Activity for Tiger Beard
-
Tiger Beard created ticket #1480
wxSmith: Contrib tool "wxImagePanel"
-
wxSmith Usability: Sort "Layout" Tabs according frequency of use
-
wxSmith Usability: Move Toolbuttons to the left
-
Version : svn-r13432 SDK Version : 2.25.0 Scintilla Version : 3.7.5 OS : Linux 4.15.0-213-generic x86_64 wxWidgets Library (wxGTK port) Version 3.0.5 (Unicode: wchar_t, debug level: 1), Runtime version of toolkit used is 3.22. Compile-time GTK+ version is 3.22.30. MyLocale: hm. not really sure. Libre office uses comma, wxControls use dot
-
wxSmith wxSpinCtrlDouble Bug
-
@precan the issue I found does not cause feezing of the interface and also the specific piece of code is not called when adding or removing a breakpoint during debug. Its only called when starting the debugger.
-
Removed.
-
I checked my local version and it does not have the "2". so it should be removed. I think I put them in to allow to distinguish my bits in thread search from all other "Cut" occurrences. Sorry about he tabs here. I only work with tabs, but I think the edit option "use TAB" will to the trick. Please say so if I should re-create the patch
-
Thanbks for your efforts. I am sorry I was not aware that spaces are required. Do you want me to redo the patch with spaces?
-
Oh, your are right, I got it backwards. Ideed, the correct text for the default behaviour is "'Match Comments' = true". Sorry for the confusion.
-
PS: the patch I am missing most for current head is 1437, the context menu addon for wxSmith. You were checking this out as well. I got so used to that feature when working with dialogs that I found it hard to use trunk for esting with that feature missing.
-
Thanks for reviewing my patch. no I did not mean "'Match Comments' = true". The line I meant is a static text line in the configuration panel telling the user what options are set when the context switch version of the command is used. Myself I was not aware up to making the patch that the context menu search is different from the current parameters set in the options setting, but I can see that in some situations it might make sense. So in order not not change the current behaviour any new option...
-
Can someone please close this ticket, see ref to patch in #1437
-
Thats good to know. So I am going to explore some more with extra care.
-
Thanks for looking into this. I tried to find the right place in the code which is called when the Edit control of that line is executed, but the wxSmith stuff is highly dynamic and I would not figure it out and gave up after a few hours. If you happen to know where I can place a breakpoint I can try to debug into it. btw: if you have a good overview of wxSmith you might help with another thing. In tickets/1437 I added a context menu for wxSmith that makes working with it a lot more fun and it works...
-
The system uses ','. The other property fields work fine using ','. I tried using '.' as well, but it was not accepted in neither field.
-
Regression somewhere after rev13411 wxSmith wxSpinCtrlDouble
-
@ollydbg any chance to look at he code yet?
-
As I am not sure how to modify OllyDbgs patch I would skip the static text change above and have the patch committed I am working since one month with it and its working fine, no side effects found. What do I have to do?
-
Thanks, I did not see that one.
-
Legacy Codecompletion SymbolBrowser not updated correctly with BottomTree option OFF
-
OK, that would require the new tree to also restore the bold status which I think it does not do. When you say bold, the GUI tree does not show anything bold but inverse for selected. I guess that is the same?
-
SymbolTree CRC32 differs with identical Tree
-
@pecan I think the current trunk is missing one required change for this to work. Without the following function the Test2 misbehaviour can still be observed. Whem the top tree has SavedItems this function parses them, but when the BottomTree is not activated the SavedItems are not applied. File: codecompletion/classbrowser.cpp old code void ClassBrowser::SelectSavedItem() { ... if (parent.IsOk() && m_ClassBrowserBuilderThread && m_Parser && m_Parser->ClassBrowserOptions().treeMembers) { m_CCTreeCtrl->SelectItem(parent);...
-
Based on the proposal above I have implemented the test code below. The fix finds out how many init command line the user has added and adds for the surplus lines the dummy command "userInit". That should balance out with the GDB Parsing replies. The code does what it should but the results are not stable. When using multiple Breakpoints with multiple debugger starts, sometimes there is still a -1 index in there. I saw that the "ouput" string to parse sometimes contained 2 GDB prompts instead of...
-
I debugged the issue and I think I have found the root cause. Its in the DebuggerGDB's Option for User Init commands. This causes the above simple procedure to sometimes work or fail. To reproduce just add two two lines in the Options Init box (e.g. the following two lines) and the test will fail in any project with the first breakpoint set before the debugger is started. It can not be removed any more while the debugee runs, because its index in CB is -1. The GDB index is different. /home/user/...
-
I debugged the issue and I think I have found the root cause. Its in the DebuggerGDB's Option for User Init commands. This causes the above simple procedure to sometimes work or fail. To reproduce just add two two lines in the Options Init box (e.g. the following two lines) and the test will fail in any project with the first breakpoint set before the debugger is started. It can not be removed any more while the debugee runs, because its index in CB is -1. The GDB index is different. /home/user/...
-
I debugged the issue and I think I have found the root cause. Its in the DebuggerGDB's Option for User Init commands. This causes the above simple procedure to sometimes work or fail. To reproduce just add two two lines in the Options Init box (e.g. the following two lines) and the test will fail in any project with the first breakpoint set before the debugger is started. It can not be removed any more while the debugee runs, because its index in CB is -1. The GDB index is different. /home/user/...
-
I debugged the issue and I think I have found the root cause. Its in the DebuggerGDB's Option for User Init commands. This causes the above simple procedure to sometimes work or fail. To reproduce just add two two lines in the Options Init box (e.g. the following two lines) and the test will fail in any project with the first breakpoint set before the debugger is started. It can not be removed any more while the debugee runs, because its index in CB is -1. The GDB index is different. /home/user/...
-
I debugged the issue and I think I have found the root cause. Its in the DebuggerGDB's Option for User Init commands. This causes the above simple procedure to sometimes work or fail. When the following two lines are entered in the Optione Init box the test will fail in any project. At least the first break point set before the debugger is started can not be removed while the debugee runs. /home/user/ wxWidgets-3.0.5.1/misc/gdb/print.py skip file string.h These two commands will be sent as a single...
-
I debugges the issue and I think I have found the root cause. Its in the DebuggerGDB's Option for User Init commands. This causes the above simple procedure to sometimes work or fail. When the following two lines are entered in the Optione Init box the test will fail in any project. At least the first break point set before the debugger is started can not be removed while the debugee runs. /home/user/ wxWidgets-3.0.5.1/misc/gdb/print.py skip file string.h These two commands will be sent as a single...
-
Terriffic, thanks a lot for your help!
-
Yes I can confirm that with "one per workspace" I also can switch between projects using he first-open-file procedure. Great. With the "one-per-workspace" I am getting a works the tree closing behaviour. It seems to be closing for each namespace for the first time. I restarted with "one per project" and in only closes on the very first selection. However, from what limited understanding I have now it seems that the tree is still only created to the root level on the very first iteration. I have not...
-
I just ran CodeCompletion and trap on OnEditorActivated, then stepped through it. It was almost instant. Thanks that was very helpful, now I know for sure that there is an issue. I use the same gdb version. First step is always awfully long and also on first break point call stack is missing which is very bad for crash debugging. Also I have this serious issue with breakpoints that I can not delete (they are still active). So I guess there somethig seriously wrong on my system. Trouble is I have...
-
I just ran CodeCompletion and trap on OnEditorActivated, then stepped through it. It was almost instant. Thanks that was very helpful, now I know for sure that there is an issue. First step is always awfully long and also on first break point call stack is missing which is very bad for crash debugging. Also I have this serious issue with breakpoints that I can not delete (they are still active). So I guess there somethig seriously wrong on my system. Trouble is I have this on multiple computers running...
-
Yes. After the change TestCase 2 happend only the first time which is expected since its the same situation as with the patch. Further changes to the header do not cause the tree to change. What I did was to comment out this patch // m_topCrc32 = CRC32_CCITT; // m_bottomCrc32 = CRC32_CCITT; // if (m_CCTreeTop) // { // delete m_CCTreeTop; // } // if (m_CCTreeBottom) // { // delete m_CCTreeBottom; // } and replaced it by this + if ( !m_CCTreeTop ) + { + m_CCTreeTop = new CCTree(); + m_topCrc32 = CRC32_CCITT;...
-
Thats would explain why I did not see that behaviour in an older version. Revert it in your code to see if this is what is causing the problem willdo PS: I have changed my post above after further testing.
-
Thats would explain why I did not see that behaviour in an older version. Revert it in your code to see if this is what is causing the problem willdo PS: I have changed my post above after further testing.
-
Thats would explain why I did not see that behaviour in an older version. PS: I have changed my post above after further testing.
