org-mode/etc/ORG-NEWS

ORG NEWS                                                       -*- org -*-

* Version 7.8

** Summary

- Jambunathan's ODT exporter is now part of Org's core.
- Nicolas' new export engine is now part of =contrib/=
- Standard code block keywords
- 10 new committers
- Many bugfixes

See below for details.

** New committers who signed the FSF copyright assigment

Here is the list of new contributors who signed the FSF papers
since Org 7.7 - welcome, and thanks for your contributions!

- Andreas Leha
- Christian Moe
- Julian Gehring
- Max Mikhanosha
- Michael Brand
- Niels Giessen
- Pieter Praet
- Sergey Litvinov
- Thomas Holst
- Thorsten Jolitz

** The ODT exporter is now part of Org's core

- Full refresh of the OpenDocument Text section in the
  manual.  All new features listed below are fully-documented.
- Associate custom styles on per-file basis using
  =#+ODT_STYLES_FILE:= directive.
- Fontify code listings using an enhanced version of
  =htmlfontify.el= and generate line numbers natively.
- Embed MathML and OpenDocument formula files.
- Use LaTeX to MathML converter -- say MathToWeb
  (http://www.mathtoweb.com/) -- for handling LaTeX Math
  fragments.
- In tables, use column width cookies to control relative width
  of columns.
- Also for tables, you can specify custom styles using
  =#+ATTR_ODT:= lines.
- Lots of bug fixes.

*Experimental*

The following features are /experimental/.  These features are
specific to the ODT export engine and their implementation and
usage could change considerably in future versions.

- Support for list tables -- see this [[http://lists.gnu.org/archive/html/emacs-orgmode/2011-09/msg00017.html][message on the list]].
- Support for annotation blocks -- see this [[http://lists.gnu.org/archive/html/emacs-orgmode/2011-10/msg01251.html][message on the list]].

Special thanks to Jambunathan for his work and his patience
through the process of integrating this vital contribution into
Org's core.

** New export engine by Nicolas

See the comment sections in org-element.el and in org-export.el
(in =contrib/lisp/=).  Also check the experimental LaTeX exporter
using in =EXPERIMENTAL/org-e-latex.el=.

Check Nicolas' announcement [[http://comments.gmane.org/gmane.emacs.orgmode/49416][on the list]] and hack around!

Thanks a lot to Nicolas for this great and promising achievement.

** Incompatible changes

*** Standardized code block keywords

Following a round of on-list discussion, many code block synonyms
have been removed.  You can safely move forward the following
syntax:

- call lines are specified with #+call:
- code blocks are named with #+name:
- results are named with #+name:, however results generated by a
  code block may still be labeled with #+results:, and tables
  named with #+tblname: will be considered to be named results

The following function may be used to update an existing Org-mode
buffer to the new syntax:

#+begin_src emacs-lisp
  (defun update-org-buffer ()
    "Update an Org-mode buffer to the new data, code block and call line syntax."
    (interactive)
    (save-excursion
      (flet ((to-re (lst) (concat "^[ \t]*#\\+" (regexp-opt lst t)
                                  "\\(\\[\\([[:alnum:]]+\\)\\]\\)?\\:[ \t]*"))
             (update (re new)
                     (goto-char (point-min))
                     (while (re-search-forward re nil t)
                       (replace-match new nil nil nil 1))))
        (let ((old-re (to-re '("RESULTS" "DATA" "SRCNAME" "SOURCE")))
              (lob-re (to-re '("LOB")))
              (case-fold-search t))
          (update old-re "name")
          (update lob-re "call")))))
#+end_src

*Note*: If an old version of Org-mode (e.g., the one shipped with
Emacs) is installed on your system, many of the important
variables will be pre-defined with a =defvar= and *will not* have
their values automatically updated -- these include the
following:

- =org-babel-data-names=
- =org-babel-result-regexp=
- =org-babel-src-block-regexp=
- =org-babel-src-name-regexp=
- =org-babel-src-name-w-name-regexp=

It may be necessary to either remove the source code of older
versions of Org-mode or to explicitly evaluate the ob.el file.

*** Removed the =#+BABEL:= keyword

Code block header arguments may now be specified using the
standard =#+Property:= keyword.  See [[#accumulating-property-keywords][Property names ending in =+=
accumulate]] for more information on the extended property syntax.

*** Deleted =org-mode-p=

This function has been deleted: please update your hacks if you
relied on it.

** New features and user-visible improvements

*** Day names are optional when you type timestamps manually

You can type "<2011-12-12>" or [2011-12-12] and modify this timestamps with
S-<right> like other timestamps.  Thanks to Carsten for implementing this.

*** Improved filtering through =org-agenda-to-appt=

The new function allows the user to refine the scope of entries
to pass to =org-agenda-get-day-entries= and allows to filter out
entries using a function.

Thanks to Peter Münster for raising a related issue and to
Tassilo Horn for this idea.  Also thanks to Peter Münster for
[[git:68ffb7a7][fixing a small bug]] in the final implementation.

*** =org-agenda-date-later= jumps to today when modifying past date

The command =org-agenda-date-later= (<S-right> in the Agenda
Mode) directly jumps to today.  Thanks to Carsten for
implementing this.

*** Use prefix arg 0 to inhibit note taking for TODO change

See Carsten's comment: "Sometimes I want to quickly make a few TODOs done
in the agenda and I want to bypass the note taking I have normally set
up.  With this change, I can press `0 t d' in the agenda to do this."

Thanks to Carsten for implementing this.

*** Various improvements to =org-indent.el=

Nicolas Goaziou worked on refactoring and improving
=org-indent.el=, which should now be faster.  It also better
combines with =visual-line-mode=.

*** Property names ending in =+= accumulate

This results in the following behavior.

#+begin_src org
  ,#+property: var  foo=1
  ,#+property: var+ bar=2

  ,#+begin_src emacs-lisp
  ,  (+ foo bar)
  ,#+end_src

  ,#+results:
  ,: 3

  ,#+begin_src emacs-lisp
  ,  (org-entry-get (point) "var" t)
  ,#+end_src

  ,#+results:
  ,: foo=1 bar=2

  ,* overwriting a file-wide property
  ,  :PROPERTIES:
  ,  :var:      foo=7
  ,  :END:

  ,#+begin_src emacs-lisp
  ,  foo
  ,#+end_src

  ,#+results:
  ,: 7

  ,#+begin_src emacs-lisp
  ,  (org-entry-get (point) "var" t)
  ,#+end_src

  ,#+results:
  ,: foo=7

  ,* appending to a file-wide property
  ,  :PROPERTIES:
  ,  :var+:      baz=3
  ,  :END:

  ,#+begin_src emacs-lisp
  ,  (+ foo bar baz)
  ,#+end_src

  ,#+results:
  ,: 6

  ,#+begin_src emacs-lisp
  ,  (org-entry-get (point) "var" t)
  ,#+end_src

  ,#+results:
  ,: foo=1 bar=2 baz=3
#+end_src

*** =org-agenda-custom-commands= has a default value

This option used to be `nil' by default.  This now has a default
value, displaying an agenda and all TODOs.  See the docstring for
details.  Thanks to Carsten for this.

*** =outline-demote/promote= points to =org-demote/promote-subtree=

Users who use this outline commands in =outline-mode= will want
them to behave the Org way in Org.  Thanks to Michael Brand for
the suggestion.

*** New escape characters for =org-log-note-headings=

The option =org-log-note-headings= now supports %d and %D for
inactive and active timestamps.  This affect the behavior of the
command =org-store-log-note=.  Thanks to John J Foerch for a
patch for this.

*** New command =org-table-transpose-table-at-point=

See the docstring.  This hack from Juan Pechiar is now part of
Org's core.  Thanks to Juan!

*** Allow recursive edit of minibuffer

You can now use the command =C-c != (=org-time-stamp-inactive=)
when prompted e.g. for a link name.  This is because the central
function =org-completing-read= now support recursive edit.  Only
=C-c != is available at the moment, but if you can think of a
useful command to use, please let me know.

Thanks to Skip Collins for the idea and to Nick Dokos for the
implementation.

*** Allow dynamic construction of the publishing destination

Thanks to Kai Tetzlaff for adding this.

*** Set the capture default time to the prompt time

In the file+datetree+promt target type, the user is being asked
for a date, where to file an entry.  In the template, he can use
the escape placeholders for active and inactive time stamps.  So
far, these were filled with todays date.  This behavior changed
so that, also at %t and %u, the date to use is the one entered at
the prompt.

Reported by Erik Hetzner and fixed by Carsten.

*** =org-agenda-skip-additional-timestamps-same-entry= defaults to `nil'
*** org-sudoku.el -- small sudoku solver

Carsten wrote org-sudoku.el, a little sudoku solver, which lives
in the contrib/ directory.

From [[http://permalink.gmane.org/gmane.emacs.orgmode/45474][his message]]: "my daughter got stuck with a couple of SUDOKU
puzzles during the vacation (where wh had no internet
connection), so I hacked a small SUDOKU solver that reads a 9x9
Org table and solves it as a sudoku puzzle.  A little silly, but
maybe fun for someone - I have pushed it into the contrib/lisp
directory."
** Code Block related features and improvements
*** Added [[http://cran.r-project.org/web/packages/tikzDevice/index.html][tikzDevice]] support to ob-R.el
*** Support for Pico Lisp code blocks

Thanks to Thorsten Jolitz Pico Lisp code blocks are now
supported.  See =ob-picolisp.el= for more information.

*** Extension to the =eval= code block header argument

The =eval= code block header argument now accepts four (six)
possible values the meaning of which is shown in the table below.

| never or no               | The code block will not be evaluated under any     |
|                           | circumstances.                                     |
| query                     | Evaluation of the code block will require a query. |
| never-export or no-export | The code block will not be evaluated during export |
|                           | but may still be called                            |
| query-export              | Evaluation of the code block during export will    |
|                           | require a query.                                   |

*** Update of intermediate results during code block evaluation

When set to t, the new =org-babel-update-intermediate= variable
will update in-buffer results for code blocks which are evaluated
in the resolution of a variable reference.  This can be used to
ensure that the latest returned results are always shown in
buffer.

*** Support for Fortran code block is now in the core

Thanks to Sergey Litvinov for contributing this support.

*** The =sbe= function allows header argument specification

If first variable is a string and not a cons cell, then interpret
it as a string of header arguments to be passed to the code
block.

*** Support for Shen code blocks

See http://www.shenlanguage.org/ for information on Shen.

A major mode for shen code blocks is available through the GNU
ELPA.

** New options and faces

*** New option =org-export-html-headline-anchor-format=

Format for anchors in HTML headlines.

It requires to %s: both will be replaced by the anchor referring
to the headline (e.g. "sec-2").  When set to `nil', don't insert
HTML anchors in headlines.

This was requested by Alan L Tyree.

*** New option =org-table-formula-field-format=

Format for fields which contain the result of a formula.
For example, using \"~%s~\" will display the result within tilde
characters.  Beware that modifying the display can prevent the
field from being used in another formula.

Thanks to Dov Grobgeld for this idea.

*** New option =org-loop-over-headlines-in-active-region=

When set to `t', some commands will loop over the active
region.  Currently, =org-schedule= and =org-deadline= uses this
option: pressing =C-c C-s= or =C-c C-d= when the region is active
will let you schedule/deadline all the visible headlines in the
region.

Thanks a lot to David Maus for implementing this.

*** New option =org-catch-invisible-edits=

This option makes it possible to check what is the right thing to do before
editing invisible regions.  Here are the possible values of this option:

: +nil              Do not check, so just do invisible edits.
: +error            Throw an error and do nothing.
: +show             Make point visible, and do the requested edit.
: +show-and-error   Make point visible, then throw an error and abort the edit.
: +smart            Make point visible, and do insertion/deletion if it is
:                   adjacent to visible text and the change feels predictable.

See the docstring for more details.

Thanks to Carsten for implementing this.

*** New option =org-export-latex-table-caption-above=

Let the user place a caption above its table in LaTeX.
Thanks to Thomas Dye for a patch to this effect.

