Hi Jack, Yeah I tried to copy into kwrite to remove any extra formatting. On Windows I use notepad for the purpose. When I originally pasted in there it was all out of shape. I thought I had fixed it. Sorry about the mess.
In any case, I believe as you do. That somehow the code I am seeing is not what is executing. I did build both ways. From within the kdevelop tool and from the command line. Neither helped. I kind of thought the c_flags was intended for C but I was not sure either. I did not execute from a console, I ran the cmake command and then the make command..then I went to the kdevelop and ran Debug. Do I need to 'install' it? Lol, I hope I didn't miss a step. Lol would explain everything. I am used to Java where we don't install, just debug from the editor. Is it the same here in c++ land. It has been a long time. Thanks, JV On Thu, Dec 26, 2019, 16:33 Jack <j...@ostroff.xyz> wrote: > Hi JV, > > On December 26, 2019 11:46 AM jvap...@gmail.com wrote: > > Thanks Jack! > > > > This is what I see in the ccmake: > > > > CMAKE_BUILD_TYPE Debugfull > > CMAKE_COLOR_MAKEFILE ON > > CMAKE_CXX_COMPILER /usr/bin/c++ > > CMAKE_CXX_COMPILER_AR /usr/bin/gcc-ar > > CMAKE_CXX_COMPILER_RANLIB /usr/bin/gcc-ranlib > > CMAKE_CXX_FLAGS > > CMAKE_CXX_FLAGS_DEBUG -g > > CMAKE_CXX_FLAGS_DEBUGFULL > > CMAKE_CXX_FLAGS_MINSIZEREL -Os -DNDEBUG > > CMAKE_CXX_FLAGS_RELEASE -O2 -DNDEBUG > > CMAKE_CXX_FLAGS_RELWITHDEBINFO -O2 -g -DNDEBUG > The -O2 shows optimization, which is not present for DEBUG. I forget > exactly what -Os does, but not sure if it's a problem or not. I believe > the -g is OK, and similar to the -ggdb I use. > > CMAKE_DLLTOOL CMAKE_DLLTOOL-NOTFOUND > > CMAKE_EXE_LINKER_FLAGS > > CMAKE_EXE_LINKER_FLAGS_DEBUG > > CMAKE_EXE_LINKER_FLAGS_DEBUGFU > > CMAKE_EXE_LINKER_FLAGS_MINSIZE > > CMAKE_EXE_LINKER_FLAGS_RELEASE > > CMAKE_EXE_LINKER_FLAGS_RELWITH > (Please be careful with copy/paste like that - I had to trim lots of > trailing spaces so it didn't look very badly wrapped.) I don't have any of > the CMAKE_CXX_... entries, but I'm using ninja instead of make, which might > account for that difference. Since KMM uses C++ and not C, the CMAKE_C_... > flags are not needed. The notation in my cmake line is from many years' > force of habit. > > > > I tried the new cmake line to see if that stops any optimization, but > truly, I don't know where to do that. > Not sure where to do what? I'd clear out the build directory and start > from scratch. Create and cd to a build directory (either under or parallel > to the base directory of the git checkout) and run the cmake command > there. You can run cmake right in the base dir of the source, but I have > found that doing that makes it hard to clean up a bad build and start > again, without messing up the source. > > > With what Jack said, I see those options in ccmake, but should I change > anything? DEBUGFULL has nothing as a value. What I don't see is a > CMAKE_C_FLAGS, not sure if that is missing or just not needed on my system. > As I mentioned above, lack of CMAKE_C_FLAGS should not be a problem as > there is no C code to compile. Missing the -g from DEBUGFULL should not > cause a problem, but I'm not sure what the default optimization level is > without an explicit -O0. You might be able to see it by looking at some > compile command withing the make output. > > > > It does seem that now the debuger Step-Over function in kdevelop works > better. So maybe it worked. I need to test it more. > > > > Still, I notice weird things, like in this image below. I stepped over a > boolean variable assignment, which sets it to true, but when I hover over > it, it says its value is false. The debug cursor is past the point where > the assignment was made. Why? Is this a kdevelop issue or maybe I just do > not understand how this tool works yet. I thought the hover over would > display the current value in the variable. > I don't use kdevelop much. but I'm wondering if the compiled code you are > actually running might be out of sync with the source code you are looking > at. Did you run the build from within kdevelop? (I don't know if it's > even possible to get out of sync like that, but something does seem odd.) > > > > Also, notice the qDebug() did not print on the application output. I did > set up my application launcher in kdevelop, and debug launcher as Thomas > showed in his screenshots. The only thing different was that I had an > application target set, where he had an executeable. I updated mine to > match his. > Did you check the console output later? I wonder if the output might have > been buffered. > > > > > > > > If anyone has an more insights, please let me know. > > > > I appreciate the time and assistance. > > > > thanks, > > > > JV > > > > > > On Thu, 2019-12-26 at 11:30 -0500, Jack wrote: > > > Pending further response from Thomas, I think the debugfull is better > for your situation. However, the actual control of optimization is in flags > to the compiler, usually set with CFLAGS or CXXFLAGS. Optimization is -Ox > where x is the level of optimization applied. For example, my cmake line > starts with 'CFLAGS="-O0 -ggdb" CXXFLAGS=$CFLAGS cmake ....' The -O0 > (capital letter o, digit zero) says no optimization at all. The -ggdb may > not be necessary for you, but at some point, helped me by explicitly > providing some additional debug information used by gdb. I am not sure > whether it makes any difference if you are debugging with kdevelop. > > > > > > Jack > > > > > > > > > On 12/26/19 11:12 AM, jvap...