Activity for Kered13
-
Kered13 committed [b9fcf5] on Code
subrepo: avoid opening console window for non-native subrepos on Windows
-
also i noticed that when switching from clockwise to anticlockwise sometimes the computer did not recognise that i had changed direction and sometimes it did recognise. its hard to predict This is a consequence of the inherent fuzziness of the algorithm. It's hard to tell the difference between tracing a small circle and reversing directions. There are some hidden parameters that you can customize in the settings file, which are documented here. You can try playing around with them to see if you...
-
Mouse pointer remains active while the chiral gesture is activated.
-
That binary is only a debugging binary intended for testing the behavior of the mouse hook API on Windows 11. That binary is also not intended as a fix for the known issues on Windows 11, again it was only a debugging test. Some users may find it an acceptable work around for the issues on Windows 11, but that is not it's purpose. It is not intended for general release and this behavior is known and full expected. (There is still no fix for the W11 issue, and unfortunately I'm not sure if a fix is...
-
I probably will not be able to fix this, but it would help if you try the couple of debug builds that I posted in The W11 ticket and report the behavior that you get on your T15. There are two debug builds posted in that thread that use different command line options to test various behaviors. It would be great if you could test your T15 with both W10 and W11, but I understand if it's too much trouble to test W10.
-
I probably will not be able to fix this, but it would help if you try the couple of debug builds that I posted in The W11 ticket and report the behavior that you get on your T15. There are two debug builds posted in that thread that use different command line options to test various behaviors.
-
I attached the wrong file. Here is the correct file.
-
Here is another debug build to try. This one uses the WH_MOUSE hook instead of WH_MOUSE_LL. As a consequence, it requires a separate DLL file. Be sure to extract both the .exe and the .dll into the same directory. The --w11DebugBehavior flag is removed, you just need to test it normally. This build also does not lock the mouse position, I suspect that is not possible using WH_MOUSE. The purpose of this build is to test if the lag is being caused by the WH_MOUSE_LL hook or if all hooks have this issue....
-
Here is another debug build to try. This one uses the WH_MOUSE hook instead of WH_MOUSE_LL. As a consequence, it requires a separate DLL file. Be sure to extract both the .exe and the .dll into the same directory. The --w11DebugBehavior flag is removed, you just need to test it normally. This build also does not lock the mouse position, I suspect that is not possible using WH_MOUSE. The purpose of this build is to test if the lag is being caused by the WH_MOUSE_LL hook or if all hooks have this issue....
-
I have attached the test build. A command line option is added to test the behavior with a couple different code changes. The option is --w11DebugBehavior=<option>. The options are normal, which is the normal behavior, nohook, and hookdoesnothing. With the second two options, the mouse cursor will not be locked while scrolling. This is not desirable behavior, but I think the problem in W11 may be tied to this. Scrolling will still work but it will be awkward because the mouse will leave the window...
-
I have a theory as to what may be causing this. No solution unfortunately, but perhaps at least an understanding of the problem. If I sent you a test build, could one of you try running it?
-
This is definitely possible to do. I don't really have time to work on it right now, but I might get around to it later.
-
Unable to build because of unresolved external symbol
-
Yeah, I tested at home and I can build with v142 using VS 2022. I even deleted the vcpk_installed directory to ensure that I would have a clean build. I'm not sure what's causing problems on your end. I won't reinstall VS 2019 to test that, but the last time I had it installed (a few weeks ago) it was building fine. Ultimately, the platform toolset shouldn't really matter much. I don't think I'm going to upgrade it right now, but I might at some point in the future.
-
Save settings to file immediately after the user changes them
-
Good change. Honestly not sure why I didn't do this in the first place. I made some changes after pulling to clean up the related code. If I were to code review, I would say that we don't want to compute the settings path twice, because this could result in the two paths getting out of sync if a change were made, and we would be reading and writing from different files. This is one of the things that I changed before merging. I'm releasing a new version with this change.
-
Good change. Honestly not sure why I didn't do this in the first place. I made some changes after pulling to clean up the related code. If I were to code review, I would say that we don't want to compute the settings path twice, because this could result in the two paths getting out of sync if a change were made, and we would be reading and writing from different files. This is one of the things that I changed before merging.
