Unable to install on Linux-Debian

Issue #61 new
Ole Kristian Ekseth created an issue

Hi,

Tried compiling (ie, running make) on the root-folder (of your source): -- The compiler aborts, complaining about "RAStateTransitions::RAStateTransitions". (For details, see the uploaded error-file).

If you would have the time helping me with he issue (ie, to get your software compiling), I'd be utmost thankful! ;)

Best,

Ole Kristian Ekseth

Comments (4)

  1. Dmitry Tsarkov repo owner

    Hi Ole,

    I don't have an access to GCC 4.6 at the moment . Could you please try the latest master branch; I hope I fix the reported error.

  2. Ole Kristian Ekseth reporter

    Hi,

    Thanks for your quick reply: -- today I finally had time to test your fix: in brief it did not work (;

    What seems to be the issue, is that your code is not compatible with older (and in many server-contexts the only available) Linux "g++" compilers*. What I base this assumption on are the observations of: (a) in-compatibility of basic code-lines, exemplified using the following code-snippet, a snippet which does not compile on a stand-alone C++ file w/"g++" version 4.6.3: namespace {std::ofstream LLprivate;} std::ostream& LL { LLprivate }; (b) the other error-messages (where the first 50 lines of it inlcuded below) describes the error when the [above] code-line(s) is commented out; (c) your make-file: the variables it uses was not recognized by my "make", ie, indicates that your approach has not been tested wrt. back-compatibility. An example is "#define Mc OS X flags"

    What I will do, is to try to install a new (and updated) Ubuntu (ie, as the 'seemed to be' compiler-version-requreiment is no longer maintained for my OS version: thankfully I do have full access to the root on my test-system); -- the alliterative is to re-write your code to be back-compatible: assumes this task may be all to time-consuming (ie, more frustrating than pleasant), ie, what do you think?

    -- * I hope I do not ruin your day by my assertion: what I am saying (ie, remember that I may be wrong) is that your software is optimized towards compilers written after March 22 2013 ("https://gcc.gnu.org/releases.html"). --

    Again, many thanks for your reply: I do believe we are closer to solving this issue, ie look forward for your reply! ;)

    Best,

    Ole Kristian

    ------------------ error ------------- In file included from modelCacheSingleton.h:23:0, from dlTBox.h:30, from dlTBox.cpp:20: modelCacheConst.h:36:42: error: expected ‘;’ at end of member declaration modelCacheConst.h:36:48: error: ‘override’ does not name a type modelCacheConst.h:54:44: error: expected ‘;’ at end of member declaration modelCacheConst.h:54:50: error: ‘override’ does not name a type modelCacheConst.h:61:64: error: expected ‘;’ at end of member declaration modelCacheConst.h:61:70: error: ‘override’ does not name a type In file included from dlTBox.h:30:0, from dlTBox.cpp:20: modelCacheSingleton.h:38:42: error: expected ‘;’ at end of member declaration modelCacheSingleton.h:38:48: error: ‘override’ does not name a type modelCacheSingleton.h:55:44: error: expected ‘;’ at end of member declaration modelCacheSingleton.h:55:50: error: ‘override’ does not name a type modelCacheSingleton.h:62:64: error: expected ‘;’ at end of member declaration modelCacheSingleton.h:62:70: error: ‘override’ does not name a type In file included from Taxonomy.h:25:0, from RoleMaster.h:27, from dlTBox.h:31, from dlTBox.cpp:20: taxVertex.h: In constructor ‘TaxonomyVertex::TaxonomyVertex(const ClassifiableEntry*)’: taxVertex.h:143:50: error: type ‘TaxonomyVertex’ is not a direct base of ‘TaxonomyVertex’ In file included from tDepSet.h:27:0, from DepSet.h:23, from ConceptWithDep.h:24, from dlDag.h:32, from dlTBox.h:33, from dlTBox.cpp:20: tHeadTailCache.h: At global scope: tHeadTailCache.h:47:37: error: ‘virtual THeadTailCache<HeadType, TailType>::~THeadTailCache()’ declared virtual cannot be defaulted in the class body In file included from DepSet.h:23:0, from ConceptWithDep.h:24, from dlDag.h:32, from dlTBox.h:33, from dlTBox.cpp:20: tDepSet.h:106:55: error: expected ‘;’ at end of member declaration tDepSet.h:106:57: error: ‘override’ does not name a type tDepSet.h:145:41: error: expected ‘;’ at end of member declaration tDepSet.h:145:43: error: ‘override’ does not name a type In file included from dlTBox.cpp:25:0: cppi.h:36:28: sorry, unimplemented: non-static data member initializers cppi.h:36:28: error: ISO C++ forbids in-class initialization of non-const static member ‘oldPercent’ cppi.h:37:28: sorry, unimplemented: non-static data member initializers cppi.h:37:28: error: ISO C++ forbids in-class initialization of non-const static member ‘curPercent’ In file included from Reasoner.h:28:0, from ReasonerNom.h:23, from dlTBox.cpp:28: modelCacheIan.h:74:42: error: expected ‘;’ at end of member declaration modelCacheIan.h:74:48: error: ‘override’ does not name a type


  3. Dmitry Tsarkov repo owner

    Hi Ole,

    Thanks for the detailed report.

    There are couple of things I can suggest here: 1) Try to replace -std=c++-11 on the 2nd line of Makefile.include with -std=c++0x. 2) You can use FaCT++ v1.6.4 rather than the latest master. Version 1.6.4 uses the old (pre-C++11) C++ standard; after it I've switch to C++11, that is the main reason of all the problems you see. There should be no functional differences between current version and 1.6.4

    Best, Dmitry.

  4. Ole Kristian Ekseth reporter

    Hi,

    Thanks for the answer: -- what I now have is (among others) a "./32bit/FaCT++" executable. However, to my surprise there is no input-option (listed when calling "./32bit/FaCT++ -get-default-options") to entail an OWL ontology: in brief, am I using the wrong executable?

    PS: based on this chat we could consider updating your wiki: what do you think?

  5. Log in to comment