Get quick announcements through the RSS feed http://www.codeblocks.org/nightly/CodeBlock_RSS.xml
Before you use a nightly make sure you understand how it works (http://forums.next.codeblocks.org/index.php/topic,3232.0.html).
A link to the unicode windows wxWidget dll for Code::Blocks : http://prdownload.berlios.de/codeblocks/wxmsw28u_gcc_cb_wx2812_gcc471-TDM.7z
For those who might need this one (when no MingW installed on your system) : the mingw10m.dll : http://prdownload.berlios.de/codeblocks/mingwm10_gcc471-TDM.7z
And the exception handler dll (for better crash reports) : http://prdownload.berlios.de/codeblocks/exchndl_gcc471-TDM.7z
The 01 October 2013 build is out.
- Windows :
http://prdownload.berlios.de/codeblocks/CB_20131001_rev9378_win32.7z
- Linux :
none
Resolved Fixed:
- find-replace-dialog: fix shortcut (was used twice)
- debugger: Improve GetEditorWordAtCaret to work correctly for non-rectangle selections (thanks Huki)
- wxSmith-plugin: immediately show changed size (after adding first image) and image count in wxsimagelisteditordkg, disable del-, clear- and save-buttons if list is empty
- wxSmith-plugin (wxsimagelisteditordlg): expand image labels, so larger numbers get not cut off; add button to save a single bitmap instead of only sving the list as one large bitmap; many differences come from autogeneration of the source files with actual wxSmith
- SDK: introduce CodeBlocksThreadEvent which is thread safe to carry wxString cross threads, also remove the mutex in CClogger class, as we don't need them. see the discussion: http://forums.next.codeblocks.org/index.php/topic,18059.0.html
- manager: Don't translate the name of all cbEVT_XXX types (thanks gd_on)
- wxSmith-plugin: fix layout issue; only the top-sizer with the title of wxsImageTreeEditorDlg was shown
- wxPropgrid: correct handling of composite flags in wxFlagsProperty; backport of my patch for wxWidgets, see: http://trac.wxwidgets.org/changeset/74808
- wxSmith- and wxSmithContribItems-plugin: correct handling of composite-flags in codegeneration of wxsRichTextCtrl, wxsRichTextFormattingDialog, wxsRichTextStyleOrganiserDialog and wxsChart
- debugger: Fix a crash when the locals or func args are disabled during debugging session
- debugger: Refresh watches, when locals or func args are added during debugging sessoin
- ui: Remove the shortcut for hiding the editor tabs, because it causes too much complaints in the forum
- CC: handle macro definition correctly, distinguish between function like macro definition and variable like definition (thanks Huki)
- CC: correctly find the end of a c style comment (thanks Huki)
- CC: only skip after single equal sign, not double equals sign (thanks Huki)
- CC: when reading the parentheses, set the correct Tokenizer mode(tsReadRawExpression), also skip comments after the first '(' (thanks Huki)
- ThreadSearch-plugin: initially load search history in toolbar combobox
- ThreadSearch-plugin: really remove last item of toolbarcombo, if max-count is reached
- ThreadSearch-plugin: use right-click event instead of menu event to invoke context-menu of logger-tree, this avoids annoying pop up of the infopanes context menu after the tree contextmenu was closed (happened at least with wxGTK)
- ThreadSearch-plugin: do not skip delete-event, to avoid double invocation of it, fixes an assert and (sometimes) crash with wx2.9+ (at least with wxGTK)
- ThreadSearch-plugin: immediately return from loggert-tree click event, if the tree was cleared; avoid annoying pop up of "Failed to retrieve file path and line number" messagebox after leaving configuration dialog (happened at least on wxGTK)
- pumped wxPDFDoc to 0.9.4 (includes many changes related to wx29)
- pumped (wx)scintilla to v3.3.5
- pumped astyle lib to v2.0.4 final
- applied patch #3478: Add line breaking option to AStyle
- applied patch #3480: Fix menu entry for scripts are not updated
- ui: Allow the user to change the colours of the linenumbers and the margin in the editor
- info_window: Improve multi monitor support
- CC: reliable working of UngetToken() when macro expansion is involved, it set the undo index value correctly(thanks Huki)
- CC: avoid take backward step (UngetToken) twice in the Tokenizer class
- colours: Use the proper constant for start page's background, wxSYS_COLOUR_BACKGROUND returns the destop's colour on windows (related to r9361)
- colours: Use wxSYS_COLOUR_WINDOWTEXT colour instead of just black for the text colour
Regressions/Confirmed/Annoying/Common bugs:
wondeful! :D
That's a lot of fixes, congrats!!
Hi,
the title of the post should be with svn number 9378, not 9295.
gd_on
Does code completition take into account the 'using' statements? I'm guessing it doesn't because cc doesn't kick in when I type 'cout.' if I use 'using std::cout;' but it does kick in when I write 'using namespace std;' or 'std::cout.'. I just wonder if that's a bug or a limitation.
CC handles using directives very poorly, unfortunately.
How long c::b uses gtk3?
Quote from: p2rkw on October 02, 2013, 10:06:35 PM
How long c::b uses gtk3?
