factor/TODO.FACTOR.txt

[error] AWT-EventQueue-0: java.lang.NullPointerException
[error] AWT-EventQueue-0:  at org.gjt.sp.jedit.Buffer.markTokens(Buffer.java:2109)
[error] AWT-EventQueue-0:  at factor.jedit.WordPreview.getWordAtCaret(WordPreview.java:95)
[error] AWT-EventQueue-0:  at factor.jedit.WordPreview.showPreview(WordPreview.java:137)
[error] AWT-EventQueue-0:  at factor.jedit.WordPreview.actionPerformed(WordPreview.java:79)
[error] AWT-EventQueue-0:  at javax.swing.Timer.fireActionPerformed(Timer.java:271)

+ inference/dataflow:

- type inference
- handle odd base cases, with code after ifte
- handle recursion with when, when* etc
- alien-call need special nodes

+ linearizer/generator:

- peephole optimizer
- tail call optimization
- getenv/setenv: if literal arg, compile as a load/store
- compiler: drop literal peephole optimization

+ compiler frontend:

- assembler opcodes dispatch on operand types
- save code in image
- compile word twice; no more 'cannot compile' error!

+ oop:

- union metaclass
- make M: order-independent
- bootstrapping generic words
- 2generic
- move generic, 2generic from kernel vocabulary
- generic = hashcode and math ops

+ ffi:

- is signed -vs- unsigned pointers an issue?
- bitfields in C structs
- SDL_Rect** type
- struct membres that are not *
- FFI float types

+ listener/plugin:

- sidekick: still parsing too much
- errors don't always disappear
- console: wrong history
- listener: if too many things popped off the stack, complain
- NPE in ErrorHighlight
- some way to not have previous definitions from a source file
  clutter the namespace
- maple-like: press enter at old commands to evaluate there
- completion in the listener
- special completion for USE:/IN:

+ kernel:

- profiler is inaccurate: wrong word on cs
- better i/o scheduler
- remove sbufs
- cat, reverse-cat primitives
- first-class hashtables
- rewrite accessors and mutators in Factor, with slot/set-slot primitive
- replace -export-dynamic with sundry-xt
- add a socket timeout

+ misc:

- some way to step over a word in the stepper
- step: print NEXT word to execute, not word that JUST executed
- perhaps /i should work with all numbers
- unit test weirdness: 2 lines appears at end
- jedit ==> jedit-word, jedit takes a file name
- command line parsing cleanup
- nicer way to combine two paths
- ditch object paths
- browser responder for word links in HTTPd; inspect responder for
  objects
- worddef props
- prettyprint: when unparse called due to recursion, write a link

+ httpd:

- log with date
- file responder; last-modified field
start dataflow optimizer 2004-12-02 22:44:36 -05:00			`[error] AWT-EventQueue-0: java.lang.NullPointerException`
			`[error] AWT-EventQueue-0: at org.gjt.sp.jedit.Buffer.markTokens(Buffer.java:2109)`
			`[error] AWT-EventQueue-0: at factor.jedit.WordPreview.getWordAtCaret(WordPreview.java:95)`
			`[error] AWT-EventQueue-0: at factor.jedit.WordPreview.showPreview(WordPreview.java:137)`
			`[error] AWT-EventQueue-0: at factor.jedit.WordPreview.actionPerformed(WordPreview.java:79)`
			`[error] AWT-EventQueue-0: at javax.swing.Timer.fireActionPerformed(Timer.java:271)`

			`+ inference/dataflow:`

stack effects for all primitives, updated to do list, literal hashtable syntax {{ [ key \| value ] ... }} 2004-11-11 16:45:55 -05:00			`- type inference`
start dataflow optimizer 2004-12-02 22:44:36 -05:00			`- handle odd base cases, with code after ifte`
			`- handle recursion with when, when* etc`
more work on optimizer 2004-12-03 22:12:58 -05:00			`- alien-call need special nodes`
a few bug fixes 2004-10-27 23:13:00 -04:00
start dataflow optimizer 2004-12-02 22:44:36 -05:00			`+ linearizer/generator:`
some cleanups and benchmarks 2004-10-27 21:21:31 -04:00
start dataflow optimizer 2004-12-02 22:44:36 -05:00			`- peephole optimizer`
stack inference fixes, do not linearize #values 2004-12-08 18:39:36 -05:00			`- tail call optimization`
stack effects for all primitives, updated to do list, literal hashtable syntax {{ [ key \| value ] ... }} 2004-11-11 16:45:55 -05:00			`- getenv/setenv: if literal arg, compile as a load/store`
			`- compiler: drop literal peephole optimization`
start dataflow optimizer 2004-12-02 22:44:36 -05:00
			`+ compiler frontend:`

			`- assembler opcodes dispatch on operand types`
			`- save code in image`
			`- compile word twice; no more 'cannot compile' error!`

don't use esi for data stack anymore; os word pushes unix or win32; don't use ansi escape codes on windows 2004-12-13 15:37:50 -05:00			`+ oop:`

			`- union metaclass`
			`- make M: order-independent`
			`- bootstrapping generic words`
			`- 2generic`
			`- move generic, 2generic from kernel vocabulary`
			`- generic = hashcode and math ops`

start dataflow optimizer 2004-12-02 22:44:36 -05:00			`+ ffi:`

working on the FFI 2004-09-19 17:39:28 -04:00			`- is signed -vs- unsigned pointers an issue?`
working on sdl binding, remove some combinators 2004-10-13 23:06:40 -04:00			`- bitfields in C structs`
			`- SDL_Rect** type`
			`- struct membres that are not *`
fix for jvm factor bootstrap 2004-11-11 21:00:28 -05:00			`- FFI float types`
ditch map, rename inject to map 2004-08-22 17:28:31 -04:00
fix invalid select() usage 2004-08-13 18:43:03 -04:00			`+ listener/plugin:`
Factor jEdit plugin! 2004-08-06 02:51:32 -04:00
predicate metaclass; prettyprint, see, unparse, ' and other words are now generic 2004-12-12 23:49:44 -05:00			`- sidekick: still parsing too much`
some dataflow IR unit tests and fixes 2004-11-29 23:14:12 -05:00			`- errors don't always disappear`
inline annotation for combinators; faster stack checker taking advantage of this fact; started dataflow IR 2004-11-27 00:33:17 -05:00			`- console: wrong history`
ExternalFactor is done 2004-11-24 21:45:30 -05:00			`- listener: if too many things popped off the stack, complain`
cwd/cd primitives, dir./pwd library words, pipe word in JVM factor' 2004-09-04 03:06:53 -04:00			`- NPE in ErrorHighlight`
gc fix 2004-09-01 21:04:16 -04:00			`- some way to not have previous definitions from a source file`
			`clutter the namespace`
html stream generates css 2004-08-22 16:04:55 -04:00			`- maple-like: press enter at old commands to evaluate there`
completion popups in factor jedit plugin 2004-08-16 23:52:52 -04:00			`- completion in the listener`
			`- special completion for USE:/IN:`
Factor jEdit plugin! 2004-08-06 02:51:32 -04:00
some cleanups and benchmarks 2004-10-27 21:21:31 -04:00			`+ kernel:`
native factor can now parse most source files that comprise it 2004-07-19 00:34:03 -04:00
some cleanups and benchmarks 2004-10-27 21:21:31 -04:00			`- profiler is inaccurate: wrong word on cs`
documentation updates 2004-08-21 02:55:37 -04:00			`- better i/o scheduler`
stack effects for all primitives, updated to do list, literal hashtable syntax {{ [ key \| value ] ... }} 2004-11-11 16:45:55 -05:00			`- remove sbufs`
			`- cat, reverse-cat primitives`
			`- first-class hashtables`
			`- rewrite accessors and mutators in Factor, with slot/set-slot primitive`
			`- replace -export-dynamic with sundry-xt`
			`- add a socket timeout`
First import of Factor 0.60 2004-07-16 02:26:21 -04:00
			`+ misc:`

start dataflow optimizer 2004-12-02 22:44:36 -05:00			`- some way to step over a word in the stepper`
			`- step: print NEXT word to execute, not word that JUST executed`
			`- perhaps /i should work with all numbers`
oop fix, split up inference 2004-11-26 22:23:57 -05:00			`- unit test weirdness: 2 lines appears at end`
some cleanups and benchmarks 2004-10-27 21:21:31 -04:00			`- jedit ==> jedit-word, jedit takes a file name`
			`- command line parsing cleanup`
			`- nicer way to combine two paths`
stack effects for all primitives, updated to do list, literal hashtable syntax {{ [ key \| value ] ... }} 2004-11-11 16:45:55 -05:00			`- ditch object paths`
			`- browser responder for word links in HTTPd; inspect responder for`
			`objects`
			`- worddef props`
			`- prettyprint: when unparse called due to recursion, write a link`
First import of Factor 0.60 2004-07-16 02:26:21 -04:00
			`+ httpd:`

			`- log with date`
gc fix 2004-09-01 21:04:16 -04:00			`- file responder; last-modified field`