- May 16, 1991
-
-
ram authored
call OUTPUT-TYPE-OR-LOSE, which uses SUBTYPEP, and guarantees to return one of LIST, STRING, BIT-VECTOR, VECTOR or NIL (the last enabled by an optional.)
-
chiles authored
Modified all the condition reporting methods to FRESH-LINE since error messages did not always end up on their own line when using Hemlock commands to drive a slave in the debugger.
-
- May 15, 1991
- May 14, 1991
-
-
ram authored
correctly. COERCE is still broken.
-
- May 09, 1991
-
-
chiles authored
Fixed an error in the report for for unhandled-conditions. It was recursively trying to print the condition.
-
- May 08, 1991
-
-
ram authored
thing, but if the result is just a call to the inverse, then directly call the inverse w/o any let binding. This makes setf inverses usable without the full interpreter.
-
ram authored
%DEFINE-SETF-MACRO, which checks for any existing definition as a slot accessor or other setf function. Changed GET-SETF-METHOD to ignore setf macros (always global) when there is a local macro or function. Changed SETF to always call GET-SETF-METHOD, instead of special-casing setf inverses. I was tired to keeping the duplicated code consistent. Generalized the special casing in SETF of APPLY. We now go for the SETF function when called on any symbol in the Lisp package. Now SETF of APPLY will work on all Common Lisp functions, instead of just two.
-
ram authored
-
ram authored
-
ram authored
%APPLY-SVSET.
-
ram authored
-
wlott authored
-
- May 06, 1991
- May 04, 1991
- Apr 28, 1991
- Apr 25, 1991
- Apr 24, 1991
-
-
ram authored
and also correctly handles multi-level displacement. Made SVREF, et al, safe so that bounds checking is done.
-
ram authored
%WITH-ARRAY-DATA in the hairy cases. Now this macro checks that (<= start end length), to prevent bounds errors in unsafe code.
-
ram authored
is more clever. Also, changed it to accept any STRINGable thing, instead of just strings and symbols. These macros now bind the offset var instead of randomly setting it.
-
ram authored
-
ram authored
able to do it, and it can't cause any problems that I can think of.
-
- Apr 23, 1991
-
-
ram authored
QUICK-INTEGER-TO-STRING.
-
ram authored
size and type. Also, printing can be limited by object count, in which case we don't print page breaks.
-
ram authored
Changed INTERNAL-LOAD to CERROR if you attempt to load a file with a fasl type, but that doesn't have a fasl header. Changed FASLOAD to use FILE-LENGTH instead of LISTEN to tell whether the file is empty, since LISTEN currently doesn't work.
-
ram authored
safe code.
-
ram authored
-
ram authored
-
wlott authored
-
wlott authored
-
wlott authored
to load the last src word or not backwards, so we didn't load it when we needed it. (And potentially read one too many words when we didn't need it.)
-
- Apr 22, 1991
- Apr 21, 1991
-
-
ram authored
trigger page protections whenever it changes the lisp *GC-TRIGGER*. This should prevent problems with GC never happening when the reclaim goal is increased. Also, merged Bill's fix from the old system whereby we only frob *GC-INHIBIT* related to the inhibit hook when the hook returns T. And I changed the default before&after hooks to not beep unless *GC-VERBOSE* is :BEEP, which is *not* the default.
-
- Apr 20, 1991
-
-
ram authored
Also, added :OPTIMIZE and :OPTIMIZE-INTERFACE for changing the "global" compilation policy within the dynamic extent. Added :CONTEXT-DECLARATIONS, which provides a way to insert declarations conditional on pattern mathcing of the context in which the definition appears. So you can compiler all external functions safe, or whatever.
-