*** New option =org-agenda-follow-indirect=

By setting `org-agenda-follow-indirect' to a non-nil value,
`org-agenda-follow-mode' will use an indirect buffer to display
only the current item, rather than the whole agenda file in which
it lives.

Thanks to Dave Abrahams for implementing this.

*** New option =org-refile-active-region-within-subtree=

Non-nil means also refile active region within a subtree.

By default =org-refile= doesn't allow refiling regions if they
don't contain a set of subtrees, but it might be convenient to do
so sometimes: in that case, the first line of the region is
converted to a headline before refiling.

Thanks to Jeff Horn for raising the issue of refiling a list
item, and to Nicolas Goaziou, Nick Dokos and Suvayu Ali for
discussing them patiently.

*** New option =org-latex-to-mathml-convert-command=

This option lets you specify a way to convert LaTeX fragments to
MathML.  See also =org-latex-to-mathml-jar-file= and the
docstring of =org-create-math-formula=.

Thanks to Jambunathan K for implementing this.

*** New option =org-properties-postprocess-alist=

See its docstring:

: Alist of properties and functions to adjust inserted values.
: Elements of this alist must be of the form
:
:   ([string] [function])
:
: where [string] must be a property name and [function] must be a
: lambda expression: this lambda expression must take one argument,
: the value to adjust, and return the new value as a string.
:
: For example, this element will allow the property "Remaining"
: to be updated wrt the relation between the "Effort" property
: and the clock summary:
:
:  (("Remaining"
:      (lambda(value)
:        (let ((clocksum (org-clock-sum-current-item))
:              (effort (org-duration-string-to-minutes
:                      (org-entry-get (point) "Effort"))))
:          (org-minutes-to-hh:mm-string (- effort clocksum))))))

This is inspired by a request from Pascal Mattia.

*** New options =org-habit-today-glyph= and =org-habit-completed-glyph=

This gives you control over the character used for displaying
today (default is `!') and days on which a task has been
completed (default is `*').  Thanks to John Wiegley for this.

*** New option =org-bibtex-type-property-name=

Configurable property name for bibtex entry types.  Thanks to
Eric Schulte for this.

*** New face =org-agenda-filter-tags=

This face is used for highlighting the tag(s) filter in the
modeline.  Thanks to Sébastien Vauban for this idea and its
implementation.

*** New faces =org-agenda-calendar-event= and =org-agenda-calendar-sexp=

=org-agenda-calendar-event= is the face used to show events and
appointments in the agenda, and =org-agenda-calendar-sexp= the
one used to show events computed from a S-expression.

Thanks to Sébastien Vauban for this addition.

** Important bugfixes

*** Respect =org-export-with-tags= when exporting a subtree

Thanks to Suvayu Ali for spotting this problem and fixing it.

*** Fix XEmacs compatibility issue when creating an indirect buffer

Thanks to Michael Sperber for this fix.

** Testing

*** New test function =org-test-with-temp-text-in-file=

* Version 7.7

** COMMENT Incompatible changes
** New features and user-visible improvements

*** New command =org-copy-visible= (=C-c C-x v=)

This command will copy the visible text in the region into the
kill ring.

Thanks to Florian Beck for this function and to Carsten for
adding it to org.el and documenting it!

*** New hook =org-ctrl-c-ctrl-c-final-hook=

This hook will be called when nothing special can be performed by
the busy =C-c C-c= key.  Thanks to Paul Sexton for this idea!

*** Allow relative time when scheduling/adding a deadline

You can now use relative duration strings like "-2d" or "++3w"
when calling =org-schedule= or =org-deadline=: it will schedule
(or set the deadline for) the item respectively two days before
today and three weeks after the current timestamp, if any.

You can use this programmatically: =(org-schedule nil "+2d")=
will work on the current entry.

You can also use this while (bulk-)rescheduling and
(bulk-)resetting the deadline of (several) items from the agenda.

Thanks to Memnon Anon for a heads up about this!

*** New functions: =org-todo-yesterday= and =org-agenda-todo-yesterday=

This is useful when you need to mark things done yesterday.

Thanks to Max Mikhanosha for this patch.

*** =org-set-property= defaults to the last interactively modified property

When setting a property with =C-c C-x p=, it will offered the
last interactively modified property as a default choice.  If
this command is called on a property line, the property in this
line will take precedence over the last set property.

*** Clock: Allow synchronous update of timestamps in CLOCK log

Using =S-M-<up/down>= on CLOCK log timestamps will
increase/decrease the two timestamps on this line so that
duration will keep the same.  Note that duration can still be
slightly modified in case a timestamp needs some rounding.

Thanks to Rainer Stengele for this idea.

*** Clock: New function =org-clock-remove-empty-clock-drawer=

This function removes empty =CLOCK= drawers and has been added to
=org-clock-out-hook=: when clocking out, if no CLOCK log has been
inserted and the drawer is empty, the drawer will be removed.

*** Capture: new escape sequence =%F= for templates

Using =%F= in capture templates will insert the full path of the
file or the directory the capture mechanism was called from
(whereas =%f= only insert the filename.)

Thanks to Nicolas Goaziou for this change.

*** Agenda: new variable =org-agenda-bulk-custom-functions= for custom bulk functions

When using =org-agenda-bulk-action= in agenda view, the user
could already call custom functions by pressing =f= and entering
the function's name.

This variable lets the user add custom choices and reach them
more quickly.  Set the variable to an alist of keys (chars) and
functions, and these keys will be accessible through the
=org-agenda-bulk-action= interface.

If there is a conflict between custom keys and hardcoded choices,
the latter ones take precedence.

Thanks to Julien Cubizolles for triggering this idea.

*** Refile: exclude irrelevant headings

When refiling from an org-mode buffer, the current heading and
its subheadings will be excluded from the list of possible
targets.

This only works when =org-refile-use-cache= is =nil= and in
org-mode buffers, not in agenda buffers.

Thanks to Jason Dunsmore for this idea!

*** Lists: new variable =org-list-use-circular-motion=

This variable allows some commands to consider lists as cyclic
structures.  For example, when non-nil, moving past the last item
of a list with S-down will bring you back to the first one.

