Trouble Installing Genometools-1.5.1 On My Mac Ox

Question

[inadamj]

I want to use annotation-sketch in order to draw gene models and protein domain architecture. It has been 2 days trying to build genometools. I installed fink to have the necessary packages however I still receive this message:

Once I type : Nadas-MacBook-Pro:genometools-1.5.1 nada$ sudo I receive:

> Password:
[compile giextract.o]
src/match/giextract.c:754:7: error: variable 'cc' is used uninitialized whenever
      'if' condition is false [-Werror,-Wsometimes-uninitialized]
  if (!haserr)
      ^~~~~~~
src/match/giextract.c:766:13: note: uninitialized use occurs here
  gt_assert(cc >= 0);
            ^~
/Users/nadaalhirabi/genometools-1.5.1/src/core/assert_api.h:37:17: note: 
      expanded from macro 'gt_assert'
          if (!(expression)) {                                               \
                ^
src/match/giextract.c:754:3: note: remove the 'if' if its condition is always
      true
  if (!haserr)
  ^~~~~~~~~~~~
src/match/giextract.c:746:10: note: initialize the variable 'cc' to silence this
      warning
  char cc;
         ^
          = '\0'
1 error generated.
make: *** [obj/src/match/giextract.o] Error 1



![enter image description here][1]


  [1]: http://postimg.org/image/mxdi5c14d/

[sst]

Use make 64bit=yes errorcheck=no (and don't forget doing a make cleanup before rebuilding). The Makefile in genometools treats compiler warnings as errors, and the errorcheck=no flag will make it ignore them. The issue here is that version 1.5.1 still triggers some warnings in newer compilers, but these warnings are already eliminated in the newest development version. You should be fine just ignoring them.

[SES]

Try to compile with: make assert=no. You won't need to sudo to compile but you will need it for installing.

[inadamj]

I tried and it started to compile then I have this error :

Nadas-MacBook-Pro:genometools-1.5.1 nada $ make assert=no
[compile giextract.o]
[compile greedyedist.o]
[compile greedyfwdmat.o]
[compile hashfirstcodes.o]
[compile idx-limdfs.o]
[compile idxlocali.o]
[compile idxlocalidp.o]
[compile idxlocalisw.o]
[compile index_options.o]
[compile initbasepower.o]
[compile initeqsvec.o]
[compile iter-window.o]
[compile kmer2string.o]
[compile lua_tools.o]
[compile marksubstring.o]
[compile merger-trie.o]
[compile mssufpat.o]
[compile myersapm.o]
[compile nullcols.o]
[compile optionargmode.o]
[compile pck-count-nodes.o]
[compile pckbucket.o]
[compile pckdfs.o]
[compile prsqualint.o]
src/match/prsqualint.c:77:7: error: variable 'readint' is used uninitialized
      whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized]
  if (!haserr)
      ^~~~~~~
src/match/prsqualint.c:109:43: note: uninitialized use occurs here
  qualint->integervalue = (unsigned long) readint;
                                          ^~~~~~~
src/match/prsqualint.c:77:3: note: remove the 'if' if its condition is always
      true
  if (!haserr)
  ^~~~~~~~~~~~
src/match/prsqualint.c:52:15: note: initialize the variable 'readint' to silence
      this warning
  long readint;
              ^
               = 0
1 error generated.
make: *** [obj/src/match/prsqualint.o] Error 1

[SES]

I have not seen this error and it may take a bit of trial and error for me to diagnose the problem. My advice is to send an email to the genometools mailing list. They will respond very quickly and I'm certain you will find a solution that way.

[inadamj]

I have been trying for days but did not solve it.. thanks for the suggestion .. I sent the problem to them and I hope there is a solution :(

[SES]

Sorry to hear that. I've always gotten a solution to my issues very quickly when I've emailed the list. Good luck to you.

[Jeremy Leipzig]

have you tried changing char cc; to char cc = '\0';

[inadamj]

From where ?? I am new at Mac and genome world :(

[inadamj]

I changed a lot of the initialization as the command line asked

but I still have this :

> ld: symbol(s) not found for architecture i386 clang: error: linker
> command failed with exit code 1 (use -v to see invocation) make: ***
> [bin/gt] Error 1

[Alex Reynolds]

Assert statements that fail usually indicate internal problems with the code that the developers need to look at and fix (as part of the development process, asserts show conditions that should never happen unless the code's logic is broken). If you don't hear from the mailing list in 3-4 days, you might at that time consider contacting the mailing list again.

[sst]

These issues have already been fixed in the current unstable version of GenomeTools, but the OP is using an older version which triggers some compiler warnings in newer compilers. The unstable version compiles fine on Mac OS X (tested up to 10.8.5). For the time being, just use errorcheck=no.