Castalia building error when changing the version of omnet++ - omnet++

I have switched from omnet++5.3 to omnet++4.6. I am using Castalia 3.3.
My installation for the first version of omnet++5.3 was ok with castalia. But when I install the second version and build castalia with it, I have got errors. When I type the command which opp_makemake to verify the path, I got /home/hana/omnet-5.3/bin/opp_makemake instead of /home/hana/omnet-6.4/bin/opp_makemake.
I have changed the path and follow the installation guide for both omnet++ and castalia but still castalia wouldn't run correctly.
Any idea or direction that can help me is appreciated !
Thank you

I have found the error:
when adding the new path of the new version of installed omnet++4.6 in the .bashrc file, I should delete the older path of the other version of omnet++5.3 (even if it is still installed)
.bashrc doesn't accept two paths for the same software. That works !

Related

". setenv" error after cloning from Github: OMNeT++6.0pre10/11 repo

Context: I am currently working on a research project to run an end to end simulation of a 5G cellular network with simu5G. I need help with the installation of Omnet++6.0pre10(or pre 11) from the command line (I don't have access to a gui). My operating system is Ubutnu 18.04.5 LTS
I am new to Omnet++ and need help with installation of Omnet++6.0pre10 (or pre11) from the command line.
I have tried looking at the manual, but the latest version is always the 5.6.1 version:
https://doc.omnetpp.org/omnetpp/UserGuide.pdf
and the steps I have taken are:
forking from the Omnetpp github repository
cloning the forked repository to the local machine
But when I try the first step of . setenv it gives me the following error:
Error: '$dir' does not look like an OMNeT++ root directory
which is the error associated with if the directory is really pointing to an omnet++ installation dir (if you look at the bash code in setenv file)
Please help in:
Advise on how to fix this error with the setenv script
how to actually install omnet++6.0pre10(or 11) through the command line and how that differs to the latest user guide (user guide 5.6.1)
Best wishes and thank you so much
You have probably forgotten creating configure.user mentioned in Instalation description, i.e.:
cd omnetpp
cp configure.user.dist configure.user
Checking out the master version of OMNeT++ is not the best choice, especially if you are doing research. I would advise to use instead the omnetpp-X.y-src-core.tgz version which does not have the IDE. However also the omnetpp-x.y-src-linux.tgz version would do fine as long as you disable the OSG, OSG_EARTH and QTENV components in configure.user, before running ./configure

can not get singularity to compile, failed to get package version

Ubuntu 20.04, downloaded singularity 3.7.3.tar.gz, sha256 matches, unzipped it and followed instructions but keep getting
Failed to get package version. Abort.
Any quick suggestions what I'm missing. I installed GO using a snap, and when I check the version I get
$ go version go version go1.16.3 linux/amd64
What am I missing. Thx, J.
I found a bug / problem!
In the file
/singularity/mconfig
is some code to check the version?
# if test -z "${package_version}" ; then
# echo "Failed to get package version. Abort."
# exit 1
# fi
I can't tell why this doesn't work because the file it is looking for is present.
If I comment out this code, the config, compile and install work seamlessly and do produce the correct version information! The syslabs.io people make it incredibly difficult to contact them to suggest a bug.
V2, got it to work, this is new to me but the quirk was I was downloading the .tar.gz into a directory that was already within a git repo. This affects where mconfig looks for the "VERSION" file and causes an error. Created a new directory outside any git repo's, dowloaded, untarred and mconfig,make, and make install all worked fine. Strange thae where it looks for VERSION file is changed?

While applying ./configure command to install omnetpp, I am getting the message about not finding the C compiler in $PATH

I am installing OMNeT++. While executing the ./configure command on the mingwenv file I am getting the message:
no acceptable C compiler found in $PATH
Kindly help
Thanks for.showing the ray of light towards solution
I had already installed omnet pp 5.4.1 on my window 8.1. Every thing was fine . Due to some reason I deleted it . When I reinstalled , I am facing this problem that means I am not able to execute ./ configure command

Error in pkg-config when try to download git repository

I was trying to make a facial recognition app using Go and Kagami/go-face repository. It requires dlib-models. And I was used MSYS tool for get the distribution in windows.
But when I try to get the go repository using go get github.com/Kagami/go-face command and it gives an error:
# pkg-config --cflags -- dlib-1
Any one had this issue before in windows 10.
As commented, the README lists the Windows pre-requisites, involving MSYS2.
This was from commit 968bbf9, following issue 5 "Windows support?"
Another approach would be to use Docker, with an Ubuntu image, which would come with dlib, where you can add Go and experiment there.
I also faced the same problem. pkg-config.exe will be found in the path:
C:\msys64\mingw64\bin
if you installed MSYS2 on default path. Export
C:\msys64\mingw64\bin
this path to environment variable then you wont get that error.

Error because file libmpi.so.1 missing

I'm using Omnet++ and Veins for simulations and it was working fine until I upgraded from Ubuntu 15 to 16.04 LTS last night. Now, I get the following error when trying to run a simulation: error while loading shared libraries: libmpi.so.1: cannot open shared object file: No such file or directory
I searched for libmpi.so.1 and it seems it is missing. There is a file libmpi.so and libmpi.so.12 in /usr/lib/openmpi/lib but not libmpi.so.1. I tried uninstalling and reinstalling the packages openmpi-bin, libopenmpi-dev as well as OpenMPI which I downloaded from the website. I also set the variable in bashrc and profile (which was recommended online) with export LD_LIBRARY_PATH:=$PATH:/usr/lib/openmpi/lib/
None of these approaches worked and I still get the same error. Any suggestions on how to fix it and how to get the file libmpi.so.1?
I could fix the problem by adding a symbolic link libmpi.so.1 in /usr/lib that points to the existing libmpi.so.12 (which again points to the location of the actual file in openmpi/lib).
Apparently, it was a problem with the version of OpenMPI since on my other system, which I didn't update, there was a symbolic link libmpi.so.1 pointing to libmpi.so.1.0.8 (but no libmpi.so.12).

Resources