[bt-devel] Hello! and GCC 3.2

[mark lybarger]

Jason Gress wrote:
> Hello!  To some of you my name may seem familiar (as I made the bug report 
> that led to solving the memory leak in Bibletime 1.2beta1).  I had been 
> hoping to help out for a while now, and I hope to start somewhere. :)  And 
> so, using Gentoo 1.4 Linux version (with GCC 3.2), I was wondering - are any 
> of the developers using it as well? 

Hi Jason! I am using Gentoo 1.4 with bibletime.  I haven't tried 
recently to build the latest cvs of bibletime.  I am looking into gentoo 
ebuild scripts which will use sword 1.5.4 and bt 1.2.2.  There's a sword 
ebuild i found on bugs.gentoo.org, it works good. I'll be testing a bt 
ebulid script, and also a script for sword-modules to install some 
modules.  My hope is that in the near future, any gentoo user will be 
able to have a working version of bibletime simply by entering "emerge 
bibletime".

  I am asking because Bibletime 1.2.1 does
> not compile properly with it, and the current (as of today) cvs simply 
> doesn't compile ;).  So, I was hoping to try to make a small patch to fix the 
> problems so at least I can use it, and maybe others too, as I may be able to 
> share the .ebuild (Gentoo's package management system) with others.  

I would like to see your ebuild script.  The one I'm about to try is 
from Pat Double.  I found Pat on the gentoo-dev mailing list.  Seems 
there's lots of interest in creating Gentoo ebuild scripts.   Now if we 
could only get some of these scripts accepted into gentoo's main 
portage, we'd be all set.

I wanted
> to run by you guys what I have found (though some may have been fixed in cvs, 
> not sure) and I would be more than happy to help you guys test, and hopefully 
> solve some more simple problems with GCC 3.2 compliance.  I do know some C++, 
> as I have taken two semesters and am in the middle of my third.  So, here is 
> what I have found so far, with my comments as I feel necessary.  I apologize 
> ahead of time if I state the obvious; at my level of C++ skill, a so-called 
> "revelation" may be "duh" obvious to you ;) but what can I do :).
> 
> All testing was done with a plain ./configure.  SWORD 1.5.4 is installed.
> 1)  in bibletime-1.2.1/bibletime/backend/cswordbookmoduleinfo.cpp: 
> cswordbookmoduleinfo.cpp:55: `cout' undeclared (first use this function)
> I simply added #include <iostream>
> 
> 2)  in bibletime-1.2.1/bibletime/printing/cprintdialogpages.cpp:
> cprintdialogpages.cpp: In constructor
>    `CStyleListPage::CStyleListPage(CPrinter*, QWidget*, const char*)':
> cprintdialogpages.cpp:57: declaration of `name' shadows a parameter
> This one stumped me at first; but it seems to be that the array "name[4]" on 
> line 57 is of the same "name" (no pun intended) as the const char * name 
> parameter for the function.  I changed the name <g> name[4] to uppercase to 
> try, and what appears to be its reference on line 64.  Don't know if that 
> fixed it, but it compiles now ;).
> 
> 3)  in bibletime-1.2.1/bibletime/frontend/presenters/cbiblepresenter.cpp,
> This one is wierd:
> cbiblepresenter.cpp: In member function `virtual void
>    CBiblePresenter::storeSettings(CProfileWindow*)':
> cbiblepresenter.cpp:411: call of overloaded `pow(int, int)' is ambiguous
> /usr/include/bits/mathcalls.h:143: candidates are: double pow(double, double)
> /usr/include/g++-v32/cmath:427:                 long double std::pow(long
>    double, int)
> /usr/include/g++-v32/cmath:423:                 float std::pow(float, int)
> /usr/include/g++-v32/cmath:419:                 double std::pow(double, int)
> /usr/include/g++-v32/cmath:410:                 long double std::pow(long
>    double, long double)
> /usr/include/g++-v32/cmath:401:                 float std::pow(float, float)
> cbiblepresenter.cpp: In member function `virtual void
>    CBiblePresenter::applySettings(CProfileWindow*)':
> cbiblepresenter.cpp:422: call of overloaded `pow(int, int)' is ambiguous
> /usr/include/bits/mathcalls.h:143: candidates are: double pow(double, double)
> /usr/include/g++-v32/cmath:427:                 long double std::pow(long
>    double, int)
> /usr/include/g++-v32/cmath:423:                 float std::pow(float, int)
> /usr/include/g++-v32/cmath:419:                 double std::pow(double, int)
> /usr/include/g++-v32/cmath:410:                 long double std::pow(long
>    double, long double)
> /usr/include/g++-v32/cmath:401:                 float std::pow(float, float)
> cbiblepresenter.cpp:423: call of overloaded `pow(int, int)' is ambiguous
> /usr/include/bits/mathcalls.h:143: candidates are: double pow(double, double)
> /usr/include/g++-v32/cmath:427:                 long double std::pow(long
>    double, int)
> /usr/include/g++-v32/cmath:423:                 float std::pow(float, int)
> /usr/include/g++-v32/cmath:419:                 double std::pow(double, int)
> /usr/include/g++-v32/cmath:410:                 long double std::pow(long
>    double, long double)
> /usr/include/g++-v32/cmath:401:                 float std::pow(float, float)
> Apparently, it can't decide which pow to overload to, even though it seems to 
> me it should clearly be an int.  Oh well, my little hack (and it is) is to 
> force it to be a double with .0's on the end of some of the numbers and with 
> the already existing (int) typecase it should work.. not happy but might do 
> the trick..
> 
> 4)  This is the one I cannot get past, and I am looking for ideas.
> In bibletime-1.2.1/bibletime/frontend/searchdialog/csearchdialog.cpp:
> csearchdialog.cpp: In member function `void
>    CSearchDialog::setModuleList(ListCSwordModuleInfo&)':
> csearchdialog.cpp:117: no match for `ListCSwordModuleInfo& !=
>    QPtrList<CSwordModuleInfo>&' operator
> /usr/qt/3/include/qcstring.h:299: candidates are: bool operator!=(const
>    QCString&, const QCString&)
> /usr/qt/3/include/qcstring.h:302:                 bool operator!=(const
>    QCString&, const char*)
> /usr/qt/3/include/qcstring.h:305:                 bool operator!=(const char*,
>    const QCString&)
> /usr/qt/3/include/qstring.h:306:                 bool operator!=(QChar, QChar)
> /usr/qt/3/include/qstring.h:311:                 bool operator!=(char, QChar)
> /usr/qt/3/include/qstring.h:316:                 bool operator!=(QChar, char)
> /usr/qt/3/include/qstring.h:828:                 bool operator!=(const
>    QString&, const QString&)
> /usr/qt/3/include/qstring.h:835:                 bool operator!=(const
>    QString&, const char*)
> /usr/qt/3/include/qstring.h:841:                 bool operator!=(const char*,
>    const QString&)
> /usr/qt/3/include/qpoint.h:151:                 bool operator!=(const QPoint&,
>    const QPoint&)
> /usr/qt/3/include/qsize.h:155:                 bool operator!=(const QSize&,
>    const QSize&)
> /usr/qt/3/include/qrect.h:139:                 bool operator!=(const QRect&,
>    const QRect&)
> It seems obvious it is another matching error of sorts, but I am not familiar 
> with the classes, etc. Bibletime uses, so I can't fix this without help.  Any 
> ideas?  I am hoping that when all is said and done, the *Great* Bibletime can 
> be GCC 3.2 compliant! :)
> 
> 	Your Brother in Christ,
> 
> 	Jason Gress
> 
>