UCyborg

The workaround relies on the codepaths still in place for Windows 7/8.x. When the support for those systems is dropped, they'll probably dump that code as well...

Yours lack the processor architecture bits, so it's actually: Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:60.0) Gecko/20100101 Firefox/60.0

Yes. STAC9921.xml

Figured that's impossible; I found the following code at widget/windows/nsWindow.cpp: // Glass hit testing w/custom transparent margins LRESULT dwmHitResult; if (mCustomNonClient && nsUXThemeData::CheckForCompositor() && /* We don't do this for win10 glass with a custom titlebar, * in order to avoid the caption buttons breaking. */ !(IsWin10OrLater() && HasGlass()) && WinUtils::dwmDwmDefWindowProcPtr(mWnd, msg, wParam, lParam, &dwmHitResult)) { *aRetValue = dwmHitResult; return true; } Stock caption buttons are effectively disabled on Windows 10. The behavior is hardcoded in xul.dll. If one searches for Windows version checks, there are some other parts of the code that will behave differently depending on the detected Windows version.

It's identical to the one that Firefox 60 ESR uses since version 56.2.8.

I did try all possible IRQ values, it always sounds the same. Again, changing it is only possible on Windows 95, on Windows 98 I'm stuck with 11. I don't know how to tell whether IRQ is exclusive or not. I also tried HDAICOUT.HDA that comes with the driver, nothing changed. I tried Mpxplay, but only got silence.

Haven't heard about that problem before. I get very obvious visual anomalies on my end, something to do with lack of certain optimizations that only later versions of Aero Glass have. The problem is more profound when ran on virtual machine due to worse graphics performance. Maybe there's something interesting in C:\AeroGlass\debug.log and in Event Viewer. You should be able to reboot to Safe Mode and access those things from there.

Doesn't work.

That wasn't directed at you specifically, but those that keep creating new threads even though they could post in the old existing threads. I've no clue what the problem on your end is because the transparency part has always worked on my end. Actually, with my custom theme in Windows, I get some transparency even without using userChrome.css (no idea why that works in the first place): The frame goes solid black after turning on dark mode: Creating the userChrome.css with the following content... #main-window { background-color: transparent !important; -moz-appearance: -moz-win-glass !important; } .titlebar-button { opacity: 0 !important; } ...gives this: Built-in CSS sets display property of .titlebar-button to none on Windows 7, but that doesn't help on Windows 10 due to some hardcoded behavior; it just completely disables the ability to use those buttons. If you turn off critical CSS files in Style Editor such as browser.css (there are 2 files called browser.css), you can still minimize/maximize/close Firefox window with the caption buttons on Windows 7/8.x, but not on Windows 10. So yeah, not all behavior can be changed with CSS alone. It would require some modifications to Firefox's source code.

People keep creating new threads for the same old problems over and over again because searching is too much of the effort for the average Joe.

@djmagnifique You're not using userChrome.css alone, you obviously also set browser.tabs.drawInTitlebar to false. I would prefer to be able to keep tabs in the title bar, but that way, there are issues with caption buttons. Also, I remember Semtex already tried these kinds of CSS tricks, but still encountered some problems.

The problem is that apparently nobody understands how this all works internally to be able to come up with good solution. I recently asked about it here.

That's because it doesn't do anything impressive, it just makes Firefox unaware of the fact that it's running on Windows 10. And since dark mode is Windows 10 thing, it won't work if it's run in the context of the older OS.

Support for UWP apps has been buggy since forever. The only thing you can do to avoid such glitches is to turn off that component of Aero Glass by running regsvr32 /U C:\AeroGlass\DWMGlass.dll from administrative Command Prompt.

I'm not an expert when it comes to IRQs, I can only tell what I've seen. No issues on Windows 7 guest. No driver I tried worked on Windows XP though, got the error that the device cannot start (code 10) in Device Manager. Doesn't help with anything, including volume control.

Sorry for the messed up post, spoiler tag seems to be buggy. Even though I only enclosed HDACFG content in the spoiler tag, the forum software trapped all the following text in there as well. The editor is royal PITA.

6.0.4

I see, 7-Zip 15.05 does show the decompiled script in this case. That remark about language files applies generally when dealing with anything where those languages are used. I merely wanted to point out that the way your screenshot looked isn't how it's supposed to look. We were discussing beta version of PotPlayer, which drops XP support, the current stable version still supports it officially.

Or you could just do what was said here: Your method is worse. Edit: Since Chrome 74, one must modify Chrome's shortcut to disable custom drawn title bar, see this post. BTW, compatibility settings for Chrome are automatically cleared on launch, at least if they are set for the current user, rather than globally for all users.

There's no easy straight-forward way to de-compile NSIS installers AFAIK, so I just extracted the files from it with 7-Zip. You can make ZIP package by extracting everything except the following: $0 $PLUGINSDIR Alarm.wav D_Exec.exe LogManager.exe PotNotify.exe PotPlayer.exe PotPlayerMini.exe uninstall.exe.nxis Then you can rename PotPlayerXP.exe to PotPlayer.exe and PotPlayerMiniXP.exe to PotPlayerMini.exe in the folder where you've extracted the files and ZIP it up. Usually, only Mini EXE files are installed. I remember that with the older releases, you got both normal and mini versions. The normal version can connect to Kakao's streaming services, so it's possible that it gets installed only if you choose Korean language during installation or if regional settings in Windows are set to South Korea, but I haven't experimented with that, so that's just my guessing. Also, when you upload it to the FTP, make it clear that it's only for use on Windows XP. The XP executables throw an error if you try to run them on Vista+. I don't have Server 2003, but the latest version does work on XP x64 Edition, which is similar under the hood. Yep, that's mandatory. Files for East Asian languages have to be installed for that text to render properly (Control Panel->Regional and Language Options->Languages).

To those that might be interested, whenever PCem v15 is done, it'll come with some speed improvements. https://pcem-emulator.co.uk/phpBB3/viewtopic.php?t=3162

Nope, no IRQ conflicts. BTW, QEMU can also emulate Intel HDA, but I haven't tried it yet. Both VirtualBox and QEMU are open-source.

Can also confirm PotPlayer 1.7.18193 actually runs on XP, just the installer was modified to output an error when ran on XP (simple version check). Regarding PotPlayer*XP.exe, they've been using separate executable for XP systems for several years now. PotPlayer.dll, where the guts of the player are, cannot be normally loaded on XP, so they use separate executable with some kind of workaround to get it to load. On Vista+, the normal workaround-free .exe is installed while .exe with a workaround is installed on XP. Loading PotPlayer.dll the usual way produces access violation exception in its initialization code due to some differences in Thread Information Block, value at FS:[0x2C] is NULL on XP.

That's true, I just assumed the possibility of using a different sound card when I said that. No, there's also AC97 and SoundBlaster 16 in VirtualBox, the point is just to see the HDA in action. Oops, I wrote "HDA Audio" in my previous post when I really meant "HDA Sound". Advanced tab was renamed to Devices on Windows 98. There, the audio driver appears like on Dave-H's screenshot. Not that I know of. It appears under System devices. I wonder if choppy audio under VirtualBox with HDA is just a simple configuration issue. AC97 works OK. I only know about the volume application that comes with the driver. Its Readme only says it may not work on some computers. And it really doesn't on that virtual computer. I see there's the volume setting in HDAcfg.ini at the end. How to interpret the value?