User Activity

  • Posted a comment on discussion Open Discussion on SASHIMI

    OK, I tried checking out the code from the SVN trunk. The build process ended with the following: /home/calvin/tpp-svn/trunk/trans_proteomic_pipeline/build/linux-x86_64/artifacts/pwiz/pwiz_tools/commandline/msconvert.o: In function _ZNK5boost15program_options11typed_valueIbcE4nameB5cxx11Ev': /home/calvin/tpp-svn/trunk/trans_proteomic_pipeline/extern/ProteoWizard/pwiz-src/libraries/boost_1_56_0/boost/program_options/detail/value_semantic.hpp:19: undefined reference to_ZN5boost15program_options3argB5cxx11E'...

  • Modified a comment on discussion Open Discussion on SASHIMI

    I am able to get 5.1.0 to consistently build now. This is with a nearly stock version of Centos 7.5 (standard development files, etc.) The only changes from the stock version: 1) I built gcc 5.4.0, then used that for compilation 2) I edited /usr/include/bits/mathcalls.h, commenting out the following line, due to a definition conflict: MATHDECL_1 (int,isnan,, (Mdouble value)) attribute ((const)); 3) I edited common.mk, adding "CXXSTD := -std=c++11" Since 5.1.0 is what we actually wanted, I think I'll...

  • Posted a comment on discussion Open Discussion on SASHIMI

    I am able to get 5.1.0 to consistently build now. This is with a nearly stock version of Centos 7.5 (standard development files, etc.) The only changes from the stock version: 1) I built gcc 5.4.0, then used that for compilation 2) I edited /usr/include/bits/mathcalls.h, commenting out the following line, due to a definition conflict: MATHDECL_1 (int,isnan,, (Mdouble value)) attribute ((const)); 3) I edited common.mk, adding "CXXSTD := -std=c++11"

  • Posted a comment on discussion Open Discussion on SASHIMI

    I made the suggested changes to common.mk and Jamroot.jam, using the latest from the trunk. I still get boost link errors: g++ -L/home/calvin/trans_proteomic_pipeline/build/linux-x86_64/lib -o /home/calvin/trans_proteomic_pipeline/build/linux-x86_64/bin/m sconvert /home/calvin/trans_proteomic_pipeline/build/linux-x86_64/artifacts/pwiz/pwiz_tools/commandline/msconvert.o /home/calvin/trans_proteomi c_pipeline/build/linux-x86_64/lib/libpwiz.a \ -L/home/calvin/trans_proteomic_pipeline/build/linux-x86_64/lib...

  • Posted a comment on discussion Open Discussion on SASHIMI

    Has anyone in this forum been able to build TPP 5.1.0 in a Red Hat/Centos 6 or 7 system?

  • Posted a comment on discussion Open Discussion on SASHIMI

    That line is already present in the common.mk which came with TPP 5.1 source. [calvin@centos64devel ~]$ grep -n BOOST common.mk 301:BOOST_CXXFLAGS := -DBOOST_NO_CXX11_SCOPED_ENUMS Can I simply substitute the common.mk file you referred to? Or does that require the use of all the code currently in the repository?

  • Posted a comment on discussion Open Discussion on SASHIMI

    I'm still getting the boost error. Is this because the version installed in the system (1.41) is too old? [calvin@centos64devel TPP_5.1.0-src]$ make all g++ -L/home/calvin/TPP_5.1.0-src/build/linux-x86_64/lib -o /home/calvin/TPP_5.1.0-src/build/linux-x86_64/bin/msconvert /home/calvin/TPP_5.1.0-src/build/linux-x86_64/artifacts/pwiz/pwiz_tools/commandline/msconvert.o /home/calvin/TPP_5.1.0-src/build/linux-x86_64/lib/libpwiz.a \ -L/home/calvin/TPP_5.1.0-src/build/linux-x86_64/lib \ -lboost_system -lboost_program_options...

  • Posted a comment on discussion Open Discussion on SASHIMI

    There was a conditional setting of that line if the user is running GCC 5.2.0. I have 5.1.0 and 7.2.0 available, so I simply set it outside of that conditional statement. I then got an error about a redefinition of "isnan" (in system it's just plain 'isnan", while in TPP code it's "std::isnan"). That error occurred with 5.1.0, but not with 7.2.0, so I'm using that now. Thanks for your help!

View All

Personal Data

Username:
caldodge
Joined:
2018-07-27 17:28:28

Projects

  • No projects to display.

Personal Tools