-
I tried to debug into a case from Test2 where the tree is closed on saving a header. What I see is the following sequence //entry is here and almost always the build tree is called ClassBrowser::UpdateClassBrowserView(..) ThreadedBuildTree(activeProject); //But while this builds the tread the first time, it initialized t every ime ClassBrowser::ThreadedBuildTree(..) m_ClassBrowserBuilderThread->Init(...) //But there the member var m_CCTreeTop is always recreated: ClassBrowserBuilderThread::Init(...)...
-
I tried to debug into a case from Test2 where the tree is closed on saving a header. What I see is the following sequence //entry is here and almost always the build tree is called ClassBrowser::UpdateClassBrowserView(..) ThreadedBuildTree(activeProject); But while this builds the tread the first time, it initialized t every ime ClassBrowser::ThreadedBuildTree(..) m_ClassBrowserBuilderThread->Init(...) But there the member var m_CCTreeTop is always recreated: ClassBrowserBuilderThread::Init(...)...
-
I tried to debug into a case from Test2 where the tree is closed on saving a header. What I see is the following sequence //entry is here and almost always the build tree is called ClassBrowser::UpdateClassBrowserView(..) ThreadedBuildTree(activeProject); But while this builds the tread the first time, it initialized t every ime ClassBrowser::ThreadedBuildTree(..) m_ClassBrowserBuilderThread->Init(...) But there the member var m_CCTreeTop is always recreated: ClassBrowserBuilderThread::Init(...)...
-
I tried to debug into a case from Test2 where the tree is closed on saving a header. What I see is the following sequence //entry is here and almost always the build tree is called ClassBrowser::UpdateClassBrowserView(..) ThreadedBuildTree(activeProject); But while this builds the tread the first time, it initialized t every ime ClassBrowser::ThreadedBuildTree(..) m_ClassBrowserBuilderThread->Init(...) But there the member var m_CCTreeTop is always recreated: ClassBrowserBuilderThread::Init(...)...
-
but I really do not want to tackle the situation when there is no open file of the project and the tree gets updated to a "closed" state. I tested that and it seems to work quite conistently. When switching projects the project tab makes it very easy to pick any file from the project and then switching to the Symbols tab will load the new tree. I think as procedure that is a very good workaround, but the point it how to know that this is the procedure. Maybe some Initial "don't show any more" Dialog...
-
The problem shows on my system Linux Ubuntu/xfce (18.04 and 22.04 shows the same behaviour). wxWidgets is: 3.0.5
-
I am using it for a while not and its speeding up things a lot. There is one thing I failed to track down, because I got lost in the code complexity a bit. When pasting an item, e.g. after copy, the current properties are not all maintained. For example the Border Settings are always reset to 5 and all sides ON. Same for Proportion that is set to 1. I think it would be better if all properties except the Varname and IDs remain as the source item.
-
Just noted that I missed one case. Context menu /"Find occurences of '..' "in the >standard editor seems to call another search function than the SeachButton in the >thead view window. The difference is that the SearchInComments setting is ignored. this is actually a feature I was not aware of: Thread Search OptionsDialog / "Use default options when 'find occurrences". Only thing that might be done is to add the 'Match Comments' = false to line 405 in /src/plugins/contrib/ThreadSearch/ThreadSearchConfPanel.cpp...
-
Just noted that I missed one case. Context menu /"Find occurences of '..' "in the >standard editor seems to call another search function than the SeachButton in the >thead view window. The difference is that the SearchInComments setting is ignored. this is actually a feature I was not aware of: Thread Search OptionsDialog / "Use default options when 'find occurrences" Only thing that could be done is to add the 'Match Comments' = false to line 405 in /src/plugins/contrib/ThreadSearch/ThreadSearc...
-
<moved to="" #1436=""></moved>
-
Just noted that I missed one case. Context menu /"Find occurences of '..' "in the standard editor seems to call another search function than the SeachButton in the thead view window. The difference is that the SearchInComments setting is ignored. To reproduce: * do a search that finds items in comments and normal code * Unselect in ThreadSearch Window/OptionButton/"Match in C++ style comments" * search -> should not list findings in comments * select the search item in the main editor and select...
-
Thanks. Is there anything expected from me now?
-
So according to the wiki, the AStyle option "ANSI" should do it, then, thanks I will try next time.
-
Hm. Interesting. I was sure this was already there as function with header but no CPP body but the compiler did not complain. In fact I added an implementation first but found that it was not called, so I deleted the CPP part again. Not sure why the compiler complains. So I did not use it and would have deleted it.
-
Thanks for checking out the patch. Attached a version w/o tabs. Am I correct that the patch format needs the tabs between the filename and the revision number. It would be easier if that could be spaces as well...
