlation />
        <Execution ProcessID="0" ThreadID="0" />
        <Channel>Application</Channel>
        <Computer>ARCADE</Computer>
        <Security />
      </System>
      <EventData>
        <Data>BigBox.exe</Data>
        <Data>11.9.0.0</Data>
        <Data>5fb3100b</Data>
        <Data>coreclr.dll</Data>
        <Data>4.700.20.26901</Data>
        <Data>5ec48a29</Data>
        <Data>c00000fd</Data>
        <Data>000000000004a4c7</Data>
        <Data>237c</Data>
        <Data>01d6f7864632d69d</Data>
        <Data>C:\Users\LAUNCHBOX\LaunchBox\Core\BigBox.exe</Data>
        <Data>C:\Program Files\dotnet\shared\Microsoft.NETCore.App\3.1.6\coreclr.dll</Data>
        <Data>5f663230-da5f-4cd8-871d-d3e20a2d77b9</Data>
        <Data>
        </Data>
        <Data>
        </Data>
      </EventData>
    </Event>

     

  22. Just found this thread, hopefully a few months does not a necro thread make? I'm running 11.8 (stable) and am experiencing these same symptoms. I foolishly left "rough duplicates" checked, and it took my machine 36+ hours to complete a full rom import. It made it to one of the very last roms (zw-something) before hanging for 8+ hours on doing the rough duplicates check on that rom. I finally gave up in the morning and hit cancel, only for it to sit hung for over an hour. I finally gave up and closed launchbox. When I went back in to see the roms that had been imported, nothing appeared. It was as if the previous import process never happened. Discouraging, to say the least. I restarted the import and unchecked rough duplicates this time. I had only two errors throughout the first attempt and neither were fatal to the process.

    I don't know how the import is structured under the hood, but maybe placing a maximum cap on the number of roms per process would help? Something like "oh I see you're trying to import 30,000 roms. I'm going to split this up into 30 1,000-rom subprocesses where it finishes the whole import process end to end for each subprocess before starting the next." This would chunk progress and make it less risky to attempt a full (all or nothing) import. Especially for users like me who are not at all proximal to the Netherlands where these imports commonly take a very (very) long time.

×
×
  • Create New...