source: project/chicken/trunk/TODO @ 13300

Last change on this file since 13300 was 13300, checked in by felix winkelmann, 11 years ago

applied read-lime-limit patch by Jim Ursetto

File size: 5.8 KB
Line 
1TODO for trunk                                                          -*- Outline -*-
2
3
4* bugs
5
6** compiler
7*** pre-optimization
8**** changes call-sites and makes them invalid for later pre-optimization
9*** check profiling (-profile vs. profile-declaration)
10**** this doesn't always seem to work, in particular with library units
11**** ##sys#finish-profile should always be invoked
12     is it in `exit'?
13*** when re-defining intrinsics, the compiler should warn and disable re-writes
14    add declaration to keep re-writes enabled for core library files
15*** check in foreign.import.scm and compiler.import.scm whether the import
16    took place in the compiler
17
18** expander
19*** expansion of r5rs_pitfall 4.2 incorrect
20    possibly due to a non-aliased implicit "begin" somewhere
21*** dirty-macros.scm loops when using `defile'
22    possibly due to unrenamed special forms
23*** extended lambda-lists refer to `optional' and `let-optionals[*]'
24    this will break, when these macros are renamed on import
25
26** modules
27*** `require-extension' fails in interpreter when extension is .so without import library
28    probably because load and import is in same toplevel form, and the
29    load is performed after the import (at runtime). A possibly fix
30    would be to load the extension immediately (at
31    compile-/expansion-time), when import is #t and o import lib
32    exists.
33*** mark import-source (module) on plist of imported symbols and use in re-import warning
34    doesn't work that easily, try to find another solution
35
36** libraries
37*** library/runtime: cyclic list checks for assq/assv/assoc/memq/memv/member
38    and C_i_list_tail
39*** finalizers on lexically ref'd data not working in interpreter (reported
40    by Jim Ursetto)
41
42** build
43*** get rid of auto-import warnings for mingw compiler/linker
44
45
46* tasks
47
48** branches
49*** try to improve performance in lazy-gensyms
50
51** expander
52*** test new implementation of `define-for-syntax'
53**** test "numbers" egg
54**** is s48-modules still working?
55**** port xlib egg
56*** at some stage remove debug-output in expand.scm
57
58** module issues
59*** need a way to omit generating module registration without generating import lib
60    for example when compiling static version, where implib already exists
61*** code-duplication in compiler and evaluator for ##core#module
62*** "scheme" module does not include some special forms ("define-syntax", etc.)
63*** import-for-syntax imports value bindings into import-env
64    should actually be a distinct meta-import-env.
65    (does it work at all?)
66    example: we need `(import-for-syntax chicken)' to have access to
67    `receive' in a procedural syntax definition.
68**** could be tested with xlib egg
69*** curried define performs expansion in empty se - problem?
70    (as comment in expand.scm indicated (##sys#register-export))
71*** check phase separation and module access
72**** see "expander" above
73**** interaction for define-for-syntax, begin-for-syntax, and macros
74***** check for different execution modes (compile-time, run-time, csi, ...)
75***** figure out visibility
76
77** compiler
78*** test define-rewrite-rule
79*** (csc) generate object-files in /tmp (or TMPDIR)?
80
81** setup/install
82*** setup-download
83**** use "HTTP_PROXY"/"http_proxy" env. var
84**** handle redirects in http-fetch
85*** automatically update db after extension installation?
86*** test sudo
87
88** library units
89*** read-mark list should be stored in read-table
90*** ports should have indicator slot about exact type (tcp, file, ...)
91*** normalize-pathname: return short name on windows? (mingw/msvc)
92
93** syntax-error
94*** if ##sys#current-module is set, add name to error message?
95*** fluidly keep track of expanded forms (extend meaning of culprit) to pprint pruned expr on error
96
97** build
98*** need script to process import libraries for generating indices for doc.callcc.org
99    then tell Toby Butzon about it
100*** test DESTDIR and proper linking
101*** test special status of setup-api.import.scm and setup-download.import.scm
102    touching .c files caused chicken-install to be rebuilt; did using
103    .SECONDARY help?
104*** using "touch" with WINDOWS_SHELL won't work (need alternative)
105** test --enable-auto-import witg mingw builds
106
107** documentation
108*** document qs, normalize-pathname
109
110
111* tests
112
113** optional longer run (env. var)
114*** syntax-rules-stress-test
115*** fully compiled ec-tests
116** 3-stage bootstrap with compiler-output comparison
117
118
119* documentation
120
121** document new .meta entries in tutorials on wiki
122   depends, test-depends
123** chicken.texi needs to be regenerated
124
125
126* optimizations
127
128** self-recursion optimization
129   what MacScheme called "benchmark-mode" (assume self-calls are recursion)
130*** needs declaration or option, >= -O2
131
132** when inlining, consing arg-list with "list" may make get-keyword possible foldable
133
134** compiler-support for get-keyword ?
135
136** lambda-fusion / "fuse-and-dispatch" (suggested by Alex Shinn)
137   convert groups of local lambdas referenced to only in operator-position into
138   looping lambda + dispatch (static variable can be used), otherwise similar to
139   a conditional
140*** new forms (after optimization, prepared language)
141    [##core#dispatch LAMBDABODY1 ... BODY]
142    [##core#goto {INDEX} ARGUMENT1 ...}
143
144** lazy gensyms (see "lazy-gensyms" branch)
145
146** handle optional args primitively
147   for example, set to distinguished val (checking argc) on C level, core
148   primitive for checking this - should be quite simple, but llist processing
149   (decompose-lambda-list) will be more expensive and cumbersome.
150
151** pre-hashed symbols (extra symbol slot)
152   The memory usage should be acceptable, performance gain is hard to guess.
153   Some experiments indicate that hashing the string is cheaper than it appears,
154   but low-level hashtables should get the most of this (and thus speed up
155   the compiler)
156
157** Lisp2-hack by Dybvig
158*** add function-cell to symbol, call on direct call without checks
159*** initialized to trap function on set!/define, which does fixup
160*** would also allow inline-caching hacks
Note: See TracBrowser for help on using the repository browser.