XMBC 2.7 Beta

[phil]

OK, I have begun the next beta phase with XMBC v2.7 Beta 1.

There are only a few changes so far (I plan much more but my time has been very limited recently - sorry about that!)

The initial beta hopefully fixes some of the more serious bugs in 2.6 to do with keyboard layouts, but if its not fixes, I hope the extra logging will help tie it down.

You can get the beta HERE. Note that this link will always get you the latest beta version!

Changes since v2.6:

• 200: Fixed scroll window under cursor in Process Explorer.
• 201: Modified code to try and fix problems when English keyboard not installed. Improved logging to help highlight any issues.
• 202: Fixed duplicate startup entry causing XMBC setup GUI to be opened on startup.

Please please test this especially if you have non-english language/keyboard layouts installed. Any problems, PM me a copy of the log file (or post a snipptet in a code block here).

Thanks,
Phil

[phil]

OK Here is 2.7 Beta 2.

I hope I have the language/keyboard bug sorted and fixed now - please test this if you use a language other than English in Windows!

You can get the beta HERE. Note that this link will always get you the latest beta version!

Changes since v2.7b1:

• #201 - Found and fixed a bug causing an incorrect/invalid language to be chosen, thus causing simulated keystrokes to fail.
• #203 - Added Magnifier toggle action to open/close the magnify application.

As usual, let me know if things go wrong!
Thanks,
Phil

[000001a]

is there no poerable versions of the betafiles?

[phil]

Nope...
Sorry, but the portable version is limited anyway (esp. on Windows it hardly seems worth it.

[lee321987]

Import is bugged.
Right-click profile list > Import > select .xmbcs file

Only imports the first profile.

Repeat import same file > asks if you want to overwrite the first profile, then imports only one more profile.

Repeat > asks if you want over write the first profile, then asks if you want to overwrite the 2nd profile, then imports only one more profile...........................

[phil]

OK Thanks for the info - I will have a look when I get a minute!

[phil]

OK Here is 2.7 Beta 3.

You can get the beta HERE. Note that this link will always get you the latest beta version!

Changes since v2.7b2:

• #204 - Added logging and error messages when the XML settings / import files are corrupt.
• #205 - Fixed a bug when importing multiple profiles where only one would import at a time.
• #206 - Fixed a problem causing RCTRL and RALT to send left CTRL/ALT not right.
• Misc internal restructuring that should make the code more maintainable!

As usual, let me know if things go wrong!
Thanks,
Phil

[tros]

Hi! Firstly, thank you for such a great software.
I tried to work with the official 2.6.6 version, but it didn't intercept my additional side buttons of Pleomax MOC-315B mouse. So, I installed a beta, and my buttons began to work!!! Unfortunately I found another bug. I set a Backspace emulation for left bending of the wheel. It works, but proceed a double press of the button. I tried all the modes of button pressing and releasing, but the problem is still present. Maybe it depends with some settings, but I didn't found any
My system is Windows 8.1 and I use Russian and English keyboard layouts.
I'll glad to know how to fix this problem.
P.S. Happy New Year!!!
P.P.S. Enter emulation with the right wheel bending also proceed double pressing

[phil]

Hi, can you confirm, what is double pressing after the key simulation.. the key of the original mouse button?
I cant think of anything obvious in that... What version are you actually running (note. there was no 2.6.6, 2.6.2 is the latest release).

I think I will need to see the debug logging, you can turn that on in the advanced settings window (and then OK/Apply on the main window). Then, after reproducing the problem, send a copy of the log to me by email or post a snippet here.

But I'm off now for new year celebrations so I wont be able to reply for a few days or so.

Thanks,
Phil

[iamntz]

Hey man, thanks for updates. I have an idea which would be awesome to see it implemented in XMBC.

Until today, i used a microsoft mouse. MS driver have this thing that is basically an accelerated scroll: you slowly move the wheel, the scroll is slow. When you quickly scroll, the page starts to scroll faster (so basically one quick flick will scroll half or more of the document you have, no matter how long it is).

What do you think, is this possible to do?

Thanks and have an awesome new year!

[tros]

Hi, again
By the double pressing I ment, that one wheel bending makes emulation of two backspace pressing. One bending delete two symbols instead of one in every modes.
You're right, I've mixed up the version number with the latest official release 2.6.2 Sorry...
Here is the latest notes of the LOG with the trying to proceed Backspace in a Microsoft Office Word 2013 with the left wheel bending:

