tmate/TODO
2014-06-21 19:45:52 +01:00

150 lines
6.6 KiB
Plaintext

- command bits and pieces:
* why doesn't command-prompt work if made read-only?
* allow multiple targets: fnmatch for -t/-c, for example detach all
clients with -t*
* add -c for new-session like new-window
* attach should take a pane and select it as well as attaching
* ' and " should be parsed the same (eg "\e" vs '\e') in config
and command prompt
* last-pane across sessions
* exact match operator for targets (or break the substring match
and require eg x* instead of just x)
- make command sequences more usable
* don't require space after ;
* options for error handling: && and ||?
- options bits and pieces:
* set-remain-on-exit is a complete hack
* way to set socket path from config file
- format improvements:
* last bits of status_replace into formats?
* option to quote format (#{session_name:quoted})
* formats need conditions for >0 (for #P)
* some way to pad # stuff with spaces, #!2T maybe
* last window update time and format for it
* formats to show if a window is linked into multiple sessions, into
multiple attached sessions, and is the active window in multiple
attached sessions?
- choose mode improvements:
* choose-pane command (augment choose-tree to do this?)
* choose-mode and copy-mode are very similar, make choose-mode a subset?
* flag to choose-* for sort order
* choose mode would be better per client than per window?
* two choices (first one then second, for swap-pane and join-pane)
- improve monitor-*:
* straighten out rules for multiple clients
* think about what happens across sessions
* monitor changes within a region
* perhaps monitor /all/ panes in the window not just one
- improve mouse support:
* bind commands to mouse in different areas?
* more fine-grained options?
* commands executed when clicking on a pattern (URL)
* mouse-select-pane will screw up with !MODE_MOUSE_STANDARD (it sets
the flag on w/o checking the others before calling tty_update_mode)
* mouse can be like normal key bindings?
- {button-{1,2,3},wheel-{up,down}}-{status,pane,border} and
drag-{start,end}-{status,pane,border} plus the modifiers
- resize and copy can be special cases - once you call something
like copy-mode -M or resize-pane -M to start the drag, it tracks
mouse until you call -m to stop the drag. or just keep drags
entirely special?
- what happens with stuff that wants mouse inside? especially for
pane clicks which need to run command AND pass event through
(like mouse-select-pane). maybe just a flag to say whether it
always runs or only if pane hasn't taken mouse? or it could be
eg bind Button1Pane "select-pane -t=; send-keys -Mt='
- also need a) some way to tell commands bound to key which
window or pane the mouse key binding applies to (maybe a new
special char in target, or pass targets through formats?) b) a
way to bind repeat count to mode keys so that wheel up/down can
do multiple lines c) send-keys -M to pass a mouse event through?
- what does the mouse->KEYC_* conversion and find-the-pane bit?
server_client_handle_key?
- hooks!
- warts on current naming:
* display-time but message-fg/bg/attr
* list-* vs show-*
* split-window -> split-pane??
- better UTF-8 support:
* window names and titles
* message display
* prompt input
* multibyte key input
- copy/paste improvements:
* incremental searching
* append to buffer
* paste w/o trailing whitespace
* command to toggle selection not to move it in copy-mode
- layout stuff
* way to tag a layout as a number/name
* maybe keep last layout + size around and if size reverts just put it
back
* revamp layouts: they are too complicated, should be more closely
integrated, should support hints, layout sets should just be a
special case of custom layouts, and we should support panes that are
not attached to a cell at all. this could be the time to introduce
panelink to replace layout_cell
* way to set hints/limits about pane size for resizing
* panning over window (window larger than visible)
* a mode where one application can cross two panes (ie x|y, width =
COLUMNS/2 but height = ROWS * 2)
- terminfo bits
* use a better termcap internally instead of screen, perhaps xterm
* use screen-256color when started on 256 colour terminal?
* need a tmux terminfo entry to document the extensions we are using in
upstream terminfo
- code cleanup
* instead of separate window and session options, just one master
options list with each option having a type (window or session), then
options on window, on session, and global. for window options we look
window->session->global, and for session we look session->global
* the way pane, window, session destroy is handled is too complicated
and the distinction between session.c, window.c and server-fn.c
functions is not clear. could we just have kill_pane(),
kill_window(), unlink_window(), kill_session() that fix up all data
structures (flagging sessions as dead) and return a value to say
whether clients need to be checked for dead sessions? sort of like
session_detach now but more so. or some other scheme to make it
simpler and clearer? also would be nice to remove/rename server-fn.c
* more readable way to work out the various things commands need to
know about the client, notably:
- is this the config file? (cmdq->c == NULL)
- is this a command client? (cmdq->c != NULL &&
cmdq->c->session == NULL)
- is this a control client?
- can i do stdin or stdout to this client?
or even guarantee that cmdq->c != NULL and provide a better way to
tell when in the config file - then we use cmdq->c if we need a
client w/o a session else cmd_current_client
* optimize pane redraws, 20120318184853.GK10965@yelena.nicm.ath.cx
- miscellaneous
* way to keep a job running just read its last line of output for #()
* link panes into multiple windows
* live update: server started with -U connects to server, requests
sessions and windows, receives file descriptors
* there are inconsistencies in what we get from old shell and what
comes from config for new sessions and windows. likewise, panes and
jobs and run-shell and lock command all start with slightly different
environments
* multiline status line?
* bind commands to key sequences -- make it so ALL keys go through a
table, first an implicit table in which C-b is the only default
binding to a command that says "next key from $othertable" and so
on. means -n can go away as well
* customizable command aliases
* any remaining clients in wait-for should be woken when server exits