diff options
author | Adrian Robert <Adrian.B.Robert@gmail.com> | 2008-07-15 18:15:18 +0000 |
---|---|---|
committer | Adrian Robert <Adrian.B.Robert@gmail.com> | 2008-07-15 18:15:18 +0000 |
commit | edfda78355c5528eee489fa8a7f9c73bf8e734f2 (patch) | |
tree | 78d2414d9791e1efc17ec9b35b438ae35602340a /nextstep/FOR-RELEASE | |
parent | 1391cd548782097e34d7856ec4f20ca90bdf2c26 (diff) | |
download | emacs-edfda78355c5528eee489fa8a7f9c73bf8e734f2.tar.gz |
merging Emacs.app (NeXTstep port)
Diffstat (limited to 'nextstep/FOR-RELEASE')
-rw-r--r-- | nextstep/FOR-RELEASE | 216 |
1 files changed, 216 insertions, 0 deletions
diff --git a/nextstep/FOR-RELEASE b/nextstep/FOR-RELEASE new file mode 100644 index 00000000000..c4405143275 --- /dev/null +++ b/nextstep/FOR-RELEASE @@ -0,0 +1,216 @@ +-*- outline -*- + +* NON-SPECIFIC + +** Remove Feval calls relating to insert working text in isearch mode. + +** free_frame_resources, face colors + +** look at nsfont_char_width and nsfont_string_width utils (Dan C.) + +** config improvements (Dan C.) + +* Mac-related: + +** open file:/// URLs + +** finish handle terminate request (user logout) + +** put frame autopositioning into C code somewhere -- if loc = same, offset + +** automap ctrl-mouse-1 to mouse-3 + +** deal with Finder aliases somehow + +** Ctrl-F2 won't pull up menus + + +* Cursor: + +** ns_cursor_blink_rate: when set in preferences, somehow save the option (but + calling custom-save-all from this causes error) + +** cursor nonerase on certain Leopard and Tiger installations + +** if I do: (mouse-avoidance-mode 'banish) and then minimize Emacs, it pops back immediately. + Also, it insists on being on top of other windows if they overlap. + +** cursor-over bugs w/some scripts (move around in HELLO to see) + + +Keyboard: + +On a German (PowerBook?) keyboard alt-` produces the correct ˚, without alt modifier only the message “<S-268632064> is undefined” is produced. Peter Maurer’s Key codes shows: + * Modifier Change: ⇧ 131330/0x20102 + * Key Down/Up event: ⇧ 24/0x18 +[note, this is += key on German KB setting on US keyboard] +[unable to reproduce w/German KB setting -- need German laptop?] + +** numeric keysetting bug + + +* Other: + +** better recog of unicode scripts / Greek / composition + +** Ctrl-g not picked up in all situations get --experimental-ctrl-g in 'compile' working better + +** Exec-dump not working on GNUstep (so slow startup). + +** occasional size too large for screen errors + +** undo for color-drag face customization + + + + +--------- +* DONE + +Try font-backend mechanism (see font.h). + +Auto font substitution? + +ns-option-modifier (alias for ns-alternate-modifier) + +ns-control-modifier, ns-function-modifier (new) + +- Frame bugs: + - create frame on second monitor fails (due to frame-parameters). + +toolbar bugs + +fix cursor disappears on italic text bug + +fix scrollbar erase on Ctrl-l bug + +Menubar emacs-21 compatibility. Make this and command-is-meta + switchable at startup. Switch mouse handling to 3-button compatible. + +mouse button normalization + + +[rc1 released] + + +shift-alt-h does produce Ó (according C-h k), but this glyph is not +always shown completely, it seems to depend on the font used + +[width estimation still off; this character is 211 = 0xD3] + +Another attempt at composite rendering. + +shift-alt-6 does not produce ^ (dead key) but &. +[adding alt continues to pick up the '&' interpretation, but in other + apps this suddenly changes it to a dead key] + +- italic in Courier is still cut off + +- dead keys end isearch w/o allowing typing +[need to insert to minibuffer if that is active] + +When using the Dvorak with Qwerty keyboard shortcuts keyboard layout +(hereon in referred to as Dvorak-Qwerty), keyboard shortcuts +erroneously use the Dvorak layout. e.g. The Qwerty H key becomes D in +Dvorak, but with Dvorak-Qwerty the same key should be used when +hitting Cmd-H to hide an application. In Emacs.app with a +Dvorak-Qwerty layout I must hit Cmd-J (Qwerty J becomes H in Dvorak) +to hide the app. + +- menus + +- consolidate FRAME_INTERNAL_BORDER extension hacks + +- mouse-1 activation problems (buttons, etc.): check keyboard.c HAVE_NS sections + +- clean up bg fill code in font->draw(), and change conditions on when called with bg=yes + +- clean up / optimize the metrics caching -- consider reducing memory load + +- non-encapsulated build + +- test color-drag face customization (accuracy and saving) + +- look at ns-mark-nav replacement + +- gnustep + +- documentation: + - "stipple" images, alpha, color-drag, etc. + - make sure everyone finds it + + +[rc2 released] + +- recheck font width determination for variable-width non-wide fonts + +- resize-drag + +- colored bitmaps (use NSCalibratedRGBColorspace for the bitmap and + add a method setBitmapForegroundColor:color to twiddle the bits + +- resize title printing bug + + +[rc2a released] + +- naming: cursor "line" -> "bar", "bar" -> underscore + +- incorporate Pete's locallisp patch + +- document needed sudo for compile --shared-lisp + + +[patch 20071010_rc2a released] + +- clean up ns-menu-bar.el and try folding into ns-win.el + +- fix jaggedy startup screen w/black bkgrd + +- Pete's fringe weirdness + +- iconification: use shrunk image for non-finder types + +- toolbar updating + + +[patch 20071015_rc2a released] + +- multi-TTY merge + +- terminal mode weird colors (real fix) + +- Frame bugs: + - a full-height frame will be shrunk on defocus, focus + [actually must be over-height, but it's a little overzealous in the + shrinking; probably happens in x_set_window_size] + - adding toolbar on full-height frame requires resize to avoid scrambling + +- Buffers menu brokenness + +- inability to use ARGB colors problem + +- cut buffer in TTY + + +[rc3 released] + +- drop MULTI_KBOARD + +- mic-paren removal (also see http://www.emacswiki.org/cgi-bin/wiki/mic-paren.el) + +- describe-key problem + +- drop need for preload advice (print-buffer workaround) + +- new frame placement ignores parameters [cannot replicate] + +- white text on background alpha + +- Multi-TTY: switching input between tty and GUI frames still needs work. + Once ns_select() and ns_read_socket() are being called, the regular terminal + input route seems ineffective. Furthermore NSApp activate events are not + picked up when attempting to read in terminal mode. + (Tests: make-frame-on-tty /dev/ttyp#, make-frame-on-display) + +- remove ns-yes-or-no-p, try again to improve looks of regular dialog in nsmenu.m |