I installed Code::Blocks on Ubuntu 9.10, but when I start it from terminal, it shows the following error:
relocation error: /usr/lib/libcodeblocks.so.0: symbol _Z18wxSafeConvertWX2MBPKw, version WXU_2.8.2 not defined in file libwx_baseu-2.8.so.0 with link time reference
What is wrong with it? It seems to me I have installed all necessary libraries. Any ideas?
P.S. I am a Linux noobie - so sorry if it is some kind of a well-known mistake.
The most likely problem is that libcodeblocks.so.0 was (apparently) linked against libwx_baseu-2.8.2.so, but the libwx_baseu you have installed is an earlier version 2.8.0.
Try locating libwx_baseu-2.8.2.so or some such and installing that.
Related
I am trying to compile a c++ code in an conan enviroment, here are details:
Ubuntu 18.04 WSL x86-64
Windowns 10 x64
python 3.7.5
protoc 3.19.4
The project is located at /mnt/c/project
The build directory is located at /mnt/c/build
The build occurs without any errors but when I try to compile with "ninja" it throws an error related to protoc-gen-nanopb, as you can see attached image.
I tried to install a new protobuf from its c++ source, i added the LD_LIBRARY_PATH that was missing, and I have also tried to clean everything and erase all the files and build again. Everything seems to fail. I believe it's related to compatibility between protobuf and my python version, but I am not sure.
Could someone help me understand how to solve it? I am new to ubuntu and stackoverflow and i am sorry if some information is missing. In this case, just ask.
Error 127
I am new here and this is my very first question. I started working on deep learning recently and decided to use caffe. I went through the steps as explained in the caffe tutorial installation: installed CUDA, cuDNN, installed OpenBLAS, the other libraries and so on. I am using Ubuntu 14.04. After that I downloaded caffe and tried to compile it. Everything seemed fine until I got this error message:
.build_release/lib/libcaffe.so/ undefined reference to boost::thread::join()
I know this is a linking problem. I installed the boost libraries from the terminal with sudo apt-get install libboost-all-dev. I have been looking around for a while, but I have not found anything specific to help me. What exactly do I need to do: create some soft-link (seemed like one of the main ways to solve this problem), change my Makefile (I already fixed it so that it looks where the boost library is) or something else?
Thanks in advance
Lately I have been messing around with OpenShift and have had some problems with my project. I have a normal Python 2.7 cartridge installed and want to use Selenium with PhantomJS but any binaries of PhantomJS I have found didnt work (Ghostdriver wouldn't work). So I installed RHEL 7 and compiled it myself, which worked perfectly on my machine but once uploaded it failed to execute.
I think this was because of missing libraries so I copied the necessary libraries and set LD_LIBRARY_PATH to where I put them. But now I've gotten a new error:
[...]/phantomjs: relocation error: [...]/lib/libc.so.6: symbol _dl_starting_up, version GLIBC_PRIVATE not defined in file ld-linux-x86-64.so.2 with link time reference
Does anyone have any ideas?
I compiled the version 1.9.8 on CentOS 6.6 x86_64 with a patch to make GhostDriver play nice with OpenShift. If you think it may help, you can find the binary here.
I searched around a fair amount and didn't see anyone else with this specific problem. I'm running Ubuntu 14.04 (amd64) and codeblocks 13.12 (from the ubuntu repositories). I've followed the directions here: "wxWidgets how to cross compile an application for windows from linux using codeblocks?" and am finding myself a bit stuck. I can successfully compile win32 CLI code and code that uses windows.h. If I create a wxwidgets project I can successfully compile it for linux, but if I try and compile it using the mingw compiler setup (per the instructions in that link) I get the following error:
unrecognized command line option "-Wno-unused-local-typedefs"
Any suggestions would be greatly appreciated!
Just remove this option, you're using a compiler too old to have it. As all -Wno-xxx options, this one just suppresses a warning, so removing it is harmless.
I'm getting error while trying to install sqlite3-ruby gem:
gem install sqlite3-ruby-1.3.1.gem
Building native extensions. This could take a while...
ERROR: Error installing sqlite3-ruby-1.3.1.gem:
ERROR: Failed to build gem native extension.
/usr/bin/ruby.exe extconf.rb
Gem files will remain installed in /usr/lib/ruby/gems/1.9.1/gems/sqlite3-ruby-1.3.1 for inspection.
Results logged to /usr/lib/ruby/gems/1.9.1/gems/sqlite3-ruby-1.3.1/ext/sqlite3/gem_make.out
Same using:
gem install sqlite3-ruby-1.3.1
And that worked:
gem install sqlite3-ruby-1.3.1-x86-mswin32-60.gem
However I discovered I had had installed sqlite-ruby-1.2.5
Well my question is how 'gem' is installing these things? Why it can't choose sqlite3-ruby-1.3.1-x86-mswin32-60.gem version automatically. Does 'gem' is aware of running on linux or windows? What is native extension for it and why it is failing to install predownloaded sqlite3-ruby-1.3.1.gem
Thanks in advance
Lots of people seem to be having this issue. I'm surprised this doesn't work out of the box, as I would have guessed cygwin+ruby+sqlite is a supercommon configuration. Anyway. Many people concentrate on sqlite3.h, which is what appears are missing in the error output. However, my problem was that gcc wasn't installed correctly. Apparently gcc can fail to install correctly under cygwin. I fixed that and it was fine. Also, make sure you install libsql3-devel in cygwin.
I've been down this road before. I failed installing Ruby in Windows to work with Cygwin and I failed on Ubuntu.
What you do to save yourself the time and trouble is get a VM running with a Linux image (Fedora seemed to work well) and work with Ruby in that.