-
That's odd though, because the project is still set to toolset v142 and still builds on my machine. Maybe some MSVS weirdness. https://sourceforge.net/p/chiralscroll/code/ci/default/tree/ChiralScroll/ChiralScroll.vcxproj
-
That other warning is innocuous, basically it's just saying that warnings are disabled for external libraries (like Abseil and wxWidgets). Otherwise they would flood the the build log with warning that I can't do anything about. Can you try Visual Studio 2022? I recently updated VS on both of my machines. I don't think that I changed anything in the project when I did that, but it's something to try. I can take a more thorough look when I get home.
-
What configuration are you trying to build? Do you have the platform toolset v142 installed? Are there any other warnings or errors when trying to build the project? When opening the project? I would guess that the problem has something to do with the standard library. Abseil and wxWidgets have included some std header that is using an internal function called std_find_trivial_1, but the linked standard library does not contain this function. I do not use the cxx17 feature for Abseil when building...
-
Hmm, try --logLevel debug. I'm not at home right now so I can't check, but something like this should work. Also -h should also show usage information.
-
Sorry, it should be logLevel=debug. Try that.
-
Sorry, it should be logLevel=trace. Try that.
-
No input options
-
No input options
-
Surface Pro X
-
Ah, that is an ARM laptop. I'm sorry but this app only supports x64. I took a look to see if compiling for ARM would be simple, but it looks like it won't be, so I'm afraid this will remain unsupported.
-
Is your new laptop Windows 11? It has been reported to me that the app does not work with Windows 11, but I can't investigate or fix the issue because I don't have W11.
-
The plugin manager has been updated with the latest version, see here. The updates to the plugin manager are released along with Notepad++ updates, and this one went line with version 8.4 of Notepad++ a couple weeks ago (I just verified that it has the latest version of the plugin), however that version has not been released to auto-update yet, it can only be manually downloaded. Unfortunately there is nothing I can do to fix the broken release in version 8.3.3.
-
python-indent error in npp8.3.3
-
The plugin manager has been updated with the latest version, see here. The updates to the plugin manager are released along with Notepad++ updates, and this one went line with version 8.4 of Notepad++ a couple weeks ago (I just verified that it has the latest version of the plugin), however that version has not been released to auto-update yet, it can only be manually downloaded.
-
A suggestion: any public forum for this project
-
I don't have the time to maintain a forum, and I don't believe it would get much use. If you have any issues or feature requests, feel free to file a ticket here. You may also use the Notepad++ community forums for discussion.
-
Updated for VS 2019.
-
Correction: The command to close ChiralScroll.exe should be "taskkill /F /IM ChiralScroll.exe". If you don't include the /F it won't actually close.
-
I just noticed behavior very similar to what you described today on my laptop, but when I looked into it more it turned out to just be that I somehow had two instances running at once. I'm not sure how that happened because I only had one icon in the taskbar, but maybe something to look into.
-
Alright, well if you feel up to it you could try the following and let me know what happens. Make sure ChiralScroll.exe is not running and that your touchpad is working. Start ChiralScroll, check if your touchpad is still working. If it stops working: Stop ChiralScroll.exe. Check if your touchpad starts working again. As a tip, you can close ChiralScroll.exe with only your keyboard by opening a command prompt (windows key + R, type "cmd.exe", press enter) then type "taskkill /IM ChiralScroll.exe"...
-
Alright, well if you feel up to it you could try the following and let me know what happens. Make sure ChiralScroll.exe is not running and that your touchpad is working. Start ChiralScroll, check if your touchpad is still working. If it stops working: Stop ChiralScroll.exe. Check if your touchpad starts working again. As a tip, you can close ChiralScroll.exe with your keyboard only by opening a command prompt (windows key + R, type "cmd.exe", press enter) then type "taskkill /IM ChiralScroll.exe"...
-
To be clear, when ChiralScroll.exe is not running, is your touchpad working or not?
-
I'm sorry, but this sounds like a problem with your computer. This application can have no effect if it's not running, it doesn't make any permanent changes to your computer or your settings. If you're made sure that ChiralScroll.exe is not running and you're still having touchpad problems, it's an issue with your computer or drivers that I cannot help you with.
