UCyborg

I mostly use 32-bit build of Basilisk, sometimes Waterfox if I encounter something that's not compatible with Basilisk (classic branch, multiprocess option disabled). Have Vivaldi installed in case I want Chromium engine feel. Multi-process browsers are very resource intensive, so I only fire up Vivaldi when I don't intend to open many tabs and/or other applications (have mere 4 GB of RAM to work with).

The credentials popup appears if the host PC has an user account which user name matches the user name of the currently logged in account on the PC accessing the share, but the passwords don't match. This applies even if eg. host PC contains a passwordless account with the user name John and you're accessing the share from another PC while you're logged in with password protected account with the user name of John regardless of whether password protected sharing is enabled or not.

I don't like any of the solutions mentioned either here or at other places. Neither tackle the problem at its core. GameUX invocation code is baked in the Shim Engine (I guess it's in %WinDir%\AppPatch\AcGenral.dll). The system compatibility database (%WinDir%\AppPatch\sysmain.sdb) contains entries for 2165 games; executable name and usually some of its properties and/or accompanying files' names for the purpose of matching the executable file to the game and invoking GameUX). It can be viewed with Compatibility Administrator (32-bit) that comes with Application Compatibility Toolkit 5.6 (for Windows 8+, there's Windows Assessment and Deployment Kit (ADK) for specific version). So the best way is to just tell the Shim Engine to ignore those entries so it doesn't even try invoking GameUX. There are two registry files in the ZIP file below, GameUX_DisableShims.reg disables the entries while the other restores the initial state. If particular game has already been launched before importing GameUX_DisableShims.reg, visiting the game's executable properties, selecting Compatibility tab and clicking OK is required for changes to take effect. https://drive.google.com/open?id=1G45mk_NUtUKGkC2TmREqGx9jhbUlN1RU GameUX shim has been removed from Windows 8 onwards, the issue is Windows 7 specific. Vista has GameUX, but no bug there.

I see, that other problem is not related. Here's a working link to Raymond Chen's blog post about this: https://web.archive.org/web/20190218192651/https://blogs.msdn.microsoft.com/oldnewthing/20120726-00/?p=7043 Probably best to wait until some more updates have already been pushed.

The old parameter still applies to Chrome 75, the new one applies to Chrome 76. I guess only time will tell about the future versions.

Wow! Might not be related, but:

I checked the code, they just changed the command line parameter for disabling custom title bar. The correct parameter in newer Chromium browsers is -disable-windows10-custom-titlebar The decision is made in chrome/browser/win/titlebar_config.cc.

Does FullEventLogView show them the same way?

If anyone's interested, Vivaldi is one of the more interesting Chromium based browsers and has the option to enable native title bar (Use Native Window under Appearance in Tools->Settings). It doesn't put tabs on it in that mode.

I think what you're trying to accomplish relies on the shell's internal logic, the option along with all supporting code would probably have to be added to the dialog that opens when you click Change desktop icons from Personalization control panel. So, since MS won't do it, the only option would be analyzing system DLLs and injecting new code into them, a non-trivial task I imagine.

Thanks, this one works. Still find it odd; GreaseMonkey 3.11 doesn't have compatibility issues with Firefox 51. Waterfox on the other hand runs them all.

There are already bugs with existing addons (at least with Basilisk). Options dialog in GreaseMonkey doesn't work. Also this particular fork's default configuration breaks ColorfulTabs (it expects Firefox version number in the user agent string).

The option is really gone in Chromium 76. They removed conditional check (feature toggle via command line parameter) that used to enable native caption bar. There's no hidden registry setting for it or anything like that.

Moonchild Productions ported the custom caption bar drawing code to their Pale Moon fork as well, see this thread. You'll find my comparison screenshots if you scroll down a bit. They say there would be visual glitches on Windows 10 without custom caption bar drawing, but I don't see anything off with the native caption bar. The caption text is always enforced by Pale Moon, the only other difference I can see is spacing between min/max/close buttons and the caption bar height (I might have reduced default caption height setting in Windows, but I'm not 100% certain). So honestly, I've no idea what they're talking about. Looks like the only option that still exists in the current nightly is setting browser.tabs.drawInTitlebar in about:config to false to enable the native caption bar (and put tabs separately). CSS moz-win-glass property is indeed gone.

Back when I wrote my last post here, policy setting for InPrivate mode was still called IncognitoModeAvailability. Now it's called InPrivateModeAvailability. There's also -inprivate command line parameter that is accepted by msedge.exe, equivalent to Chromium's -incognito.

Recently experienced the same issue. I was trying to boot one of OSes I have on the USB flash drive and it failed. Plugging the drive in a running Windows machine showed it as unformatted. Could access the data from Linux, TestDisk showed invalid boot sector and some unimportant files were corrupted. I could probably avoid reformatting if I knew the correct numbers for drive geometry (they were different than after format). At the end, I also had to change drive's unique ID with diskpart because it apparently changed to the exact same one my laptop's disk has, so Windows refused to mount it. I'm certain the actual cause was loose connection. If you even slightly touch the drive so it moves just a little to the left or right when plugged in, it can lose power. That's how my laptop's ports are and the fronts ones on my desktop aren't much better.

Yes, it seems they have to be made specifically for non-default taskbar configurations. When I have more time and if I figure out how to do file selection and such.

I wonder how much effort it actually takes to provide 32-bit builds. Is it just the compilation part or are there also significant chunks of processor specific code? Graphics cards with 2 GB of video RAM were already common in low-mid price range 5 years ago. So that takes the half of the 32-bit addressable space, so one can use 2 GB RAM max. 4 GB card would take that number down to zero. Then there's PAE, which, AFAIK, on today's Windows can only be utilized by unofficial patches to the kernel (assuming drivers in use on the particular system aren't buggy).

I noticed clicking on the thread title from the index page (https://msfn.org/board/) takes me to the last post. Obviously, the selection of threads is limited to specific one per each sub-forum.

Compatibility section is irrelevant when it comes to Windows Vista because it's only recognized on Windows 7+ and even there, the Vista supportedOS entry alone is meaningless and equivalent in behavior as if compatibility section is absent. The OS looks for the entry of the highest Windows version it recognizes (so either itself or that of the older version) and modifies the behavior of certain APIs accordingly. https://docs.microsoft.com/en-us/windows/desktop/win7appqual/compatibility---application-manifest

https://techcommunity.microsoft.com/t5/Microsoft-USB-Blog/What-is-new-with-Serial-in-Windows-10/ba-p/270855

chkdsk is useful to get certain file system structures back in order if they were corrupted, though it can't ensure data integrity. There are also cases when TestDisk fails. Never had problems with USB thumb drives unless the one in question was actually faulty.

Disregard my previous post. Yup, that's it. Works as intended. When you think about it, it would be silly if clicking thread title would take you to the last post instead of that being separate action.

Looks like users of certain forks aren't affected. Waterfox, Basilisk and Pale Moon have xpinstall.signatures.required set to false by default.

Right, should've said that it's supposed to be stored on the server. Actually, I think it still is, but I guess some of the old data was lost. At least in this very thread, I see the "New content begins here" marker, while it's absent if I view any of the old threads I've visited before, but have new posts added since. Edit: Or instead of some data being actually lost, could just not be interpreted by the updated forum software.