-
Thanks for your feedback. The "Cut2" was an leftover from debugging (my first implementation used wxSmith menu items but that proved much less efficient than the context menu. I'll correct that. The tabs. I had to change the patch manually, because it had loads of changes from my AStyle setting which seems very different than the code style CB svn is using. I think the editor tab preferences also affect hat. I have not found anything on the wiki about the settings to use for SVN commits. Obviously...
-
This was my first patch, so please bear with me if I got not everything right.
-
See patch #1437 for a working wxSmith tree context menu with copy/paste options. It has cut, copy, PasteBefore, PasteInto and PasteAfter.
-
See patch #1437 for a working wxSmith tree context menu with copy/paste options. It has cut, copy, PasteBefore, PasteInto and PasteAfter.
-
Patch: wxSmithTreeContextMenu CopyPaste
-
Took quite a while but now i managed to get a working compile environment and implemented the proposal. See patch #1436 The patch adds the option "Match In Comments". Its checked by default because then the behaviour is the same as today. It only works with C++ style comments "//" both full line and end of line. This info was put into the tooltip string. Tested in Linux 22.04 with wxWidgets3.0.5 only. i hope you find it a useful addition.
-
Patch ThreadSearch "MatchInComments"
-
Watch Window Context/Watch created invalid code
-
Yes I use that a lot. Afaik the menu the only way to get whole sections copied or moved around. As I tried to explain using the menu is very tedious when you need o change the copy options all he time. The mouse travel distance is really a lot.
-
I came across many of those cases while moving old dialogs from xrc to wxSmith. Bascally there is nothing I could not handle with Edit/Cut, Edit/Paste but its quite tedious. With the Ctrl-C and Ctrl-V working the mouse could stay on the left hand side to press the "inside" "before" "after" buttons and save quite a bit of time already. To get a context Up/Down would be the luxury version, of course...
-
Thanks for the hints and your time. I see what I can do.
-
Forgot to add Type:Feature_request. I do not have he right to edit the ticket (no edit button). sorry for that
-
[wxSmith] Usability Improvement Context Menu II
-
[wxSmith] Usability Improvement ContextMenu
-
Thread search: New Search option "Exclude comments"
-
OK, I sovled the problem. The repository did install the printer drivers only for gimp with the gutenprint packages. After I did sudo apt install printer-drivers-gutenprint the drivers also popped up in CUPS and then I could also see the parameters in the standard printer dialogs.
-
I am a bit lost with installing an Epson Styuls Pro 4880 printer on Ubuntu22.04. I want to use the Gutenprint driver for my printer for all applications, not only Gimp. Currently it seems I have it installed in Gimp but CUPS seems to be a different driver which confuses me. I have installed the "gutenprint" and "gimp-gutenprint" packages. Printing from Gimp seems to work fine. I would have the Gutenprint printer dialog in any any application (e.g. Inkscape) instread the default printer dialog - is...
-
Hello I use a Stylus Pro 4880 printer with eco-solvent inks under Linux with the Gutenprint printer driver. Normally printing graphics on non-paper material which works fine. But I would be interested to also print photos on standard paper as "Photo Luster", "Photo Semimatt" and "Photo HiGloss". A first test showed that the printer used too much ink for the paper so fine details were lost. Does anyone is using solvent inks with Gutenprints and what is your experience? If I need to make own ICC profiles,...
-
GUI: Size and margin seetings ignore integers
-
Looking at the code Ticket 49 has not been introduced. Its not clear why, though. Interestingly I never noticed that thead search was forgetting the current target or project. Since the patch is from @bluehazzard anyway hopefully he should know...
-
-
This is from a forum post about this issue. The log covers a Breakpoint set in CB but not removed in gdb. The GUI side, both Editor and BP Window work consistently (all BPs I added or removed in the editor are added or removed in the BP window). Only the gdb side gdb differs. I included "<======" where I think the key points are. I can not attach the program and I did not manage to reproduce it with something small yet. The trace action was: The program starts and hits one specific BP. User selects...
-
++
-
I think is a good idea to collect wxSmith feature ideas to a single ticket. 8) Clone dialogs and panels I think that is a great feature, but with some complexity. When refactoring code I am missing that one very often. I think involved in cloning would be those five items: .cbp file, .wxs file, .cpp and .h file and Classname. Maybe its hard to make it foolproof, e.g. when working with existing code files as one cpp/h may contain several classes. But maybe thats not really needed if its clear what...