Hm, you'll have to try again, because it is not clear what are you asking.
When I go to properties on a file and tick "Use custom command to build this file" in Advanced tab, it will automatically tick "File is read-only" in General tab. Is this an intended behavior?
Quote from: rathapolk on October 07, 2013, 08:35:29 PM
When I go to properties on a file and tick "Use custom command to build this file" in Advanced tab, it will automatically tick "File is read-only" in General tab. Is this an intended behavior?
If the file is read only (Operating System Setting), yes it is the intended behavior.
Tim S.
Quote from: stahta01 on October 07, 2013, 09:00:56 PM
Quote from: rathapolk on October 07, 2013, 08:35:29 PM
When I go to properties on a file and tick "Use custom command to build this file" in Advanced tab, it will automatically tick "File is read-only" in General tab. Is this an intended behavior?
If the file is read only (Operating System Setting), yes it is the intended behavior.
Tim S.
I am also experiencing this and unfortunately the files are not read only. I think that's a bug.
It is a bug...
Should be fixed in trunk.
What about #include filenames completion? This feature seems to work sometimes and sometimes it doesn't.
Assume I want to include the file at "../dialogs/CustomDialog.h"
So, I type:
#include "../d" (note the closing quote mark)
and here comes CC suggesting "../dialogs/CustomDialog.h". I press 'enter' and get
#include "../dialogs/CustomDialog.h"" (extra quote mark)
or even
#include "../d../dialogs/CustomDialog.h" (my input wasn't overwritten)
But sometimes headers completion works perfectly.
Should this be reported as a bug?
fIf you can provide simple project and the exact steps to reproduce it - the yes report it as a bug.
1. create the new wxWidgets project. Let's name it 'completion_test'.
- open the generated completion_testMain.cpp, place the cursor next line to #include <wx/msgdlg.h>
- type #include ""
- move the cursor between two quote marks and start typing comple
- when the completion variants show up select any and press enter. You'll get an extra quote mark at the end of the line.
2. create a new folder in project directory. Place any header file into it.
- open the generated completion_testMain.cpp
- type #include "folder/
- select your header file from the completion variants. You'll get something like #include "folder/"folder/header.h"
OS: win7 64
wxWidgets 2.9.5
code::blocks rev. 9378
Under linux and wx28, 1. confirmed, 2. cannot reproduce.
Can you try applying this patch?
Index: src/plugins/codecompletion/codecompletion.cpp
===================================================================
--- src/plugins/codecompletion/codecompletion.cpp (revision 9398)
+++ src/plugins/codecompletion/codecompletion.cpp (working copy)
@@ -1562,7 +1562,6 @@
{
--startPos;
}
- const int endPos = control->WordEndPosition(curPos, true);
bool needReparse = false;
if (control->IsPreprocessor(control->GetStyleAt(curPos)))
@@ -1606,7 +1605,7 @@
}
needReparse = true;
- int pos = startPos;
+ int pos = startPos - 1;
wxChar ch = control->GetCharAt(pos);
while (ch != _T('<') && ch != _T('"') && ch != _T('#') && (pos>0))
ch = control->GetCharAt(--pos);
@@ -1618,11 +1617,14 @@
else if (ch == _T('<'))
itemText << _T('>');
}
+ else
+ {
+ const int endPos = control->WordEndPosition(curPos, true);
+ const wxString& alreadyText = control->GetTextRange(curPos, endPos);
+ if (!alreadyText.IsEmpty() && itemText.EndsWith(alreadyText))
+ curPos = endPos;
+ }
- const wxString alreadyText = control->GetTextRange(curPos, endPos);
- if (!alreadyText.IsEmpty() && itemText.EndsWith(alreadyText))
- curPos = endPos;
-
control->AutoCompCancel();
int positionModificator = 0;
Thank you for the reply.
I use precompiled binaries for code::blocks, so it would take me some time to check the patch.
I'll try to download and compile sources tomorrow at night, still can't promise that :(
I've seen both of them... ::)
Quote from: Alpha on October 13, 2013, 05:47:35 PM
Can you try applying this patch?
[...]
I have not noticed problems with this patch, and will commit soon if there are no further comments. (I would appreciate if someone who has observed problem 2. can report whether or not this patch resolves it.)
After updating from 9158 codeblocks drop all compiler settings.
It's too sad, will stay on 9158.
what do you mean exactly? I'm on 9393 and it's working fine* :P
(and I've got lots of custom but still mandatory compiler settings... -m32, -m64, -flto, -static, -fPIC etc.)
*besides those nasty crashes every C::B build with code-completion includes^^
Ahem. Greetings all, I have an announcement to make. I can get C::B to consistently crash by opening the watch properties while a watch is being evaluated. I'm not sure if it's just me, but If anyone is interested in the steps to reproduce said phenom, or would just like to observe the action, I will be making a demo video sometime today. That is all.
Quote from: Static on October 13, 2013, 11:42:46 AM
What about #include filenames completion? This feature seems to work sometimes and sometimes it doesn't.
