It looks like you're new here. If you want to get involved, click one of these buttons!
Matthias,
In the file antConfigPage.cc, there is :
#include "RulerConfigPage.h"
#include "RulerConfigPage2.h"
#include "RulerConfigPage3.h"
#include "RulerConfigPage4.h"
But those files do not exist, there are only RulerConfigPagexxx.ui
The compilation works thanks to the option MM
I tried to rename those files in xxx.ui and to comment them, but they are needed.
Can you explain this ?
Thanks,
OkGuy
Comments
Hi okguy,
these files are generated from the .ui files in the Makefile using Qt's uic tool. They are generated in the object directory and cannot be found in the source tree. If they do not exist, the uic step failed for some reason. Please check whether uic is in your path and try a complete rebuild by removing the build folders.
Matthias
Thank you Matthias.
I noticed that I was not using Makefile.rules
BTW, I am compiling with CodeBlocks but I cannot find in your scripts how MakeFiles.rules is called ?
I have re-build a Makefile.env (STLPORT, QT64DIR and RUBY64 are defined elsewhere):
Other question : as the coordinates are defined as float , would these options help to improve Klayout speed ?
BRgds,
OkGuy
Hi OkGuy,
the build script will use a top level Makefile which just includes "Makefile.env", "Makefile.body" and "Makefile.rules".
I cannot support every kind of build system. I'd love to support qmake and QtCreator (which in my perception is one of the best lightweight IDE's for C++). Right now I found it too tedious to configure all the smartness like detecting Ruby installation into .pro files so that has not proceeded further. The build script is a pragmatic approach to cover the Linux part and for Windows there is a Visual Studio solution.
BTW: it's not true that coordinates are double's. For the layout representation and the majority of layout processing KLayout uses 32bit integers (or 64bit if you configure that option). Only upon drawing or on transformation into a double coordinate space the shapes become versions using floating-point coordinates. The reason for having integer coordinates is not performance (there is no performance advantage of integers over floating types today) but avoiding rounding issues that are inherent in floating point arithmetic. You can try the tuning options but I doubt they give a noticeable performance boost.
Regards,
Matthias