Index: Lib/decimal.py
===================================================================
--- Lib/decimal.py	(revision 76570)
+++ Lib/decimal.py	(working copy)
@@ -3462,10 +3462,6 @@
 
         if other._exp != 0:
             return context._raise_error(InvalidOperation)
-        liminf = -2 * (context.Emax + context.prec)
-        limsup =  2 * (context.Emax + context.prec)
-        if not (liminf <= int(other) <= limsup):
-            return context._raise_error(InvalidOperation)
 
         if self._isinfinity():
             return Decimal(self)
Index: Lib/test/test_decimal.py
===================================================================
--- Lib/test/test_decimal.py	(revision 76570)
+++ Lib/test/test_decimal.py	(working copy)
@@ -68,6 +68,16 @@
 skipped_test_ids = [
     'scbx164',  # skipping apparently implementation-specific scaleb
     'scbx165',  # tests, pending clarification of scaleb rules.
+
+    # the specification requires that the second argument to scaleb is
+    # in the range +-2 * (Emax + precision) inclusive.  This rather
+    # arbitrary restriction, which doesn't appear in IEEE 754, appears
+    # to exist solely for the benefit of decimal implementations in
+    # low-level languages.  For Python, we ignore this restriction
+    # and the associated tests.
+    'ddscb122', 'ddscb123', 'ddscb126', 'ddscb127',
+    'dqscb122', 'dqscb123', 'dqscb126', 'dqscb127',
+    'scbx122', 'scbx123', 'scbx126', 'scbx127',
 ]
 
 # Make sure it actually raises errors when not expected and caught in flags