>>>>> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> I think it is also logically more valid.
Yes.
Martin> BTW get rid of the
Martin> if (it < end() && ...
Martin> condition! It is redundant.
Indeed. I did that.
Please test. Personally I cannot crash it anymore, but my imagination
may be too limited.
JMarc
Index: src/mathed/ChangeLog
===================================================================
--- src/mathed/ChangeLog (revision 13268)
+++ src/mathed/ChangeLog (working copy)
@@ -1,3 +1,9 @@
+2006-02-23 Jean-Marc Lasgouttes <[EMAIL PROTECTED]>
+
+ * math_data.C (x2pos): Do not try to enter in non-nest insets;
+ avoids misplaced cursor (and subsequent crash) when placing the
+ mouse near a symbol inset.
+
2006-02-17 Georg Baum <[EMAIL PROTECTED]>
* math_xymatrixinset.[Ch]: Readd
Index: src/mathed/math_data.C
===================================================================
--- src/mathed/math_data.C (revision 13268)
+++ src/mathed/math_data.C (working copy)
@@ -393,7 +393,7 @@
#endif
if (it != begin()) {
--it;
- if (it < end() && (*it)->getChar())
+ if (!(*it)->asNestInset())
++it;
}