*** Lists: New variable =org-list-indent-offset=

This variable helps improving readability of sub-items by
increasing their indentation.  E.g., if =org-list-indent-offset=
is set to 2, you may see the following list:

#+begin_example
- First item
	  - Sub-item 1
	  - Sub-item 2
- Second item
#+end_example

*** Table: New flag ";t" to tailor the display of computed durations

See the new variable =org-table-duration-custom-format= and the
updated example from the manual:

|  Task 1 |   Task 2 | Total |
|---------+----------+-------|
| 3:02:20 | -2:07:00 |  0.92 |
#+TBLFM: @2$3=$1+$2;t

In this example, =0.92= is a fraction of hour, the default for
=org-table-duration-custom-format=.

Thanks to Daniel E. Doherty for discussions about this.

*** Table: display field's coordinates when editing it with =C-c `=

When editing a field with =C-c `=, the field's coordinate will
the displayed in the buffer.

Thanks to Michael Brand for a patch to this effect.

*** Babel: support for =java= code blocks

Evaluation of blocks of Java code is now possible.  Currently
only external evaluation of Java code is supported (i.e., no
session evaluation) and only the =:results output= results are
collected.  Code blocks are compiled to a Java class file which
are then evaluated and the values printed to STDOUT are returned.

Java code blocks required a =:classname= header argument which is
used by the compiler to name =.java= and =.class= files.

Example Java code block:
#+begin_src org
  ,#+begin_src java :classname myfirstjavaprog
  ,  class myfirstjavaprog
  ,  {
  ,      public static void main(String args[])
  ,      {
  ,          System.out.println("Hello World!");
  ,      }
  ,  }
  ,#+end_src

  ,#+results:
  ,: Hello World!
#+end_src

*** Babel: support for =fortran= code blocks by Sergey Litvinov

The =contrib/babel/langs/= contains the new library
=ob-fortran.el= by Sergey Litvinov.

Thanks to him for this addition.

*** Babel: support for the [[http://fomus.sourceforge.net/][fomus]] language by Torsten Anders

=ob-fomus.el= has been added to =contrib/babel/langs/=.

Thanks to Torsten Anders for contributing this functionality.

*** Publish: hide .orgx files and use theindex.org directly

When :makeindex is `non-nil' in the publishing options, Org will
export an index.  It populates the directory with .orgx files:
those files are now hidden (.file.orgx).  Also the index is
directly stored in the file =theindex.org=, not in the file
=theindex.inc= (which you can delete).

Thanks to Nathan Neff and Carsten for pointing at the problems
and solutions.

*** Export: new variable =org-export-html-divs= to allow custom divs

The default value of =org-export-html-divs= is =("preamble"
"content" "postamble")= and is used to define the =<div "%s">=
for the main structure of the exported HTML file.

Note that modifying the default value will break compatibility
with the =org-info.js= script.

Also note that the variable =org-export-html-content-div= will
still be checked for compatibility reasons but is not a custom
variable anymore.

Thanks to Sébastian Vauban for a preliminary version of this
patch.

*** Export: new variable =org-export-latex-quotes= to customize quotes in LaTeX export

This allows users to define what quotes they want to use as a
replacement of english double-quotes while exporting to LaTeX.

In particular, if you use the csquote package, you can configure
Org to output something like \endquote{some quoted text} instead
of "some quoted text".

Thanks to Frederik for bringing this issue up, and to Thomas S.
Dye, Nick Dokos and Stefan Nobis for elaborating this solution.

** Important bugfixes

*** Duration computations now work for complex ranges in tables

Thanks to Christian Moe for signaling this problem.

*** Handle recursive setup files correctly

Thanks to Stefan Vollmar for mentioning this problem.

** Details

*** =org-map-entries= can now have 'region as its scope

Thanks to David Maus for suggesting related improvements to
=org-map-entries=.  This one prepares the possibility of letting
some commands to loop over the active region.

*** org-depend.el: new chain-find-next trigger option

See the updated docstring of this file.

Thanks a lot to Max Mikhanosha for this!

*** Export: new experimental exporter to MoinMoin by Puneeth Chaganti.

The =EXPERIMENTAL/= directory contains a new file org-mm.el that
allows you to export an Org file to a MoinMoin file.

Thanks to Puneeth Chaganti for this addition.

*** Export: new default for =org-export-html-preamble=

The default value included the title.  It now defaults to the
empty string, as the title is hardcoded and included in the
"content" div.  Including the title here is necessary to let
=org-info.js= find it and display the page properly.

*** Export: new variable =org-lparse-use-flashy-warning= defaulting to =nil=

Flashy and verbose warnings while exporting to ODT have been
disabled.  Set this to =t= if you want it back.

*** Export: new default value for =org-export-latex-image-default-option=

This used to be =width=10em= and this is now width=.9\linewidth,
which makes more sense.

Thanks to Sebastien Vauban for this suggestion.

*** Export: allow =org-export-latex-href-format= to have only one "%s"

This is useful when you want to use \url{link} instead of the
default \href{link}{path}.

Thanks to Henri-Paul Indiogine for bringing this up.

** Known issues

*** The ODT exporter will choke when using some Org-defined strings

For example, using the hungarian version of the table of
contents, as defined in =org-export-language-setup=, the ODT
exporter will complain about a problem with the translated
string.

The workaround is to customize =org-export-language-setup=
and to use accents directly.

* Version 7.6

** Incompatible changes
** New features and user-visible improvements

*** Integration of Jambunathan's OpenDocumentText Exporter

**** Activation

Org-mode 7.6 supports exporting to OpenDocument Text (=odt=)
format using org-odt.el.  Depending on how you installed Org,
this module can be enabled in one of the following ways:

1. If you have downloaded the Org from the Web, either as a
   distribution =.zip= or =.tar.gz= file, or as a Git archive,
   enable the =odt= option in the variable =org-modules=.

2. If you are using Org-mode 7.6 that comes bundled with
   Emacs-24.0.50 (or future Emacs-24.1), then you can install the
   OpenDocumentText exporter using the package manager.  Check
   the list of available packages with =M-x list-packages= and
   install the =org-odt= package.

Thanks a lot to Jambunathan K for this great contribution.

**** Keybindings

The following interactive commands are provided:

1. =C-c C-e o= (=org-export-as-odt=): Export as an =odt= file.

2. =C-c C-e O= (=org-export-as-odt-and-open=): Export as an =odt=
   file and open the resulting file.

See the =contrib/odt/README.org= file for further details; you
may check in particular the commands =M-x org-lparse= and =M-x
org-export-convert=.

*** Ob-Lilypond -- new Babel language to allow score generation

ob-lilypond - an org-babel language, provided to allow LilyPond
music score generation, complete with optional auditioning via
midi, whilst leveraging the full power of org mode, and literate
programming.  See https://github.com/mjago/ob-lilypond for more
documentation.

Thanks to Martyn Jago for this addition.

*** Org-Bibtex -- major improvements

Provides support for managing bibtex bibliographical references
data in headline properties.  Each headline corresponds to a
single reference and the relevant bibliographic meta-data is
stored in headline properties, leaving the body of the headline
free to hold notes and comments.  Org-bibtex is aware of all
standard bibtex reference types and fields.

The key new functions are

- org-bibtex-check :: queries the user to flesh out all required
     (and with prefix argument optional) bibtex fields available
     for the specific reference =type= of the current headline.

- org-bibtex-create :: Create a new entry at the given level,
     using org-bibtex-check to flesh out the relevant fields.

- org-bibtex-yank :: Yank a bibtex entry on the kill ring as a
     formatted Org-mode headline into the current buffer

- org-bibtex-export-to-kill-ring :: Export the current headline
     to the kill ring as a formatted bibtex entry.

*** Spreadsheet computation of durations and time values

If you want to compute time values use the =T= flag, either in
Calc formulas or Elisp formulas:

| Task 1 | Task 2 |   Total |
|--------+--------+---------|
|  35:00 |  35:00 | 1:10:00 |
#+TBLFM: @2$3=$1+$2;T

Values must be of the form =[HH:]MM:SS=, where hours are
optional.

Thanks to Martin Halder, Eric Schulte and Carsten for code and
feedback on this.

*** Links within inlined footnotes.

It as also possible to have footnotes side-by-side correctly
exported. New variables =org-export-latex-footnote-separator=,
=org-export-html-footnote-separator= and
=org-export-docbook-footnote-separator= are used to separate them
in that case.

Fontification of footnotes is also more accurate.

*** New variable =org-export-with-tasks=

Non-nil means include TODO items for export.

This may have the following values:

- t                    include tasks independent of state.
- todo                 include only tasks that are not yet done.
- done                 include only tasks that are already done.
- nil                  remove all tasks before export
- list of TODO kwds    keep only tasks with these keywords

Thanks to Carsten for implementing this!

*** New variable =org-export-latex-timestamp-inactive-markup=

This variable allows the user to define the LaTeX markup for
inactive timestamps.  It defaults to the same markup than active
timestamps.  Thanks to Eric S Fraga for this patch.

*** New =org-default= face

=M-x customize-face RET org-default RET= will let you define the
default face for =org-mode= buffers.

*** Babel improvements
**** In line code block call syntax
It is now possible to call code blocks from within blocks of
prose.

The new syntax is exactly analogous to the existing =#+call:=
line syntax, only it may be present embedded in a block of prose
for example =call_double(num=8)= would call the =double= code
block assigning the =num= variable to the value =8=.

**** Optional variable names in code block calls
Variable names are now optional when passing variables to a code
block reference.  Un-named variables will be assigned in order as
shown below.

: #+source: minus
: #+begin_src emacs-lisp :var a=0 :var b=0
:   (- a b)
: #+end_src
:
: #+call: minus(a=8, b=4)
:
: #+call: minus(8,4)

**** Sub-tree ID as valid code block variable reference
It is now possible to assign the textual contents of an Org-mode
subtree to a code block variable using the ID of the subtree.
Both custom IDs and Org-mode IDs may be used.  For example;

#+begin_src org
  ,#+begin_src sh :var text=foo
  ,  echo "$text"|wc
  ,#+end_src

  ,#+results:
  ,: 8      58     415

  ,* example foo
  ,  :PROPERTIES:
  ,  :CUSTOM_ID: foo
  ,  :END:

  ,Lorem ipsum dolor sit amet, consectetuer adipiscing elit. Donec
  ,hendrerit tempor tellus. Donec pretium posuere tellus. Proin quam
  ,nisl, tincidunt et, mattis eget, convallis nec, purus. Cum sociis
  ,natoque penatibus et magnis dis parturient montes, nascetur ridiculus
  ,mus. Nulla posuere. Donec vitae dolor. Nullam tristique diam non
  ,turpis. Cras placerat accumsan nulla. Nullam rutrum. Nam vestibulum
  ,accumsan nisl.
#+end_src

**** =org-babel-tangle-body-hook= for reprocessing code block bodies during tangling
**** =padline= header argument controls newline padding during tangling
**** Maxima code blocks are now supported

Thanks to Eric Fraga for contributing this support.

**** =awk= code blocks are now supported
**** Added =xmpfilter= to Ruby code blocks for annotated code output
**** New =noweb-ref= header argument

This header argument may be used to concatenate the bodies of
many code blocks into a single noweb reference.  This brings
Org-mode's tangling functionality in line with traditional noweb
tangling.

A no web reference like the following

#+begin_src org
  ,#+begin_src sh
  ,  <<the-ref>>
  ,#+end_src
#+end_src

will now expand to include the bodies of all code blocks which
are named =the-ref=, as well as all code blocks which have a
=:noweb-ref= header argument set to the value =the-ref=.

*** New tests

The =tests/= directory has been extensively updated.

** Important bugfixes

*** Org-exp-blocks --- proper handling of recursively nested blocks

During export pre-processing org-exp-blocks will now ensure that
all matched blocks contain a proper balanced number of
recursively nested blocks.

Before this fix nested blocks such as the following would break
during export.

#+begin_src org
  ,#+begin_src org
  ,  ,#+begin_example
  ,  ,  nested example
  ,  ,#+end_example
  ,#+end_src
#+end_src

*** List handling

Fix an infinite loop when a list has an end of block string
without the corresponding beginning.

Auto-filling cannot happen at a location where it would otherwise
insert a new item.

** Details

*** Footnotes have gone through some bug-fixing:

- properly ignore footnotes in comments,
- export calls to previously defined footnotes in LaTeX using
  \footnotemark,
- export footnotes before first heading (LaTeX),
- export footnotes when selecting a subtree not holding their
  definition (LaTeX).

*** Many small bug fixes have been applied to list handling

- fix `org-timer-item',
- fix insertion of a new item with a non-nil `indent-tabs-mode',
- fix use of `fill-region' in an item,
- correct export lists within footnotes and footnotes within lists,
- correctly export lists containing macros,
- don't ignore with-case specification when sorting a list,
- better indentation handling when changing an item to an headline
  or the other way,
- fix check-boxes' cookies updating.

* Version 7.5

** Incompatible changes

*** Code block variable initialized with Emacs Lisp code in tables and lists

It is no longer possible to assign code block variables using
executable Emacs Lisp statements contained in tables or lists.
As per the following example.
#+tblname: table
| (a b c) |

#+begin_src perl :var data=table[0,0]
  $data
#+end_src

#+results:
: (a b c)

Thanks to Vladimir Alexiev for raising this issue.

*** `org-bbdb-anniversary-format-alist' has changed

Please check the docstring and update your settings accordingly.
** New features and user-visible improvements

*** Implement formulas applying to field ranges

Carsten implemented this field-ranges formulas.

: A frequently requested feature for tables has been to be able to define
: row formulas in a way similar to column formulas.  The patch below allows
: things like
:
: @3=
: @2$2..@5$7=
: @I$2..@II$4=
:
: as the left hand side for table formulas in order to write a formula that
: is valid for an entire column or for a rectangular section in a
: table.

Thanks a lot to Carsten for this.

*** Improved handling of lists

Nicolas Goaziou extended and improved the way Org handles lists.

1. Indentation of text determines again end of items in
   lists. So, some text less indented than the previous item
   doesn't close the whole list anymore, only all items more
   indented than it.

2. Alphabetical bullets are implemented, through the use of the
   variable `org-alphabetical-lists'. This also adds alphabetical
   counters like [@c] or [@W].

3. Lists can now safely contain drawers, inline tasks, or various
   blocks, themselves containing lists. Two variables are
   controlling this: `org-list-forbidden-blocks', and
   `org-list-export-context'.

4. Improve `newline-and-indent' (C-j): used in an item, it will
   keep text from moving at column 0. This allows to split text
   and make paragraphs and still not break the list.

5. Improve `org-toggle-item' (C-c -): used on a region with
   standard text, it will change the region into one item. With a
   prefix argument, it will fallback to the previous behavior and
   make every line in region an item. It permits to easily
   integrate paragraphs inside a list.

6. `fill-paragraph' (M-q) now understands lists. It can freely be
   used inside items, or on text just after a list, even with no
   blank line around, without breaking list structure.

Thanks a lot to Nicolas for all this!

*** Modified link escaping

David Maus worked on `org-link-escape'.  See [[http://article.gmane.org/gmane.emacs.orgmode/37888][his message]]:

: Percent escaping is used in Org mode to escape certain characters
: in links that would either break the parser (e.g. square brackets
: in link target oder description) or are not allowed to appear in
: a particular link type (e.g. non-ascii characters in a http:
: link).
:
: With this change in place Org will apply percent escaping and
: unescaping more consistently especially for non-ascii characters.
: Additionally some of the outstanding bugs or glitches concerning
: percent escaped links are solved.

Thanks a lot to David for this work.

*** Simplification of org-export-html-preamble/postamble

When set to `t', export the preamble/postamble as usual, honoring
the =org-export-email/author/creator-info= variables.

When set to a formatting string, insert this string.  See the
docstring of these variable for details about available
%-sequences.

You can set =:html-preamble= in publishing project in the same
way: `t' means to honor =:email/creator/author-info=, and a
formatting string will insert a string.

*** New command `org-agenda-append-agenda'

You can now use `org-agenda-append-agenda' to dynamically add new
agendas views to the current one.  It is particularily useful to
compare multiple small agendas.

*** Localized clock tables

Clock tables now support a new new =:lang= parameter, allowing
the user to customize the localization of the table headers.  See
the variable =org-clock-clocktable-language-setup= which controls
available translated strings.

*** New sorting options when publishing projects

The =:sitemap-sort-file= option now allows sorting the sitemap
file (anti-)alphabetically and (anti-)chronogically.  Thanks a
lot to Manuel Giraud for a patch to this effet.

*** Testing with ERT

Martyn Jago added new tests to =testing/= - thanks to him!
*** New file in contrib/: org-notmuch.el

Org is now distributed with =org-notmuch.el=, by Matthieu
Lemerre.  See explanations in the header of =org-notmuch.el=:

: =org-notmuch.el= implements links to notmuch messages and
: "searchs". A search is a query to be performed by notmuch; it is
: the equivalent to folders in other mail clients. Similarly, mails
: are refered to by a query, so both a link can refer to several
: mails.

*** org-gnus.el now allows link creation from messages

You can now create links from messages.  This is particularily
useful when the user wants to stored messages that he sends, for
later check.  Thanks to Ulf Stegemann for the patch.

** Important bug fixes

*** Capturing to narrowed buffers

You can now safely capture entries to narrowed buffers.  Thanks a
lot to Memnon Anon for bringing this up.

*** Better handling of the new `org-agenda-span' variable

Agendas were a bit confused by the introduction of this variable,
in particular block agendas.  This is now fixed.

Thanks to Julien and Carsten for helping find the right fix for
this issue, and to Michael Brand and Matt Lundin for their
patient testing and reporting.

*** Security warning: using org-crypt with auto-save

To prevent Emacs from auto-saving encrypted entries in clear
text, the user should not use auto-save with org-crypt.el.  We
now send a warning when users are both using auto-saving and
org-crypt.el.  Thanks to Peter Jones for bringing this up.

** Details

*** Babel
**** :file argument causes results to be written to file for all languages
:file <filename> should be understood as saying "write the result
to <filename> and return a link to <filename>".

This works for all languages. For graphics languages (e.g. ditaa, dot,
gnuplot) there is no change in behavior: "result" in the above is the
graphics, and a link to the image is placed in the org buffer. For
general-purpose languages (e.g. emacs-lisp, python, R, ruby, shell),
the "result" written to file is the normal org-babel result (string,
number, table).

In order to return a file link from a src block without telling babel
to save any results to that file, use :results <filename> and do not
use :file. The code block can of course write arbitrary content to
<filename>.

Some examples:

Save the output of ls -l as a .csv file (recall that :results value is
the default):

#+begin_src sh :file dirlisting.csv :sep ,
 ls -l
#+end_src

Send the text output of ls -l directly to file:

#+begin_src sh :results output :file dirlisting.txt
 ls -l
#+end_src

**** R requires :results graphics :file filename when generating graphics
":results graphics" is now required in addition to ":file
filename" in order for graphical output to be sent automatically
to file. If :file is supplied, but not ":results graphics", then
non-graphical, "value" or "output" results are written to file,
depending on which of those options is in effect.
**** Calc code blocks can now accept vectors
For example;

#+begin_src calc :var y=[1 2 3]
  3 y
#+end_src

#+results:
: [3, 6, 9]

Thanks to Eric S. Fraga for raising this issue

**** Code blocks with empty bodies are now acceptable

Previously these caused errors on export.  Thanks to Martyn Jago
for this patch.

**** Emacs Lisp variable assignments which don't eval cleanly passed literally

This makes it possible to easily pass through non-elisp variable
assignments which may initially look like valid elisp.

**** Unified naming of =c++= functions to =C++=
Thanks to Martyn Jago for this patch.

**** `org-babel-execute-buffer' and `org-babel-execute-subtree' now eval inline code blocks as well

**** New :mkdirp header argument creates parent dirs of tangle targets

**** New ":comments noweb" option for wrapping noweb references in comment links

This can be useful to allow backward linking from tangle code
files to the original code block holding noweb-expanded content.

**** Allow detangling of text containing '\'s -- Thanks to Seth Burleigh

**** =:sep= specifies table separator when opening or writing tabular results

**** `org-edit-src-content-indentation' can now be a buffer-local variable
*** All export configuration variables can now be buffer-local variables
*** org-complete.el has been renamed to org-pcomplete.el

In case you were manually loading =org-complete.el= (which is
*not* necessary anyway), please be aware that the name of this
library was changed to =org-pcomplete.el=.

*** New user options for LaTeX source code export via minted and listings packages

New variables `org-export-latex-listings-options' and
`org-export-latex-minted-options' allow package options to be
controlled; `org-export-latex-custom-lang-environments' allows
arbitrary configuration on a per-language basis.

*** Effort durations now support 2d, 2m, etc.

Effort duration can now be set as 2h (for 2 hours), etc.  This
will be converted to minutes automatically when clocking in an
entry with an effort property.   See the =org-effort-durations=
variable.

Thanks a lot to Lawrence Mitchell for this patch.

*** New option :clock-keep for capture templates

A capture template with =:clock-keep t= will prevent the refiling
process from clocking out the entry.  If =:clock-resume= is also
`t', =:clock-keep= will take precedence and =:clock-resume= will
be ignored.

So now =:immediate-finish t :clock-in t :clock-keep t= makes
sense: it will capture a new task and clock it.

*** Misc

**** New command `org-agenda-bulk-mark-regexp'

=M-x org-agenda-bulk-mark-regexp RET= will mark agenda entries
which headings match against a regular expression.  You can call
this command with the `%' key from an agenda buffer.

**** New command `org-agenda-reset-view'

Julien Danjou implemented this:

: This new command lets you switch to day/week/month/year view.
:
: When switching to day or week view, this setting becomes the default for
: subsequent agenda refreshes.  Since month and year views are slow to
: create, they do not become the default.  A numeric prefix argument may be
: used to jump directly to a specific day of the year, ISO week, month, or
: year, respectively.  For example, `32 d' jumps to February 1st, `9 w' to
: ISO week number 9.  When setting day, week, or month view, a year may be
: encoded in the prefix argument as well.  For example, `200712 w' will jump
: to week 12 in 2007.  If such a year specification has only one or two
: digits, it will be mapped to the interval 1938-2037. `v SPC'' will reset to
: what is set in `org-agenda-span'.

Thanks a lot to Julien for this.

**** New options for ignoring past or future items in the global todo list

This patch gives users greater control over which past or future items
they would like to ignore in the global todo list. By setting
org-agenda-todo-ignore-scheduled to 7, for instance, a user can ignore all
items scheduled 7 or more days in the future. Similarly, by setting
org-agenda-todo-ignore-scheduled to -1, a user can ignore all items that
are truly in the past (unlike the 'past setting, which ignores items
scheduled today).

See the docstrings of these variables:

- org-agenda-todo-ignore-deadlines
- org-agenda-todo-ignore-scheduled
- org-agenda-todo-ignore-timestamp

Thanks a lot to Matt Lundin for implementing this and to Paul
Sexton for the idea.

**** New variable `org-export-table-remove-empty-lines'

When set to `nil', don't remove empty tables when exporting
tables.  This was requested by Eric S Fraga.

**** New variable `org-table-fix-formulas-confirm'

Sometime, editing the structure of a table should not edit the
corresponding formulas.  This new variable lets the user decide
whether he wants to confirm formula fixes or not.

**** New variable `org-export-initial-scope'

This variable controls the initial scope when exporting with `org-export'.
It can be set to 'buffer or 'subtree.  If there is an active region, tell
it when prompting the user for an export command.

**** Show and use the default refile location

M-x org-refile RET now shows the default refile location.  Thanks to
Tassilo Horn for a patch to this effect.

**** New variable `org-archive-subtree-add-inherited-tags'

Non-nil means append inherited tags when archiving a subtree.

**** New variable `org-export-current-backend'

This variable is dynamically set by exporters.  You can check
against its value anytime in your code to see if you are
exporting to HTML, LaTeX, etc.  Possible values are 'html,
'latex, 'ascii, 'docbook.  Thanks to Eric Schulte and Dan Davison
for ideas and patches in this area.

**** New hook `org-clock-before-select-task-hook'

Hook called in task selection just before prompting the user.

Thanks to Benjamin Drieu for the patch.

**** = = emphasis now uses \protectedtexttt
**** Author's email now included in the LaTeX title

When `org-export-email-info' is non-nil, the LaTeX title will
also include the author's email.  Thanks to Lawrence Mitchell for
the patch.


**** Update contrib/scripts/ditaa.jar to ditaa v0.9 of 2009-11-24

**** New variable `org-mobile-files-exclude-regexp'

This variable lets you exclude files that you don't want in
org-mobile-files.

**** New variable `org-confirm-elisp-link-not-regexp'

Set this to a regexp if you want to skip the confirmation step for
Elisp/Shell code matching this regexp.

**** New variable `org-attach-store-link-p'

When set to `t', store link to the attached file, at its original location.

**** `org-table-use-standard-references' now defaults to 'from

**** Better `org-agenda-repeating-timestamp-show-all'

When this is set to a list of TODO keywords, the agenda will only show
occurrences of repeating stamps for these TODO keywords.

**** New command `org-narrow-to-block'

This command (`C-x n b') will narrow the buffer to the current block.
* Version 7.4

** Incompatible changes

*** Agenda: rework ndays and span handling

The variable =org-agenda-ndays= is obsolete - please use
=org-agenda-span= instead.

Thanks to Julien Danjou for this.

** Details

*** Improvements with inline tasks and indentation

There is now a configurable way on how to export inline tasks.  See
the new variable =org-inlinetask-export-templates=.

Thanks to Nicolas Goaziou for coding these changes.

*** Agenda: Added a bulk "scattering" command

=B S= in the agenda buffer will cause tasks to be rescheduled a random
number of days into the future, with 7 as the default.  This is useful
if you've got a ton of tasks scheduled for today, you realize you'll
never deal with them all, and you just want them to be distributed
across the next N days.  When called with a prefix arg, rescheduling
will avoid weekend days.

Thanks to John Wiegley for this.

*** In-buffer completion is now done using John Wiegleys pcomplete.el

Thanks to John Wiegley for much of this code.

*** Sending radio tables from org buffers is now allowed

Org radio tables can no also be sent inside Org buffers.  Also,
there is a new hook which get called after a table has been sent.

Thanks to Seweryn Kokot.

*** Command names shown in manual

The reference manual now lists command names for most commands.
Thanks to Andreas Röhler who started this project.

*** Allow ap/pm times in agenda time grid

Times in the agenda can now be displayed in am/pm format.  See the new
variable =org-agenda-timegrid-use-ampm=.  Thanks to C. A. Webber for
a patch to this effect.

*** Rewriten clock table code

The entire clocktable code has been rewritten to add more options and
to make hacking time reports easier.

Thanks to Erwin Vrolijk for a patch introducing clock tables for
quarters.

*** Babel
**** Add =msosql= engine to sql code blocks
SQL code blocks can now be executed using the =myosql= engine
using the osql command (from MS SQL Server) on Windows systems.

Thanks to Sébastien Vauban for this contribution.

**** Python code blocks now accept a =preamble= header argument
This allows specification of coding declarations and library imports
which must take place in the beginning of a file of executed python
code (note this header argument is used during code block evaluation
unlike the =shebang= header argument which is used during tangling).
For example

#+begin_src org
  ,#+begin_src python :preamble # -*- coding: utf-8 -*- :return s
  ,s = "é"
  ,#+end_src
#+end_src

Thanks to Vincent Beffara for this idea.

**** Code block name is shown during evaluation query
When the user is queried about the evaluation of a named code block
the name of the code block is now displayed.

Thanks to Tom Dye for this suggestion.

**** Clojure code blocks results insertion
The results of Clojure code blocks have been improved in two ways.
1. lazy sequences are now expanded for insertion into the Org-mode
   buffer
2. pretty printing of results is now possible with both "code" and
   "data" pretty print formats

Thanks to Rick Moynihan for suggesting these changes.

**** Python code blocks now accept a =:return= header argument
This alleviates the need to explicitly insert return statements into
the bode of Python code blocks.  This change both
- allows the same python code blocks to be run both in sessions and
  externally
- removes the floating =return= statements which violated python
  syntax

Thanks to Darlan Cavalcante for proposing this feature.

**** =:results wrap= header argument wraps code block results
The new =:results wrap= wraps code blocks results in a custom
environment making it possible to offset their contents during
export.  For example

#+begin_src org
  ,#+begin_src emacs-lisp :results wrap
  ,  "code block results"
  ,#+end_src

  ,#+results:
  ,#+BEGIN_RESULT
  ,: code block results
  ,#+END_RESULT
#+end_src

Thanks to Sébastien Vauban for persistently suggesting this enhancement.

**** Code block error buffer wiped clean between executions
Previously the code block error buffer accumulated errors making it
difficult to distinguish between previous and current errors.  This
buffer is now cleaned before every interactive code block evaluation.

**** Lists now recognized by code blocks
It is now possible for code blocks to both read and write list
contents from and to Org-mode buffers.  For example

#+begin_src org
  ,#+results: a-list
  ,- babel
  ,- and
  ,- org-mode

  ,#+source: a-list
  ,#+begin_src emacs-lisp :var lst=a-list :results list
  ,  (reverse lst)
  ,#+end_src
#+end_src

**** Calc added as a supported code block language
The Emacs Calc package can be used through =calc= code blocks allowing
both regular arithmetic operations as well as stack based
calculation.  For example

#+begin_src org
  ,#+source: calc-stack
  ,#+begin_src calc
  ,  8
  ,  1
  ,  '+
  ,  9
  ,  '*
  ,#+end_src

  ,#+results: calc-stack
  ,: 81

  ,#+source: calc-arithmetic
  ,#+begin_src calc :var in=calc-stack
  ,  in / 9
  ,#+end_src

  ,#+results: calc-arithmetic
  ,: 9
#+end_src

**** "org-babel-detangle" propagates change to source code files into code blocks
`org-babel-detangle' can be used to propagate changes to pure source
code files tangled from embedded code blocks in Org-mode files back to
the original code blocks in the Org-mode file.  This can be used on
collaborative projects to keep embedded code blocks up to date with
edits made in pure source code files.

* Version 7.02

** Incompatible Changes
*** