0.7.1.28:

[sbcl.git] / BUGS

diff --git a/BUGS b/BUGS
index bdf4781..b0d7dae 100644 (file)
--- a/BUGS
+++ b/BUGS
@@ -337,9 +337,11 @@ WORKAROUND:
           blows up at the level of SPECIFIER-TYPE with
           "Lower bound (0) is greater than upper bound (0)." Probably
           SPECIFIER-TYPE should return the NIL type instead.
-       g: The type system isn't all that smart about relationships
-          between hairy types, as shown in the type.erg test results,
-          e.g. (SUBTYPEP 'CONS '(NOT ATOM)) => NIL, NIL.
+       g: The type system [still] isn't all that smart about relationships
+          between hairy types. [The original example from PVE was
+          (SUBTYPEP 'CONS '(NOT ATOM)) => NIL, NIL, which was fixed
+          by CSR in sbcl-0.7.1.28, but there are still
+           plenty of corner cases out there.]
 
 51:
   miscellaneous errors reported by Peter Van Eynde July 25, 2000:
@@ -1117,6 +1119,11 @@ WORKAROUND:
   T
   T
 
+  This is probably due to underzealous clearing of the type caches; a
+  brute-force solution in that case would be to make a defclass expand
+  into something that included a call to SB-KERNEL::CLEAR-TYPE-CACHES,
+  but there may be a better solution.
+
 141: 
   Pretty-printing nested backquotes doesn't work right, as 
   reported by Alexey Dejneka sbcl-devel 2002-01-13: