Hello Gabor! On Sunday, May 20, 2007 12:29:49 PM Gabor ROCZEI wrote:
> This is a short summary (I wrote it for the knowarcs people), use it: Did you even read my original mail? This is the error you sent patches for that are long in CVS (which I explicitly mentioned testing with): > ../../libMRML/include/CAttributeList.h:82: error: extra qualification > 'CAttributeList::' on member 'toXML' A trivial extra qualification. I've patched that away from the tarballed versions long before posting here. This is the error *I* get: CAFPerl.cc:8: error: cannot allocate an object of abstract type 'CAcPerl' ../../libGIFTAcPerl/include/CAcPerl.h:72: note: because the following virtual functions are pure within 'CAcPerl': ../../libMRML/include/CAccessor.h:109: note: virtual void CAccessor::getRandomIDs(std::list<int, std::allocator<int> >&, size_t) const ../../libMRML/include/CAccessor.h:119: note: virtual void CAccessor::getRandomAccessorElements(std::list<CAccessorElement, std::allocator<CAccessorElement> >&, size_t) const Here gcc complains about purely virtual member functions from which it can't generate an object. This is *completely* different from the error you're referring to. I'm sorry if I sound ungrateful or impatient but I've been battling with gift for several days (and nights) now, I've done extensive searching on Google, the gift mailinglists (back to 2003, btw.), I've waded through tons of spam on the bug-gift mailinglist hoping to find something meaningful there, I've bugged people with better C++ skills than myself, I've searched Debian, ArchLinux, OpenPKG and lots of other (meta-) distributions to find how they deal with this problem (they don't have a patch for this either), I've consulted lots of bugtrackers to see how people dealt with other applications with the same problem. I've checked *all* the patchsets posted by Julia "risc" Longtin to this very list. I've experimented with the code of 0.1.14 and 0.1.15b myself and patched away lots of other issues (yes, I've compared my patches with latest CVS and they're all in there or I would have submitted them). Heck, I've even checked the PRs for gcc 4.1.2 to see what changed in gcc to work around it. Only then did I post here because I didn't want to bother anyone with stuff I could easily fix myself. So maybe you can understand now why I'm not too happy about your answer. :-) No offence intended and best regards, Wulf
pgpLQEvTWq3t9.pgp
Description: PGP signature
_______________________________________________ help-GIFT mailing list [email protected] http://lists.gnu.org/mailman/listinfo/help-gift