@gmail.com wrote: > > > > > > > > > > Hi Thomas, > > > > > > > > > > > > > > > > > > > > Thanks for the response. > > > > > > > > > > > > > > > > > > > > Where can I check this? > > > > > > > > > Did you build with optimization? Turn them off. > > > > > > > > > > > > > > > > > > > > My last cmake command was: > > > > > > > > cmake .. -DCMAKE_INSTALL_PREFIX=/usr > -DKDE_INSTALL_PLUGINDIR=/usr/lib64/qt5/plugins > > > > > > > > > > > > > > > > > > > > I will try the following: > > > > > > > > cmake .. -DCMAKE_INSTALL_PREFIX=/usr > -DKDE_INSTALL_PLUGINDIR=/usr/lib64/qt5/plugins > > > > -DCMAKE_BUILD_TYPE=Debugfull > > > > > > > > > > > > > > > > In case this is the problem, do you know what this is for then? > > > > > > > > > > > > > > > > > > > > The default value is: 'RelWithDebInfo' > > > > > > > > > > > > > > > > I read that as Release-with-Debug-and-Info. Is this still optimizing > for release? > > > > > > > > > > > > > > > > > > > > thanks, > > > > > > > > > > > > > > > > > > > > JV > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Thu, 2019-12-26 at 11:59 +0100, Thomas Baumgart wrote: > > > > > > > > > Jesus, > > > > > > > > > > > > > > > > > > > > > > > > > On Donnerstag, 26. Dezember 2019 08:54:38 CET jvap...@gmail.com > wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > Everyone... in case you have run into this before. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I cannot seem to get Kdevelop to work correctly. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Problems: > > > > > > > > > > > > - I set breakpoints but the breakpoint is set a few lines off. > > > > > > > > > > > > - When running in Debug mode, I am watching a variable named > > > > > > > > > > > > invertedValue which I hard coded to = true; and still the > variable does > > > > > > > > > > > > not set to true, as if it is not debugging the code I see in the > IDE. > > > > > > > > > > > > > > > > > > > > > > > > > Did you build with optimization? Turn them off. Optimization > causes all kinds of weird looking side effects when debugging. > > > > > > > > > > > > > > > > > > > > > > > > > > - I set qDebug() and qInfo() methods but they did not show > anything in > > > > > > > > > > > > the output view. Where should this show up? Anything I can > configure in > > > > > > > > > > > > the Kdevelop to get it to work? > > > > > > > > > > > > > > > > > > > > > > > > > Shows up in the Debug view. See attached screenshot: shows an > active breakpoint in the source code and the output in the Debug view. > > > > > > > > > > > > > > > > > > > > > > > > > > Any help would be greatly appreciated. > > > > > > > > > > > > > > > > > > > > > > > > > Also attached how I configured the launches in KDevelop > > > > > > > > > > > @Thomas, I am focusing on the Cashflow report. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > thanks, > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > JV > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Sun, 2019-12-22 at 08:51 +0100, Thomas Baumgart wrote: > > > > > > > > > > > > > Jesus, > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Samstag, 21. Dezember 2019 18:28:05 CET Jesus Varela wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Dev team, > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I am trying to become more familiar with the code and am > looking > > > > > > > > > > > > > > > > for the > > > > > > > > > > > > > > > > code that would control whether transfers are graphed and > displayed > > > > > > > > > > > > > > > > as > > > > > > > > > > > > > > > > negative numbers in the reports. I want to display absolute > values > > > > > > > > > > > > > > > > on > > > > > > > > > > > > > > > > charts and tables. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Not sure which report you want to modify, but there is e.g. a > method > > > > > > > > > > > > > > called > > > > > > > > > > > > > > PivotTable::coloredAmount(). It is responsible to render the > color > > > > > > > > > > > > > > for HTML > > > > > > > > > > > > > > reports. The callee of this method provides the value. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I figured I would use this personal desire to get acquainted > with > > > > > > > > > > > > > > > > the code. > > > > > > > > > > > > > > > > Any hints would be appreciated. I found the > kreportchartview.cpp > > > > > > > > > > > > > > > > and in > > > > > > > > > > > > > > > > there is a variable to figure out if expenses should be > negative or > > > > > > > > > > > > > > > > not, > > > > > > > > > > > > > > > > but I do not see anything for transfers. If it is there, I > will > > > > > > > > > > > > > > > > keep > > > > > > > > > > > > > > > > looking. If I am on the wrong file, please let me know if > you know > > > > > > > > > > > > > > > > where I > > > > > > > > > > > > > > > > should be looking. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > KReportChartView::drawPivotChart receives all values in the > parameter > > > > > > > > > > > > > > 'grid'. > > > > > > > > > > > > > > You may want to look into the construction of the grid to find > the > > > > > > > > > > > > > > spot where > > > > > > > > > > > > > > the values of the transfers are inserted. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > The spot you found is not negating the value of each expense, > but > > > > > > > > > > > > > > simply > > > > > > > > > > > > > > negates the value contained in the grid if the underlying > account for > > > > > > > the cell > > > > > > > is an expense account. So I wonder if that would be the spot to > > > > > > > achieve what > > > > > > > you are trying. > > > > > > > > > > > > > > > Just trying to make sure I am understanding the structure. I > > > > > > > > haven't used > > > > > > > > c++ in a long time. > > > > > > > > > > > > > > > > Thanks in advance for any assistance. > > > > > > > Hope that helps. The reporting section is not my area of > expertise. >
