Fixes#153. AMD display profiles with rotated displays will now show that rotation properly in the display layout. Last commit included NVIDIA and Windows, and this now means all supported video cards now have this feature so the issue can be closed.
Added display rotation awareness to the Screen awareness parsing. This code is used to create the generic 'Screens' structures that are used to generate the display profile icons. This partially fixes#153. I still have to add the AMD configuration equivalent!
There was an error that occurred in NvAPI_SetDisplayConfig when attempting to go from an NVIDIA Surround to a non-NVIDIA Surround setup on a machine that has multiple video cards installed in it. This is due to the fact that the NVIDIA driver only sees the displays connected to NVIDIA video adapters.
The previous DM logic tried to set the DisplayConfig after switching to or from a Surround display profile, but this would fail when returning from a a Surround display profile on Win 10 devices. It appears that the NVIDIA driver AUTOMATICALLY disables all additional displays in windows when it returns from a Surround profile. This simple fact means that the DisplayConfig won't be applied properly, and it errors with an NVAPI_INVALID_ARGUMENT error. This doesn't actually matter though, as the WinLibrary comes to the rescue.
WinLibrary can see all the video adapters available, and so will turn the required displays back on and set them up just right! We *may* lose are some specific DisplayConfig parameters abeing set as part of this process, but I'm not totally sure about that as WinLibrary is basically feature parity with the DisplayConfig settings as far as I can tell.
The fix is to look specifically for the NVAPI_INVALID_ARGUMENT error when attempting to set the NvAPI_DisplayConfig, and if this happens we check if we were going from a surround profile to a non-surround profile. If that is true, then we simply ignore that error. WinLibrary then clears up that problem and everything proceeds as normal.
This should (fingers crossed) fix#119!
This change also makes it far faster to grab and set the tabaskbar settings from registry, though this logic may not detect some Windows 10 formats which appear to be LOCALDISPLAY(\d,\d,\d\d) settings which I've never seen before. It should be generally much faster and more reliable.
We now properly set the default JSON values for everything except Display Profiles, which should avoid crashes to desktop if we miss setting a value from it's default in the future. This should help with application robustness.
WinLibrary currently waits 5 seconds if it can't read the taskbar registry, and then it tries again. This is because based on my testing, if a screen layout changes, windows takes up to 20 seconds to update registry to record this fact. We have to wait until windows has finished 4 times before we are sure to have passed the 20 second window.
This is likely the delay you have mentioned. I *think* that I can slightly speed this up. We only MUST to do this delay when we are recording the config (i.e. creating a new display profile), and other times it's kind of a nice to have. So I've attempted to speed this up using a 'fastScan' option for the WinLibrary GetActiveConfig function. This will enable it to only query once for the general scans of the active config, and if there is a problem getting the data it will just accept that fact and will still return quickly. But it will still take up to 20 seconds when creating a new display profile as it is REALLY important we get that data correctly.
Fixes#129
Updated the NLog exception handling to make use of the additional logging options released recently. This will help provide me with more details when users create the Support ZIP File.
There is a possibility that there are two displays with the same UID. This happens with software created displays such as SpaceDesk and Superdisplay. This means that we need to try and figure out which screen is the right screen, and then when we know which screen we were really wanting to
TaskBarLayout now catches exceptions recording the taskbar location and handles them properly. It will skip any screenss that it cannot access the taskbar information for. The Screen layout generator for NVIDIA, AMD and Windows have all been updated to handle having no taskbar layout information (it just assumes the taskbar is down the bottom of the screen).
Hopefully fixes#114
There was a flaw in the way that ProgramSettings was being used that meant there was a lot of reloading and saving going on which was unnecessary. Additionally there were some errors in the way that the Settings file was upgraded which mean that the old settings were keeping on getting overwritten. Fixes#104.
Made some adjustments to when the MainForm is loaded, and how the NotifyIcon works for system tray.
Also Adjusted the way that the Windows Screens are computed so that we can reuse some of the code in AMD and NVIDIA screen generation so that the additional non-NVIDIA and non-AMD connected screens are still shown.
This will ensure that all screens show up in the images even if they aren't connected through an NVIDIA device. I also made the same changes to the AMD mode.
This is a subtle error, caused by either the NVIDIA driver being slow, or windows being slow in updating the location of the windows taskbar edge. This was causing an error in the GetNVIDIASCreenPositions function. Have added a 0.5 second delay to allow windows to update iit's display layout before we try to read it which should help, and also if we can't find the taskbar location in the settings, we just assume it is at the bottom. This should avoid exception.
Added specific checks to find and alert on 0 width or 0 height screens. This will then break other parts of the display layout.
Also moved the cloned screen check higher up to ensure we skip them properly.
Had some mistakes in the way that the profile graphics images were calculated on NVIDIA video cards. This meant that in some situations it calculated the image structures incorrectly, and didn't put the displays in the right places. This now changes that fact.
There is still an outstanding issue that the GetNVIDIAScreenPositions function still doesn't put the taskbar location in the right place yet. I need to fix that tomorrow.
I have been attempting to force Windows to refresh the notification tray using various Win32 API methods by sending messages to those notification tray windows to get them to refresh, but I haven't been able to get the right combination of setings. I am now leaving it for the next littlw while to concentrate on other issues.
There was a bug in the way that NVIDIA screen layouts were calculated. This error was due to a misunderstanding from me on how additional screens were listed within a mosaic config. Turns out additional screens are still recorded in the mosaic config, only just as 1x1 display grids. The screen finding logic now takes this into account.
Partially through a different refactor to completely change how the taskbar settigns are stored, and the logic behind the application. Should make the taskbar more robust to apply.
Found that I was combining Threads and Tasks together...never a good idea. Have moved fully over to a Task based model. Now still having issues with parallelism.
This corrects some errors with the WinLibrary and how it handles cloned displays.
Also removed the 0.5 second delay between the end of the NVIDIALibrary call, and the start of the WinLibrary one, because they do not appear to be needed based on the NVIDIALibrary testing we've been doing with NVIDIAInfo.
Also corrected some log entries and comments.
I *think* that NVIDIA Surround and AMD Eyefinity changes are so great that WinLibrary gets confused and gives up. NVIDIA Surround and AMD Eyefinity affects the windows display layout so much. This change adds a windows config refresh after the NVIDIA or AMD config is applied to make sure that the winLibrary calls are using the latest updated active config that results from the NVIDIALibrary and AMDLibrary config changes.
Found some calls that were bypassing the video library state buffering I'd put in place earlier. This should generally make DisplayMagician faster and minimise the number of log entries written to DisplayMagician.log.
The profile updating logic wouldn't work with the new way of buffering the display config, which means that when swapping back to DisplayMagician after updating the display layout wouldn't automatically update the display profile. Now, it forces this.
Changed Profiles to use new methods for profiles detection provided by the new video card library updates for NVIDIALibrary, AMDLibrary and WinLibrary. This should make DisplayMagician faster as there is less repetition.
Weird error where having a clone causes an extra 'zeroed out' entry in the Windows Display Modes sequence. This change modifies the Windows mode info checks to allow for this situation.
Updated the profile so that it applies the new split video application process properly to DisplayMagician profiles. Should fix#45 (we shall soon see).
AMD Eyefinity now works just as well as the NVIDIA implementation... probably even more robustly than the NVIDIA one. Fixes#37 and #21. I'll now start on some other bugfixing and features in preparation for DisplayMagician v2.0.0
