- * Then (as shown below) the author needs to use, not the simple
- * JMPENV_PUSH, but several of its constitutent macros, to arrange for
- * the Perl internals to call thrower() rather than longjmp() to
- * report exceptions:
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2004-07/msg00378.html
+ *
+ * Joshua's original patches (which weren't applied) and discussion:
+ *
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1998-02/msg01396.html
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1998-02/msg01489.html
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1998-02/msg01491.html
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1998-02/msg01608.html
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1998-02/msg02144.html
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1998-02/msg02998.html
+ *
+ * Chip's reworked patch and discussion:
+ *
+ * http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/1999-03/msg00520.html
+ *
+ * The flaw in these patches (which went unnoticed at the time) was
+ * that they moved some code that could potentially die() out of the
+ * region protected by the setjmp()s. This caused exceptions within
+ * END blocks and such to not be handled by the correct setjmp().
+ *
+ * The original patches that introduces flexible exceptions were: