BUGS

** Open BUGS **
- Cutting the line at the top of the screen recenters to center (71)
- Segfault editing at COLS presumably due to new color syntax 
  highlighting (80)

** Fixed BUGS **
- Marked cutting sometimes leaves a newline in the file unintelligently,
  such as when all of a line is selected but the mark doesn't proceed to
  the new line. (8) [FIXED/IRRELEVANT]
- Certains are not lined up properly when there are tabs in them at 
  certain col values. (9) [FIXED]
- edit_refresh() and update_line() do not handle selecting text when the
  cursor is beyond COLS (10) [FIXED]
- No way to do a replace with the empty string (11) [FIXED, yay!]
- Spelling support is not elegant like Pico's integration of the 'spell'
  program.  nano only uses ispell (for now) (12) [FIXED]
- Moving to the end of a line when close to a multiple of COLS and at
  least COLS * 2 does not make the screen jump early like it would for
  if we were around COLS (bugs in edit_refresh, update_line) (13)
  [FIXED, mostly]
- When at the very bottom of the edit window, do_wrap() goes berserk and
  puts the cursor somewhere bad; subsequent keystrokes crash the program
  (14) [FIXED, mostly]
- Doing a replacement of a substring of the replace string (e.g.
  replacing "ed" with "fred" causes an infinite loop). (15) [FIXED]
- Cutting a file with marked text and both marker ends on the same line
  causes a random segfault (16) [FIXED]
- Cutting more than one line resets the KEEP_CUTBUFFER flag in 0.9.3.
  Error is in bitwise assignment (nano.h). (17) [FIXED]
- The wrapping code does not work right for lines like the following:
 *                                                                     *
  Error is in do_wrap, must be rewritten. (18) [FIXED]
- nano fails to follow symlinks, even though -l isn't being used (20).
  [Bug in global flag init, FIXED]
- When using --help or --version, the SIGINT character gets lost. (21)
  [FIXED]
- edit_refresh() and update_line() (and related functions) have 
  trouble when a tab is the character that is the boundary at COLS (23)
  [FIXED]
- There is an off-by-one error in keeping track of totsize.  It is
  caused by the fact that we count the newline at the end when we read
  in a file but we do not, in fact, display this newline.  This should
  go away implicitly when the "Magic Line" returns, but it is noted here
  for documentation's sake. (24) [FIXED]
- In replace with completely adjacent strings, every other search string
  will be missed (try replace ':' in a string like ':::::::::') (28)
  [FIXED]
- If nano opens a read-only file with the --tempfile option, and a
  change is made to the buffer, and a write is attempted, nano will not
  give you a dialog about not being able to write the file, and it will
  not let the user exit (29, discovered by Joshua Jensen) [FIXED]
- Using nano -k, marked text is not cut properly. (31) [FIXED]
- Invoking -t or -k has the effect of invoking both options. (32)
  [FIXED]
- totsize becomes incorrect after word-wrapping (25) [FIXED]
- Wrapping a line with autoindent mode sometimes causes a segfault (19)
  [FIXED]
- When inserting files, the display sometimes fails to display properly
  until a pageup/down occurs (22) [FIXED]
- In search/replace code, there is too much refreshing in bottomwin (26)
  [FIXED]
- In replace, there is no way to accept the default replace string. (27)
  [FIXED]
- Using nano -t, user can not exit until a filename is given via ^O.
  (30) [FIXED]
- totsize problems still abound in do_justify (33) [FIXED]
- Using -k, cut text is not pasted properly. (34) [FIXED].
- Using -k, pasted text is not updated properly if it goes beyond
  editbot. (35) [FIXED]
- Doing a cut with -k can screw up the filestruct; fault is in cutting
  code. (36) [FIXED]
- Hitting enter on the magic line makes new lines, but they are not
  written out to disk when saved..... (37). [FIXED]
- Page up and page down do not work the same way as in Pico (# of
  lines). (38) [FIXED]
- When doing a search and the marker is set, the screen does not always
  properly update the inverted text (39). [FIXED]
- Searches for a string that only exists on one line multiple times will
  fail after finding the last occurrence (discovered by Ken Tyler) (40).
  [FIXED]
- Meta-Z is currently broken to toggle suspend.  I guess I still don't
  know signals very well =-) (41) [FIXED].
- Unable to cut the entire file using the marker (discovered by Ken
  Tyler) (42). [FIXED]
- The keypad does not work when nano runs in the Gnome terminal (43).
  [FIXED]
- When reading in a file, if the file is a directory, the contents of
  the file being edited are blown away (discovered by  Chris Pimlot)
  (44). [FIXED]
- In certain terms, nano will leave a "ghost" of screen upon exit when
  called from inside mutt (among other settings) (45). [FIXED]
- In replace, hitting the Goto line shortcut key does nothing after a
  search string is entered (discovered by Rocco Corsi) (46) [FIXED].
- When typing in a string in search or replace and hitting
  CASE_SENSITIVE or the other search string, the current string edit is
  blown away in favor of the last stored search (47) [FIXED]
- If nano fails to open a file when it starts up, doing almost anything
  causes a segfault (discovered by Ben Roberts) (48). [FIXED]
- In certain terminals, nano would not work properly with keypad().
  Turned out to be the silly timeout(0) call, which is completely
  unneeded, anyway. (49) [FIXED]
- With less than a page of text, doing a page down will move the current
  line to the top of the screen, which it shouldn't do. (50) [FIXED]
- With PDCurses, running Meta-X turns off the keypad. (51) [FIXED]
- Resizing the window completely screws up the display if in any other
  mode than normal editing (help screen, search and replace, file
  browser...) (52) [FIXED]
- Alt speller argument (-s, --speller) does not take a string argument
  of more than one word. (53) [FIXED].
- Cut to end cutting (-k) causes segfaults (try cutting "- Backup making
  (filename~)?" line in TODO file) (discovered by
  higuita@cadernoverde.com) (54) [FIXED].
- When using autoindent (-i), wrapped text does not get autoindented
  (55, discovered by Mark Senior) [FIXED].
- When using -R (regex) and -p (Pico mode), subsequent searches after
  the first fail if no string is entered (56) [FIXED].
- Page down on a file of editwinrows fails (again). Reported by Ryan
  Krebs (57) [FIXED].
- File browser aborts on Solaris in qsort() call. (Reported by Matthias
  Andree) (58) [FIXED].
- Can modify the current file in view mode with ^W^R (discovered by
  Rocco Corsi) (58) [FIXED].
- When page up is used after two page downs, the screen doesn't update
  properly (discovered by David Lawrence Ramsey) (59) [FIXED].
- On BSD systems, marked cutting and paste often screws up the last line
  in the cutbuffer (discovered by Barry Pederson) (60) [FIXED]
- Blank lines are not kept when cutting with -k (discovered by Rocco)
  (61) [FIXED].
- nano will not suspend properly inside of mutt (62) [FIXED].
- When switching from Pico mode to normal mode, the previous search is
  not displayed until cancelling the search (63) [FIXED].
- If you change search options but don't change the search string in
  normal mode, hitting Enter causes the search/replace to abort (64)
  (Jordi Mallach) [FIXED].
- Cutting one line of text causes the screen to recenter the line
  (reported and fixed by David Lawrence Ramsey) (65) [FIXED].
- When cutting marked text including the bottom of the file, a new
  "magic line" is not created" (reported by David Lawrence Ramsey, fixed
  by David Lawrence Ramsey & Chris) (66) [FIXED].
- ^C does not work after a suspend in tcsh (discovered by Trevor Cordes)
  (68) [FIXED].
- Home and End control keys (^A, ^E) do not always work in filename
  prompt (bug found by Ian Turner) (69) [1.0 series only] [FIXED].
- Trying to insert a file of 0 bytes will hang nano (70) [FIXED].
- Meta-space and ^Space are not yet handled  in new backend code (77) [FIXED]
- F-keys do not work with new shortcut backend (72) [FIXED]
- Many menu items are probably missing from the new backend (73) [FIXED]
- Need a 'check-vitals-mapped' function to check that the end used didn't
  unbind all the keys for exit or cancel before starting up the editor (74) [FIXED]
- Browser and prompt code still implement old switch on raw key input instead
  of new if block for subnfunc values (75) [FIXED]
- New backend code probably does not compile under anything but default options, 
  if that (76) [FIXED]
- -enable-tiny does not work with new code (78) [FIXED]
- If a user only binds meta sequences to a function like left, right
  page up/down, insert, and unbinds all other control and F keys for it,
  nano will do the wrong thing when reading the key which is normally
  assigned to it (79 - may not be worth fixing) [FIXED]


$Id$