3 * building with CLISP (or explaining why not)
4 * urgent EVAL/EVAL-WHEN/%COMPILE/DEFUN/DEFSTRUCT cleanup:
5 ** made inlining DEFUN inside MACROLET work again
6 ** (also, while working on INLINE anyway, it might be easy
7 to flush the old MAYBE-INLINE cruft entirely,
8 including e.g. on the man page)
9 ** fixed bug 137 (more)
10 * faster bootstrapping (both make.sh and slam.sh)
11 ** added mechanisms for automatically finding dead code, and
12 used them to remove dead code
13 ** moved stuff from warm init into cold init where possible
14 (so that slam.sh will run faster and also just because
15 ideally everything would be in cold init)
16 ** profiled and tweaked
17 * fixed (TRACE :REPORT PROFILE ...) interface to profiling
18 * more EVAL/EVAL-WHEN/%COMPILE/DEFUN/DEFSTRUCT cleanup:
19 ** made %COMPILE understand magicality of DEFUN FOO
20 w.r.t. e.g. preexisting inlineness of FOO
21 ** used %COMPILE where COMPILE-TOP-LEVEL used to be used
22 ** removed now-redundant COMPILE-TOP-LEVEL and
23 FUNCTIONAL-KIND=:TOP-LEVEL stuff from the compiler
24 ** (ideally, but perhaps too hard, given what I've discovered
25 about the godawful internals of function debug names):
26 made FUNCTION-NAME logic work on closures, so that
27 various public functions like CL:PACKAGEP which
28 are now implemented as closures (because
29 they're structure slot accessors) won't be so
31 * rewrote long-standing confusing error restarts for redefining
33 * outstanding embarrassments
34 ** cut-and-pasted DEF-BOOLEAN-ATTRIBUTE (maybe easier to fix
35 now that EVAL-WHEN works correctly..)
37 ** :IGNORE-ERRORS-P cruft in stems-and-flags.lisp-expr. (It's
38 reasonable to support this as a crutch when initially
39 bootstrapping from balky xc hosts with their own
40 idiosyncratic ideas of what merits FAILURE-P, but it's
41 embarrassing to have to use it when bootstrapping
43 ** weird double-loading (first in GENESIS, then in warm init)
44 of src/assembly/target/*.lisp stuff, and the associated
45 weirdness of the half-baked state (compiler almost but
46 not quite ready for prime time..) of the system after
48 * fixups now feasible because of pre7 changes
49 ** ANSIfied DECLAIM INLINE stuff (deprecating MAYBE-INLINE)
50 * miscellaneous simple refactoring
52 ** renamed %PRIMITIVE to %VOP
53 * These days ANSI C has inline functions, so..
54 ** redid many cpp macros as inline functions:
55 HeaderValue, Pointerp, CEILING, ALIGNED_SIZE,
56 GET_FREE_POINTER, SET_FREE_POINTER,
57 GET_GC_TRIGGER, SET_GC_TRIGGER, GetBSP, SetBSP,
58 os_trunc_foo(), os_round_up_foo()
59 ** removed various avoid-evaluating-C-macro-arg-twice
61 * added mechanisms for automatically finding dead symbols in
62 package-data.lisp-expr (i.e. those symbols not bound,
63 fbound, defined as types, or whatever), and used them
64 to remove dead symbols
65 * Either get rid of or at least rework the fdefinition/encapsulation
66 system so that (SYMBOL-FUNCTION 'FOO) is identically equal to
68 =======================================================================
71 * refactored in preparation for moving CLOS into cold init and merging
72 SB-PCL:FOO with CL:FOO (for FOO=CLASS, FOO=CLASS-OF, etc.)
73 ** systematized support for MOP (new regression tests, maybe
74 new SB-MOP package..) to try to make sure things don't
75 get mislaid in the upcoming CLOS restructuring
76 ** extracted type system from SB-KERNEL into new SB-TYPE
78 ** reimplemented GENERIC-FUNCTION as a primitive object (or
79 maybe made SB-MOP:FUNCALLABLE-STANDARD-OBJECT the
80 primitive object, and then let GENERIC-FUNCTIONs
81 inherit from that) instead of structures with
82 :ALTERNATE-METACLASS and funcallableness. Now
83 FUNCALLABLE-INSTANCE can go away. (And now the new
84 funcallable primitive objects need to go into
85 collections like *FUN-HEADER-WIDETAGS* where
86 FUNCALLABLE-INSTANCE objects used to be.)
87 ** reimplemented CONDITIONs as primitive objects instead of
88 structures with :ALTERNATE-METACLASS. Now (between
89 this and the change to GENERIC-FUNCTIONs)
90 DEFSTRUCT :ALTERNATE-METACLASS can go away.
91 ** (maybe) Now INSTANCE_POINTER_LOWTAG can become just
92 STRUCTURE_POINTER_LOWTAG, and the concept of
93 SB-KERNEL:INSTANCE (including INSTANCEP,
94 (SPECIFIER-TYPE 'INSTANCE), etc.) can go away.
95 * moved CLOS into cold init, in order to allow CLOS to be used in the
96 implementation of the core system (e.g. the type system and the
97 compiler) and in order to support merger of CL:CLASS with
99 * (maybe) eliminated warm init altogether in favor of cold init
100 * (maybe, especially if warm init can be eliminated) rationalized
101 the build process, fixing miscellaneous pre-0.5.0 stuff that's
102 transparently not the right thing
103 ** removed separate build directories, now just building in
104 place with .sbclcoldfasl extensions
105 * (maybe) more refactoring in preparation for merging SB-PCL:FOO
106 into CL:FOO: reimplemented type system OO dispatch
107 (!DEFINE-TYPE-METHOD, etc.) in terms of CLOS OO dispatch
108 * merged SB-PCL:FOO into CL:FOO (and similarly CLASS-OF, etc.)
109 * added some automatic tests for basic binary compatibility, in hopes
110 that it might be practical to maintain binary compatibility
111 between minor maintenance releases on the stable branch (but no
112 promises, sorry, since I've never tried to do this before, and
113 have no idea how much of a pain this'll be)
114 ========================================================================
115 for 1.0 (fixes of lower priority which I'd nonetheless be embarrassed
116 to leave unfixed in 1.0):
117 * all too many BUGS entries and FIXMEs
118 =======================================================================
119 other priorities, no particular time:
121 * bug fixes, especially really annoying bugs (ANSI or not) and any
122 ANSI bugs (i.e. not just bugs in extras like the debugger or
123 "declarations are assertions", but violations of the standard)
124 * better communication with the outside world (scratching WHN's
125 personal itch): I don't want socket-level stuff so much as I
126 want RPC-level or higher (CORBA?) interfaces and (possibly
127 through RPC or CORBA) GUI support
128 =======================================================================
129 important but out of scope (for WHN, anyway: Patches from other people
130 are still welcome!) until after 1.0:
132 * sadly deteriorated support for ANSI-style block compilation
133 (static linking of DEFUNs within a single file or
134 WITH-COMPILATION-UNIT)
135 * various GC issues (exuberant cut-and-paste coding,
136 possibly dangerously over-conservative handling
137 of neighbors of function objects, general GC efficiency)
138 * package issues other than SB!TYPE, SB!MOP, and dead exported
140 * Any systematic effort to fix compiler consistency checks is
141 out of scope. (However, it still might be possible to
142 determine that some or all of them are hopelessly stale
144 =======================================================================
145 other known issues with no particular target date:
147 bugs listed on the man page
149 hundreds of FIXME notes in the sources from WHN
151 various other unfinished business from CMU CL and before, marked with
152 "XX" or "XXX" or "###" or "***" or "???" or "pfw" or "@@@@" or "zzzzz"
153 or probably also other codes that I haven't noticed or have forgotten.
155 (Things marked as KLUDGE are in general things which are ugly or
156 confusing, but that, for whatever reason, may stay that way
158 =======================================================================
159 "There's nothing an agnostic can't do as long as he doesn't know
160 whether he believes in anything or not."
163 "God grant me serenity to accept the code I cannot change, courage to
164 change the code I can, and wisdom to know the difference."
167 "Accumulation of half-understood design decisions eventually chokes a
168 program as a water weed chokes a canal. By refactoring you can ensure
169 that your full understanding of how the program should be designed is
170 always reflected in the program. As a water weed quickly spreads its
171 tendrils, partially understood design decisions quickly spread their
172 effects throughout your program. No one or two or even ten individual
173 actions will be enough to eradicate the problem."
174 -- Martin Fowler, in _Refactoring: Improving the Design of Existing
177 "I wish I didn't know now what I didn't know then."