Assume I want to include the file at "../dialogs/CustomDialog.h"
So, I type:
#include "../d" (note the closing quote mark)
and here comes CC suggesting "../dialogs/CustomDialog.h". I press 'enter' and get
#include "../dialogs/CustomDialog.h"" (extra quote mark)
or even
#include "../d../dialogs/CustomDialog.h" (my input wasn't overwritten)
But sometimes headers completion works perfectly.
Should this be reported as a bug?
I have a similar issue with header files. For example, if I have typed
<ios> by mistake and take my cursor inside header name, and continue typing from
's', it will pop up options like
iosfwd and
iostream. When I choose the latter by pressing the tab key, it automatically adds an extra
> character.
Is this an expected behavior, or a bug?
Cheers.
For all it's worth, to me, that's an annoying bug(feature?). You eventually learn to avoid it.
Quote from: CrispiestMushroom on October 21, 2013, 07:31:23 AM
Ahem. Greetings all, I have an announcement to make. I can get C::B to consistently crash by opening the watch properties while a watch is being evaluated. I'm not sure if it's just me, but If anyone is interested in the steps to reproduce said phenom, or would just like to observe the action, I will be making a demo video sometime today. That is all.
please make a separate topic, and please give us the full debug log (you have to enable it in the optiions) also a procedure to reproduce it would be nice.
greetings
Quote from: ToApolytoXaos on October 21, 2013, 08:21:03 AM
I have a similar issue with header files. For example, if I have typed <ios> by mistake and take my cursor inside header name, and continue typing from 's', it will pop up options like iosfwd and iostream. When I choose the latter by pressing the tab key, it automatically adds an extra > character.
Should already be fixed in trunk (as of rev. 9401).
Quote from: Alpha on October 21, 2013, 02:16:25 PM
Should already be fixed in trunk (as of rev. 9401).
As of rev. 9414, the aforementioned problems got solved.
Thank you very much Alpha et al. for your valuable work and feedback :)
Quote from: CrispiestMushroom on October 21, 2013, 07:31:23 AM
Ahem. Greetings all, I have an announcement to make. I can get C::B to consistently crash by opening the watch properties while a watch is being evaluated. I'm not sure if it's just me, but If anyone is interested in the steps to reproduce said phenom, or would just like to observe the action, I will be making a demo video sometime today. That is all.
Can you reproduce this problem with a minimal-sample project and then provide the steps needed to do it, so I can take a look at it?
On openSUSE 12.3 and higher cannot build RPM package with error.
Quote
[ 1981s] W: codeblocks strict-aliasing-punning ../../../../src/include/scripting/sqplus/sqplus.h:420, 1801
[ 1981s] W: codeblocks strict-aliasing-punning sqcompiler.cpp:655, 701
[ 1981s] W: codeblocks strict-aliasing-punning sqstdblob.cpp:161, 169
[ 1981s] W: codeblocks strict-aliasing-punning sqstdsystem.cpp:55
[ 1981s] W: codeblocks strict-aliasing-punning sqvm.cpp:740
[ 1981s] W: codeblocks strict-aliasing-punning src/pdfencrypt.cpp:188, 189
[ 1981s] W: codeblocks strict-aliasing-punning src/pdfrijndael.cpp:1088, 1089, 1090, 1091, 1096, 1109, 1110, 1111, 1112, 1118, 1119, 1120, 1121, 1182, 1235, 1236, 1237, 1238, 1243, 1244, 1245, 1246, 1251, 1252, 1253, 1254, 1264, 1265, 1266, 1267, 1273, 1274, 1275, 1276, 1345, 1356, 1396, 1407, 1430, 1435, 1443, 1450, 1469, 1471, 1473, 1475, 1484, 1485, 1486, 1487, 1488, 1489, 1490, 1491, 1492, 1493, 1494, 1495, 1496, 1497, 1498, 1499, 1500, 1501, 1502, 1503, 1506, 1507, 1508, 1509, 1511, 1512, 1513, 1514, 1515, 1516, 1517, 1518, 1519, 1520, 1521, 1522, 1523, 1524, 1525, 1526, 1528, 1529, 1530, 1531, 1548, 1549, 1550, 1551, 1559, 1560, 1561, 1562, 1564, 1565, 1566, 1567, 1568, 1569, 1570, 1571, 1572, 1573, 1574, 1575, 1576, 1577, 1578, 1579, 1582, 1583, 1584, 1585, 1586, 1587, 1588, 1589, 1590, 1591, 1592, 1593, 1594, 1595, 1596, 1597, 1598, 1599, 1600, 1601, 1604, 1605, 1606, 1607, 1624, 1625, 1626, 1627
[ 1981s] E: codeblocks 64bit-portability-issue bzlib.c:1431
I have a serious issue! I use 9145 in OpenSUSE. It always uses one core and does not want to exit completely. The UI disappears BUT the process exists and does not want to stop. It always uses a core!
Quote from: Grom on October 27, 2013, 10:57:36 PM
I have a serious issue! I use 9145 in OpenSUSE. It always uses one core and does not want to exit completely. The UI disappears BUT the process exists and does not want to stop. It always uses a core!
Install gamin-server it need for FileManager plugin.