01-01-2014 12:54:26.0953> CMyMessageWnd::OnSetup - Opening setup dialog.
01-01-2014 12:54:26.0968> Setup dialog init took 0ms
01-01-2014 12:54:26.0968> Load Settings took 0ms
01-01-2014 12:54:26.0140> MouseHookData: Message=0x0202 (WM_LBUTTONUP), X=1412, Y=739, mouseData=0x00000000, flags=0x00000000, time=92628140, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:26.0140> Translating Left Button Up Keystate 0x1: Action 40 [** No Change (Don't intercept) **]
01-01-2014 12:54:27.0781> GetProcessName: Got name for PID 4248, 'winword.exe' for HWND 0x00020644 using GetModuleFileNameEx
01-01-2014 12:54:27.0781> CProcessHelper: Got name for PID 4248, 'winword.exe' for HWND 0x00020644 using GetModuleFileNameEx
01-01-2014 12:54:27.0828> GetProcessName: Got name for PID 3872, 'xmousebuttoncontrol.exe' for HWND 0x0003078C using GetModuleFileNameEx
01-01-2014 12:54:27.0828> CProcessHelper: Got name for PID 3872, 'xmousebuttoncontrol.exe' for HWND 0x0003078C using GetModuleFileNameEx
01-01-2014 12:54:28.0312> MouseHookData: Message=0x0201 (WM_LBUTTONDOWN), X=449, Y=677, mouseData=0x00000000, flags=0x00000000, time=92630296, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:28.0312> Translating Left Button Down Keystate 0x0: Action 40 [** No Change (Don't intercept) **]
01-01-2014 12:54:28.0421> MouseHookData: Message=0x0202 (WM_LBUTTONUP), X=449, Y=677, mouseData=0x00000000, flags=0x00000000, time=92630421, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:28.0421> Translating Left Button Up Keystate 0x1: Action 40 [** No Change (Don't intercept) **]
01-01-2014 12:54:29.0531> MouseHookData: Message=0x0201 (WM_LBUTTONDOWN), X=664, Y=199, mouseData=0x00000000, flags=0x00000000, time=92631515, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:29.0531> Translating Left Button Down Keystate 0x0: Action 40 [** No Change (Don't intercept) **]
01-01-2014 12:54:29.0625> MouseHookData: Message=0x0202 (WM_LBUTTONUP), X=664, Y=199, mouseData=0x00000000, flags=0x00000000, time=92631625, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:29.0625> Translating Left Button Up Keystate 0x1: Action 40 [** No Change (Don't intercept) **]
01-01-2014 12:54:30.0687> MouseHookData: Message=0x0201 (WM_LBUTTONDOWN), X=863, Y=573, mouseData=0x00000000, flags=0x00000000, time=92632687, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:30.0687> Translating Left Button Down Keystate 0x0: Action 40 [** No Change (Don't intercept) **]
01-01-2014 12:54:31.0796> MouseHookData: Message=0x0202 (WM_LBUTTONUP), X=863, Y=573, mouseData=0x00000000, flags=0x00000000, time=92632781, dwExtraInfo=0x0, MouseHookPtr=0x10261
01-01-2014 12:54:31.0796> Translating Left Button Up Keystate 0x1: Action 40 [** No Change (Don't intercept) **]

Have a good holidays!

[phil]

Hi Tros, I think I might know what the problem is...

You log does not actually show any left press of the wheel (left tilt) but I think this is due to the design of the horizontal scrolling of the wheel. You see, the wheel tilt does not work like a normal button, it actually repeats a single tilt message while you hold the wheel left (or right).

A normal button has a discreet DOWN and UP message. The tile wheel has a single TILT message that is repeated (by the device or the device driver) and every device is different.

Right now, XMBC can has some "de-bounce" timing options that you can play around with to try and ignore repeated messages from your device (they are in Advanced Settings) but they are far from ideal, and I suspect, the repeated backspace is actually because the underlying message is repeated.

Have a look at those advanced settings and changing the timings to see if you can reduce or remove the problem. If not, I'm afraid you will have to wait until I have redesigned XMBC to better handle this.

Thanks,
Phil

[phil]

Hey man, thanks for updates. I have an idea which would be awesome to see it implemented in XMBC.

Until today, i used a microsoft mouse. MS driver have this thing that is basically an accelerated scroll: you slowly move the wheel, the scroll is slow. When you quickly scroll, the page starts to scroll faster (so basically one quick flick will scroll half or more of the document you have, no matter how long it is).

What do you think, is this possible to do?

Thanks and have an awesome new year!

Hi, I'm not sure if this wheel acceleration is possible outside of the mouse driver. I guess it could be, especially as other applications seem to offer it, maybe by injecting more (or blocking existing) scroll messages depending on how many it sees, but that's going to be a bit tricky to integrate with the rest of XMBC reliably I think. I will add it to the todo list but please don't expect anything too quickly as right now I'm concentrating on bugs over new features (esp. ones that might need significant changes).

[phil]

OK Here is 2.7 Beta 4.

You can get the beta HERE. Note that this link will always get you the latest beta version!

Changes since v2.7b3:

• #181 - Make "Disabled because of scroll lock" more visible by changing the icon and red text on setup GUI.
• #208 - Add ability to detect and select Modern UI apps in window profiles.
• #209 - Added ability to disable scroll window under cursor for specific profiles (ie METRO/Modern UI apps)

As usual, let me know if things go wrong!
Thanks,
Phil

[tros]

Hi, Phil!

Thanks for your reply, but those settings also don't influence to the problem. Now I've set the max value (500 ms) for de-bounce, and other measurements (like Repeat tilt rate and delay between simulations, but I still have a problem.

I think, for now it's impossible to suspend a repeating behavior of a wheel, so I'll wait for the next beta or release!

By the way, I've written an article about your useful application on the Russian freeware soft portal BestFREE. I think, it may be a good advertisement

Thanks again for your work! Wish you success in the new year!