-
This should produce a notification. Scintilla receives a WM_KILLFOCUS when switching to another window so calls CancelModes which calls AutoCompleteCancel which sends the notification. Which version of Scintilla is being used here? I'm embarrassed to say that this turned out to be a bug in my code. Notifications were added to inform applications of events that occur outside of their direct control. When the application calls SCI_AUTOCCANCEL it knows that the list will be closed so can perform whatever...
-
Provide notification (like SCN_AUTOCCANCELLED) in all events where autocompletion closes
-
SCI_AUTOCSHOW crashes on empty itemList with SC_ORDER_CUSTOM (Windows)
-
Thank you for reporting this. Unfortunately I do not have Windows 11 so I cannot take any action on this at the moment. Perhaps some other user with Windows 11 could attempt to investigate. If any other Windows 11 users can confirm this issue, that would also be very helpful. I can say that this sounds like the behavior when the debugger is paused in the middle of the mouse hook function (WinScroller.cpp, line 13), so perhaps it is related to that. The purpose of this hook is to block mouse move...
-
Thank you for reporting this. Unfortunately I do not have Windows 11 so I cannot take any action on this at the moment. Perhaps some other user with Windows 11 could attempt to investigate. I can say that this sounds like the behavior when the debugger is paused in the middle of the mouse hook function (WinScroller.cpp, line 13), so perhaps it is related to that. The purpose of this hook is to block mouse move events, therefore locking the mouse in place, while chiral scrolling.
-
I am already in the credits ("Derek Brown"). On Mon, Oct 18, 2021, 8:48 AM Neil Hodgson nyamatongwe@users.sourceforge.net wrote: OK, committed as [6d6294]. Credited to "Derek". If you prefer a different name in the credits please tell me what you would like. [bugs:#2288] SCN_DWELLEND incorrectly called after SCI_CALLTIPSETPOSITION, SCI_CALLTIPSHOW Status: open Group: Bug Created: Sat Oct 16, 2021 10:40 AM UTC by Kered13 Last Updated: Sat Oct 16, 2021 10:40 AM UTC Owner: nobody Attachments: 8985.patch...
-
I'm actually already in the credits ("Derek Brown").
-
SCN_DWELLEND incorrectly called after SCI_CALLTIPSETPOSITION, SCI_CALLTIPSHOW
-
Crash on legitimate exception "Too many/few contacts in frame"
-
Sorry I've been neglecting this, I got busy with other stuff. I never could find a cause for this, and I still suspect it is your driver misbehaving (counting contacts incorrectly). It's hard to tell for sure though since I can't reproduce it on my laptop. I have added a flag to control this panicking behavior. If --panicOnUnexpectedInput is provided, then these errors will cause a panic. The default behavior is to ignore them. The newest version is uploaded to the Files section.
-
crash on mouse click
-
Thanks, I appreciate it. But I don't need anything.
-
Ahaha, I just realized I can reproduce this error on my laptop by using a pencil to click the touchpad so it clicks without registering a touch. I think this has also made me realize that when I was getting this error before, it was caused by the laptop warping slightly making the touchpad click. So this thread has been at least somewhat productive. Anyways, I think I have a pretty good idea of what is going on now. Pressing the buttons causes input reports. I could try to track the state of these...
-
Okay, try this, I believe it will fix your problem. This ignores "invalid frames" when a button has been pressed. If you still get the crash, let me know if it's happening on press or on release. I'm not entirely sure how this will behave when you release the button, since I can't test this on my laptop. I'm still not sure if this is a correct solution or just a hacky workaround.
-
Okay, try this, I believe it will fix your problem. This ignores "invalid frames" when a button has been pressed. I'm still not sure if this is a correct solution or just a hacky workaround.
-
Scrolling direction not changed when rotation changed
-
Thank you. Does it also crash for (bottom) right click as well? Either way, can you post a log of pressing right click only? (If it doesn't crash you can just close it normally.)
-
Thank you. Can you do it one more time please? Here's a new version.
-
Appreciated. I have added a little more logging, can you please download this version, crash it, and post the log? My suspicion at this point is that I won't find real fix for this. As I said, I've seen a similar issue before but without a consistent cause. The check that is causing the crash is not strictly necessary, I only included it to check for an "impossible" condition, since that should indicate a bug in my code. However from what I have seen on my laptop, and what I think is happening here,...