-
++
-
I think that is a great solution and a fully transparent change that does not conflict with the current look&feel.
-
Yeah, actually I had been thinking about suggesting that. But honestly I have been using solutions like that but I did not because you do not have focus all the time in TS. When you just switch tabs and come back or switch to the editor and back, wouldn't it be weird to have the focus on that line? I would not see a "selected" line as default, i.e. when I am have the TS tab open but I am not searching anything but coding in the editor there is a high risk of mis-click. Ether you want to change the...
-
Yeah, actually I had been thinking about suggesting that. But honestly I have been using solutions like that but I did not because * you do not have focus all the time in TS. When you just switch tabs and come back or switch to the editor and back, wouldn't it be weird to have the focus on that line? * I would not see a "selected" line as default, i.e. when I am have the TS tab open but I am not searching anything but coding in the editor * there is a high risk of mis-click. Ether you want to change...
-
I am not sure that two "collapse" context menu settings are confusing, maybe an option under the "Logger Typ" is an alternative. Uncollapsing Is nothing I have missed yet, but it makes sense to have that as well.
-
Thread Search context/CollapseAll in ListSTC Type
-
New ThreadSearch Button: Re-Do Last Search
-
I installed V12765 and tested it. In large projects crashes are a lot better. CB does not crash as regulary, but it still does. I checked the procedure above (same environemt) and it still crashes quite consistently. The test had Setting CC/UpdateParseWhenTyping OFF and ON. Same result. So I guess the fix was the right thing, but still there might be some situations
-
Sorry: Forgot to change the title (I see no edit button to modify it). The request Is about 1 button, not 2. I initially had a second button but found a workaround for that... If there is a way to edit a post, I would appreciate a hint.
-
Thread Search 2 New buttons
-
Just some feedback for people reading this later. I have been using the setting "listSTC" for the Trread Search in Programs with many classes (>100) for about 3 weeks. My personal impression was that the usabiliy of the tread search was improved compared to the default view. I also found that it took a couple of days to get comfortable with the "interrupted look & feel of the list, so its good that the default setting is as it is today. However this is outweighted by far by the ability to fold single...
-
@ollidbg seems you are implying that the token tree access functions are not thread safe?
-
Indeed, never noted that second option tab. Knowing these options is a great find and they offer interesting advantages. Maybe a button to switch between them would bring them to the attention of many more users. I will try to switch to "listSTC" and see how that works. I would miss the sorting feature of the list but for long file paths that is a lot better display. Thanks a lot!
-
Didn't check on Windows. Even it that would be behaving differently it would be only half a solution. Do you know that wxGrid would behave differently or is it since it allows owner drawing?
-
Unfortunately it does not work as intended. The colum is right aligned, but when the column is resized and gets smaller than its content, the text always starts at the begin which fails to deliver the desired effect The line column that has been right aligned before has the same behaviour. Maybe this is rather a wxListCtrl issue than a CodeBlocks issue, because its counter-intuitive from what the user expects?
-
Fantastic. Thanks for sharing!
-
Right-aligned FilePaths in ThreadSearch
-
This is the Crash XML that was produced. It is from a non debug version, the call stack is not as complete as the ne provided above.
-
C::B crashes in Linux in a minimal HelloWorld program
-
I found the source of the problem. It was a hardware issue. The Linux system had a PCIe USB3.0 card hat caused the problem. I swapped it for a new USB3.0 PCIe card and I could open the veracrypt volume like in windows. This is very interesting, because any other UBS drives and sticks worked fine with that adapter. The error "Bad superblock" only appeared with veracrypt. I did not expect that this kind of behaviour could be caused by faulty hardware. Thanks Tiger PS: I still would be interested to...
-
@idrassi All these elements show that nothing can be done on the software side since this is really a communication issue with the disk for which I have no explanation. Is it possible for you to put the disk in different encluser and see if it works better? A test that you can do is to use the cross-platform application wxHexEditor to read the content of the first sector of the encrypted partition on both Linux and Windows. If my analysis is correct, you will not get the same value. If official Linux...
-
A test that you can do is to read the content of the first sector of the encrypted partition on both Linux and Windows. If my analysis is correct, you will not get the same value. I tried to do that. Worked on Linux. Failed on Windows. How to read the 1st sector of an unmounted disk in windows? To mount it I would need to run VeraCrypt. I checked Linux against the veracrypt mounted ones and they are totally different. Does that makes sense? I have the same issue as the OP with a standard WD Passport...
1 >
