OMNeT++ Wiki | Main / OMNeT++ 4.x

How To...

Java and Eclipse plug-in development

General

Tutorials

Developing simulation models

Debugging C++ code

Debugging with Tkenv

Debugging simulations with Tkenv (article on omnetpp.org)
Logging from modules (article on omnetpp.org)

Database integration

Using MySQL and PHP for web-based query of simulation results (article by Patrick Haeflinger)

Parallel distributed simulation

Misc

How to write regression tests for protocol models (article on omnetpp.org)
How to make Fast mode faster
How can I capture movies from my simulation?

Known Problems

Issue: On Windows 7, running ./configure in the MinGW terminal gave several of the following errors: rm: cannot lstat 'conftest.exe': Permission denied

Solution: This issue is tracked by http://dev.omnetpp.org/bugs/view.php?id=154 and happens when "application experience service" is disabled on a Windows 7 installation. See http://www.blackviper.com/wiki/Application_Experience for further details.

Issue: On some Linux systems, the IDE will not start; instead, an empty dialog titled "Omnetpp" appears after the splash screen, and it cannot be closed. The log file contains SWTError: XPCOM error -2147467262.

Solution: This is related to xulrunner (1.9 beta versions, possibly others); see https://bugs.eclipse.org/213194. xulrunner is part of Firefox; installing another version of Firefox (and making sure the wrong xulrunner version has been uninstalled!) should solve the problem. Resolved in 4.1

Issue: Sometimes, after opening an existing project, C++ sources are not recognized properly by the IDE (file icons depict contoured "h" and "c" instead of solid ones).

Solution: Close and re-open the project. This is likely an issue in CDT; we are investigating it. Resolved in 4.1

Issue: The simulation fingerprint can differ across machines or between valgrind and non-valgrind runs.

Solution: The value of the calculated fingerprint in a simulation is heavily dependent on the accuracy of the floating point arithmetic. There are differences between the floating point handling of AMD and Intel CPUs. By default GCC uses the i387 instruction set on intel processors and the SSE2 instruction set on AMD processors. Running under a processor emulator software like valgrind may also produce a different result. If your CPU supports SSE2 or later, add the -mfpmath=sse -msse2 options to CFLAGS_RELEASE and CFLAGS_DEBUG to get the same results on AMD, Intel and valgrind. See related bug report in valgrind: http://bugs.kde.org/show_bug.cgi?id=197915

To search or report issues, visit the OMNeT++ Bug Tracker