On Tue, Jan 27, 2004 at 10:09:12PM +0100, Georg Baum wrote: > I get this error when compiling latest CVS with gcc 2.95 + stlport: > > g++ -DHAVE_CONFIG_H -I. -I../../src -I. -I../../boost > -I/usr/include/stlport -I/usr/X11R6/include -DCXX_GLOBAL_CSTD > -ftemplate-depth-30 -nostdinc++ -g -W -Wall -c -o cursor.o `test -f > '../../src/cursor.C' || echo '../../src/'`../../src/cursor.C > ./../src/cursor.C:545: warning: #warning FIXME > ./../src/cursor.C:616: warning: #warning FIXME > ./../src/cursor.C:1084: warning: #warning FIXME > ./../src/cursor.C:1266: warning: #warning look here > ./../src/cursor.C:1282: warning: #warning look here > ./../src/cursor.C:1552: warning: #warning FIXME > ./../src/cursor.C: In method `bool LCursor::openable(const MathAtom &)': > ./../src/cursor.C:838: comparison of distinct pointer types `const > MathInset *' and `InsetBase *' lacks a cast > make[3]: *** [cursor.o] Fehler 1 > > Is this expected?
Not really. But we've had that problem somewhere else before. My take is that this is a bug in 2.9x, not in the code as such. > It compiles if I change line 838 of src/Cursor.C from > > if (t.operator->() != anchor_[depth()].inset()) > > to > > if (t.operator->() != > const_cast<InsetBase const *>(anchor_[depth()].inset())) Do that if it helps. Andre'