-
Interesting, I've had sporadic crashes with this error message (which I haven't been able to fix, I actually was planning on just getting rid of the error message), but I've never had a reproduceable crash with it. I believe the Dell Latitude 5590 has two left click buttons, can you confirm that it crashes on both of them? The second thing I would like for you to do is to run the program with the flag: --logLevel=debug You can either run this from the command line or create a shortcut with the flag,...
-
That doesn't reproduce it on my laptop, but that's not surprising since my laptop reports touches differently. Based on the logs, It looks like your touchpad is reporting the wrong number of contacts when one or more contacts are lifted and a new contact is made on the same frame. For example in the second log we see: Expecting 4 contacts. Report: id=0, link=5, isTouch=false, x=0, y=0 id=0, link=1, isTouch=true, x=404, y=911 id=3, link=4, isTouch=true, x=1097, y=407 id=2, link=3, isTouch=true, x=969,...
-
Yes, I've seen errors like this on a few occasions on my laptop as well. Unfortunately I haven't been able to reproduce it reliably, so it's been difficult to track down. I'm leaving it crashing, at least for now, so I can know when it's happening and try to investigate it. If you can run the program with the "--logLevel devel" flag and then upload a log from one of these crashes, that will hopefully help find the cause.
-
Yes, I've seen errors like this on a few occasions on my laptop as well. Unfortunately I haven't been able to reproduce it, so it's been difficult to track down. I'm leaving it crashing, at least for now, so I can know when it's happening and try to investigate it. If you can run the program with the "--logLevel devel" flag and then upload a log from one of these crashes, that will hopefully help find the cause.
-
In short, this is intended behavior. I thought for awhile about how to handle the direction changing logic. I don't think a perfect solution exists, but I wanted something that would work most of the time with the least surprises. One thing I wanted is that if you move in a straight line scrolling should continue and the direction should not change. So you might start scrolling down, make a clockwise half turn, then continue dragging up. The scroll direction should continue to be down. But of course...
-
Working on some launches, not on others, on some laptop
-
Then it sounds like you were experiencing the same issue that the other user was having. That was fixed in the June 4th commit, and released around the same time. I will close this ticket then. It's good to have these logging options anyways.
-
Here's some reading on how the Windows Precision Touchpad interface works, if you'd like to understand better what these reports, links, and HID functions are doing. https://docs.microsoft.com/en-us/windows-hardware/design/component-guidelines/windows-precision-touchpad-required-hid-top-level-collections I will add to this some things that I have learned by debugging another user's issues. It will be useful to have this in writing for myself, even if you don't use it. A device has a number of links...
-
I have added a flag to control the logging level. This should make debugging easier in the future. There are two flags to be aware of: --logToConsole: Log to a command prompt instead of a file. --logLevel [level]: Sets the logging level. Most of the logging statements I have added are at the "debug" level. You can pull these changes from the latest commit, or download the .zip file, I have updated it again. Give it a try and let me know what you see in the logs. If you add any additional logging,...
-
Here's some reading on how the Windows Precision Touchpad interface works, if you'd like to understand better what these reports, links, and HID functions are doing. https://docs.microsoft.com/en-us/windows-hardware/design/component-guidelines/windows-precision-touchpad-required-hid-top-level-collections I will add to this some things that I have learned by debugging another user's issues. It will be useful to have this in writing for myself, even if you don't use it. A device has a number of links...
-
This is happening because the first line has the prefix "// " (note the space). Therefore when wrapping the second line the first three characters ("//s") are ignored and the rest is wrapped. Line prefixes are recalculated every line, which is why the rest of the comment has the two character prefix "//" after wrapping. I'll have to think about what the best way to fix this is.
-
Provide a mechanism to read EDGE_MULTILINE setting
-
Access violation
-
Thank you, I can confirm that I can reproduce the bug and will start looking into it. From an initial glance, it looks like the vertical edge settings in NP++ have changed, and this might be related to the problem, as my plugin reads these settings. However setting an explicit wrap column does not act as a workaround, which should bypass reading those settings. So it may be something else.
-
Please provide: Notepad++ version (including 64 bit or 32 bit). Sample text to reproduce the error. What language are you using?
1
Gen AI apps are built with MongoDB Atlas
Atlas offers built-in vector search and global availability across 125+ regions. Start building AI apps faster, all in one place.