Make doesn't work for Allpaths LG

Question

[cybercat]

Hello, I want to install and use Allpaths LG on my laptop with Ubuntu 20 LTS. However, after "./configure" with "sudo make" it always gives me this error message:

make[1]: *** [Makefile:2654: CleanEfasta.o] Error 1

make[1]: Leaving directory '/home/yvonne/allpathslg-52488_/src'

make: *** [Makefile:284: all-recursive] Error 1

The log looks like that with almost every line:

./system/System.h:32:36: error: ambiguous overload for ‘operator<<’ (operand types are ‘std::basic_ostream<char>’ and ‘__pid_t’ {aka ‘int’})    32 |      cout << "\nFatal error (pid=" << getpid() << ") at " \
      |      ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^~ ~~~~~~~~
      |           |                                 |
      |           std::basic_ostream<char>          __pid_t {aka int} ./paths/long/Logging.h:35:16: note: in expansion of macro ‘FatalErr’   35 |           {    FatalErr( "Illegal logging " << x << "." );    }  \
      |                ^~~~~~~~ ./paths/long/Logging.h:231:1: note: in expansion of macro ‘LoggingBool’   231 | LoggingBool( KEEP_LOCS, False,
      | ^~~~~~~~~~~

Trying to fix it, I only find results where I would need to change something on the code, but I am not the developer, I just want to use it. Please let me know if you need additional information. I'm really getting desperate here. I just want to assemble some NGS + ONT Data for my master thesis :(

[Pierre Lindenbaum]

what is your compiler version ?

[cybercat]

The gcc compiler version is 9.3.0

[h.mon]

ALLPATHS-LG latest version is very old (from 2015), and doesn't combine nicely with more recent compilers. There is a Conda channel with ALLPATHS-LG available ( https://anaconda.org/biobuilds/allpathslg ).

In case you really want to compile, you can:

1) use an older GCC, like so:

CC=gcc-4.9 CXX=g++-4.9 ./configure

2) Try to fix the errors. I do remember reading a blog / github gist with fixes allowing ALLPATHS-LG compilation with more recent GCC versions, but I can find it again. This page is not very encouraging: 04.Running_Heidi_Lischer's_Pipeline.md

[cybercat]

Thank you! Using it with Conda finally works for me!