Activity for Code::Blocks
-
Miguel Gimenez committed [r13843] on SVN Repo
- clangd_client: Fix compilation error with wxWidgets 3.0 on Linux.
-
[r13841] should fix the wxsresourcetree part. EDIT: tested OK on Fedora 44
-
lcs posted a comment on ticket #1596
OK. Thank you for your hard work on this.
-
- wxSmith: Enable direct loading after tree construction.
-
[r13841] should fix the wxsresourcetree part.
-
* wxSmith: Delay-load images in the resource tree (ticket #1561).
-
pecan posted a comment on ticket #1596
I've spent another day on this, and I'm calling it quits. I cannot get a clean compile of CodeBlocks using the AutoMake system from Msys2 ucrt64. I strongly suggest that you download a nightly version of CodeBlocks and use that to compile CodeBlocks with wx33 and the compiler in Msys2. I've done that and CodeBlocks does not crash using that method. https://wiki.codeblocks.org/index.php/Installing_Code::Blocks_from_source_on_Windows I'll close this ticket soon.
-
homertp created ticket #1599
Command line parameter --user-data-dir broken .
-
Carlo Bramini posted a comment on ticket #1598
The patch to codeblocks-code\src\build_tools\autorevision\Makefile.am:
-
The auto_revision script:
-
PATCH: new auto_revision tool for unix-like shells
-
* dragscroll: fix Makefile.am to set correct location of manifest.xml
-
* Dragscroll: Fix DragScroll_Windows.cbp to zip correct position of the mainifest
-
OK. I used wxWidgets 3.3 under MSYS2 UCRT (because I wanted Dark Mode) and clangd (mingw-w64-ucrt-x86_64-clang-tools-extra). All other dependencies were also provided by MSYS2. In the file src/src/Makefile.am, I added the line codeblocks_LDFLAGS = -mwindows before line 74. I built Code::Blocks using autotools with this configure command: ./configure --prefix=/d/App/codeblocks --enable-silent-rules --disable-pch --with-contrib-plugins=all,-Valgrind
-
OK. I used wxWidgets 3.3 under MSYS2 UCRT (because I wanted Dark Mode) and clangd (mingw-w64-ucrt-x86_64-clang-tools-extra). All other dependencies were also provided by MSYS2. I built Code::Blocks using autotools with this configure command: ./configure --prefix=/d/App/codeblocks --enable-silent-rules --disable-pch --with-contrib-plugins=all,-Valgrind
-
@lcs Please attach the project that invokes the crash so I can be as close to your environment as I can. Thanks.
-
@lcs Please attach the project that invokes the crash so I can be a close to your environment as I can. Thanks.
-
Gerard DURAND modified a comment on ticket #1597
I don't think that your problem is with slashes or backslashes. C::B and gcc admit both characters for path separator. You should check the real full path to your "other" gcc compiler and set accordingly the path to that compiler in your compiler's configuration (as told in the window, without the bin subfolder).
-
I don't think that your problem is with slashes or backslashes. C::B and gcc admit both characters for path separator. You should check the real full path to your "other" gcc compiler and set accordingly the path to that compiler in your compiler configuration (as told in the window, without the bin subfolder).
-
I don't think that your problem is with slashes or backslashes. You should check the real full path to your "other" gcc compiler and set it accordingly the path to that compiler in your compiler configuration (as told in the window, without the bin subfolder).
-
Quote: "Could it be possible that m_parser is not null but invalid?. The OP has just enabled clangd_client, may be m_parser is not initialized yet." Ah... Excellent suggesting. I'll have a look at this.
-
* DragScroll: 1.4.19 260430 Fix crash in Linux OnTimerEvent() using garbage event object
-
* DragScroll: Fix: Remove windows only function from compiling on unix
-
* DragScroll: 1.4.18 26/4/16 Refactoring and cleanup with -Wextra
-
I used Claude Sonnet 4.6 to analyze the RPT file, and the LLM guided me through GDB debugging. In the end, the LLM concluded that the problem is in wxXmlResource::DoGetXRCID inside wxmsw332u_xrc_gcc_custom.dll. I am using the package provided by MSYS2 UCRT64. The report it generated is attached, for reference only. (I don't understand assembly, so I don't know if the AI's analysis is correct.)
-
Could it be possible that m_parser is not null but invalid?. The OP has just enabled clangd_client, may be m_parser is not initialized yet.
-
I got it. Many thanks for your work on this.
-
wx3.3.2 is really buggy. Especially within wxAUI. Attached: Video of non-crashing function. You might want to re-install wx 3.3.2 to see if they've fixed anything. I'm using a download from yesterday. You might also re-download clangd_client. I had to make changes before it would cleanly compile. It had an "OVERRIDE" macro that wx 3.3.2 did not like. (Ref: commit 13835). The changes might help. Since this is an un-stable version of wxWidgets, I'll close this ticket soon, but I hope that won't discourage...
-
i've spent 16 hours on this and cannot re-create the problem. I've done my best to clone a matching environment using wx332. Whatever the problem it occurs here (see attached). The posted RPT file shows the crash at line 128. Hard to believe, but possible with such a buggy wx library.
-
wx3.3.2 is really buggy. Especially within wxAUI. Attached: Video of non-crashing function. You might want to re-install wx 3.3.2 to see if they've fixed anything. I'm using a download from yesterday. You might also re-download clangd_client. I had to make changes before it would cleanly compile. It had an "OVERRIDE" macro that wx 3.3.2 did not like. (Ref: commit 13835). The changes might help.
-
wx3.3.2 is really buggy. Especially within wxAUI. Attached: Video of non-crashing function. You might want to re-install wx 3.3.2 to see if they've fixed anything. I'm using a download from yesterday.
-
wx3.3.2 is really buggy. Especially within wxAUI. Attached: Video of non-crashing function.
-
wx3.3.2 is really buggyl. Especially within wxAUI. Attached: Video of non-crashing function.
-
i've spent 16 hours on this and cannot re-create the problem. I've done my best to clone a matching environment using wx332. Whatever the problem it occurs here:
-
* clangd_client: changes and cleanup for wx3.3.2
-
No. CodeBlocks crashed immediately. I just did a quick 23-second screen recording.
-
massimo soricetti created ticket #1597
C::B 25.03 uses wrong slashes in path to gcc under windows
-
@lcs Do you ever get this message: "You have changed the C/C++ parser search paths for this project.\n" "These paths will be taken into account for next parser runs.\n" "If you want them to take effect immediately, you will have to close " "and re-open your project."
-
Crash when adding a path in "C/C++ parser options" after using the Clangd_Client
-
Just as an aside about the "Database Error" when posting on the forum. It usually means that there is an utf-8 character in the post that does not map into a regular ascii character and the forum parser has a fit about it. The forum parser is old and and refuses to deal with the modern world. I'll look into the crash.
-
I wanted post a topic on code::blocks forum, but I got Database Error. Correction: I meant “I wanted to post a topic, but I got a Database Error when I submitted it.” Sorry for the typo.
-
I wanted post a topic on code::blocks forum, but I got Database Error. Sorry, I made a typo. I meant to say: "I had posted a topic on the Code::Blocks forum, but I got a Database Error."
-
Crash when adding a path in "C/C++ parser options" after using the Clangd_Client
-
Morten MacFly committed [r13834] on SVN Repo
* allow to collapse all projects at ones quickly (after re-opening a workspace with many project, for example, that are expanded)
-
ollydbg committed [r13833] on SVN Repo
* wxSmith: Add an option for whether we need to use the Create() function for the top level wxWindow class. (Thanks LR83)
-
MINGW: Cannot build C::B, missing link to boost_system library.
-
I compiled and installed Boost 1.90.0 and the problem is gone. I think you can close this issue. I also agree that configure should also return an error if boost < 1.69.
-
Adding -lboost_system would fix your issue, but break compilation for users with recent boost versions, namely >= 1.89. Boost 1.66 is almost 9 years old, IMHO is time to upgrade. Just as we removed support for wxWidgets < 3.0.0, consider this as removing support for boost < 1.69.
-
Since Boost C++ libraries don't provide support for pkg-config or others (unless provided by the package maintainers for a particular distro), perhaps it would be worth to support some variables for letting the users to handle this event. The configure script already provides support for some variables defined by the user, as you can see from the output on the console, see attached screenshot. So, I would like to suggest to add something like BOOST_CXXFLAGS and BOOST_LIB for letting the user to select...
-
Since Boost C++ libraries don't provide support for pkg-config or others (unless provided by the package maintainers for a particular distro), perhaps it would be worth to support some variables for letting the users to handle this event. The configure script already provides support for some variables defined by the user, as you can see from the output on the console, see attached screenshot. So, I would like to suggest to add something like BOOST_CXXFLAGS and BOOST_LIB for letting the user to select...
-
Thanks. As far I have tested, it's OK now.
-
Fixed svn 13832. Garbage in unintialized variable causing the first mouse right-click to be interpreted as a mouse drag. A missing event.Skip() was causing the request for context popups to be ignored (Linux). On high DPI monitors some mouse right clicks were missed because mouse jitter was being interpreted as a beginning drag.
-
Fixed svn 13832. Garbage in unintialized variable cause the first mouse right-click to be interpreted as a mouse drag. A missing event.Skip() was causing the request for context popups to be ignored (Linux). On high DPI monitors some mouse right clicks were missed because mouse jitter was being interpreted as a beginning drag.
-
Right clic does not work the 1st time in the managment window
-
Fixed svn 13832. Garbage in unintialized variable cause the first mouse right-click to be interpreted as a mouse drag. An missing event.Skip() was causing the request for context popups to be ignored (Linux). On high DPI monitors some mouse right clicks were missed because mouse jitter was being interpreted as a beginning drag.
-
* DragScroll: Fix missed first Manager mouse Right click.
-
I'm using boost-1.66.0-1
-
In recent boost libraries the system library is header-only, so detection of the library failed. Which version of boost are you using? EDIT: the change was made in 1.69, a binary stub was left for compatibility until 1.89. In 1.70 and 1.88 the binary exists and only exports dummy_exported_function()
-
In recent boost libraries the system library is header-only, so detection of the library failed. Which version of boost are you using?
-
MINGW: Cannot build C::B, missing link to boost_system library.
-
Code::Blocks released /Binaries/Nightlies/2026/CB_20260409_rev13831_win64.7z
-
-
-
-
I'm still using some old screens with the resolution 1920x1080, some of my clients are still using Windows 7 with smaller screen. Update the Scintilla code is hard.
-
I'm still using some old screen with the resolution 1920x1080, some of my client are still using Windows 7 with smaller screen. Update the Scintilla code is hard.
-
Mehdi Chinoune posted a comment on ticket #1471
Come on, 2026 still no support for HiDPI? What dispalays are you using?
-
Cool. Hope to test this work soon!
-
Thanks for the help chasing this down guys. The cause appears to be (not a missing event.Skip()), but an uninitialized boolean variable that had an int(30) in it. After the first right-click in the management tree, the var got it's first true/false value resulting in correct performance. That would also explain why the difference in compilers would provoke or hide the problem. While I'm here, some changes for high DPI need examination. I'll work with this for awhile to make sure I'm not "halluci...
-
Thanks for the help chasing this down guys. The cause appears to be (not a missing event.Skip()), but an uninitialized boolean variable that had an int(30) in it. After the first right-click in the management tree, the var got it's first true/false value resulting correct performance. That would also explain why the difference in compilers would provoke or hide the problem. I'll work with this for awhile to make sure I'm not "hallucinating".
-
Thanks for the help chasing this down guys. The cause appears to be (not a missing event.Skip()), but an uninitialized boolean variable that had an int(30) in it. After the first right-click in the management tree, the var got it's first true/false value resulting correct performance. I'll work with this for awhile to make sure I'm not "hallucinating".
-
Thanks for the help chasing this down guys. The cause appears to be (not a missing event.Skip()), but an uninitialized boolean variable that had an int(30) in it. After the first right-click in the management tree, the var got it's first value resulting correct performance. I'll work with this for awhile to make sure I'm not "hallucinating".
-
When I looked at the source code of the dragscroll.cpp, especially in the class MouseEventsHandler, do you mean that before each return statement, there is a call to the event.Skip();? I see some functions do not have that function call.
-
When I looked at the source code of the dragscroll.cpp, especially in the class MouseEventsHandler, do you mean that before each return statement, there is a call to the event.Skip();? I see some function does not have that function call.
-
After disabling all the plugings and re-enabling them one by one, I think the problem I have comes from the dragscroll plugin: when desactivated, no problem, when activated the problem is there (of course, on my msys2 config, because it seems that with other compilers there is not systematically this problem). To test this, each time, I quit C::B then restart it to be sure that I have this 1rt time behavior.
-
I've also experienced this problem. (Windows 11; CB svn 13831; wx328). Could those having the disappearing mouse right-click problem please list the plugins they are using along with the plugins version. This problem is usually caused by a missing "event.Skip();" statement in a plugin event. A list of plugins to inspect would be very useful.
-
I've also experienced this problem. (Windows 11; CB svn 13831; wx328). Could those having the diappearing mouse right-click problem please list the plugins they are using along with the plugins version. This problem is usually caused by a missing "event.Skip();" statement in a plugin event. A list of plugins to inspect would be very useful.
-
I've also experienced this problem. (Windows 11; CB svn 13831; wx328). Could those having the diappearing mouse right-click problem please list the plugins they are using along with the plugins version. This problem is usually caused by a missing "event.Skip();" statement in a plugin event. A list of plugin to inspect would be very useful.
-
- codesnippets: Don't open duplicate temporary .txt files. Fix crash possibility.
-
clangd_client: support versioned executable auto detection
-
Thanks Applied SVN rev 13830
-
* clangd_client: Apply ticket 1593 - Auto detect versioned clangd executable (Thanks Christo)
-
Prompt to Gemini AI: For CodeBlocks, what does a contributor mean by ticket 1593 when he states: Installing clangd from https://apt.llvm.org/ results in verisioned libraries, for eg. /usr/bin/clangd-20 in my recent installation. Attached patch helps to autodetect this. With patch code: ticket 1593 Gemini said In Code::Blocks ticket 1593, the contributor is addressing a common issue where the IDE fails to find the clangd executable (the Language Server Protocol provider) because the filename isn't...
-
Prompt to Gemini AI: For CodeBlocks, what does a contributor mean by ticket 1593 when he states: Installing clangd from https://apt.llvm.org/ results in verisioned libraries, for eg. /usr/bin/clangd-20 in my recent installation. Attached patch helps to autodetect this. With patch code: ticket 1593 Gemini said In Code::Blocks ticket 1593, the contributor is addressing a common issue where the IDE fails to find the clangd executable (the Language Server Protocol provider) because the filename isn't...
-
clangd_client: add option to parse only opened files
-
Great addition, thanks ! I'll be using this even on a new powerful laptop. Applied SVN rev 13829
-
Great addition, thanks ! I'll be using this even on a new powerful laptop. Applied SVN rev 18829
-
I Posted to wrong ticket.
-
Thanks, great addition ! Applied svn rev 13829
-
* clangd_client: Apply ticket 1592 parse only open files option (thanks Christo)
-
-
- Settings: Fix assertion due to specifying wxALIGN_CENTER_HORIZONTAL in an horizontal sizer.
-
- Remove wxALIGN_CENTER_HORIZONTAL in horizontal sizers and wxALIGN_CENTER_VERTICAL in vertical ones.
-
clangd_client: add option to parse only opened files
-
clangd_client: support versioned executable auto detection
-
Christo created ticket #1593
clangd_client: support versioned executable auto detection
-
clangd_client: add option to parse only opened files
-
I'm using the rev13826, plus I have some patches locally mainly on the codecompletion plugins. I'm under Windows 10 64bit, and I used the latest GCC version(15.2) in msys2/mingw64 platform. My wxWidgets version is 3.3.2. I don't have such issue.
-
I'm using the rev13826, plus I have some patches locally mainly on the codecompletion plugins. I'm under Windows 10 64bit, and I used the latest GCC version in msys2/mingw64 platform. I don't have such issue.
1 >
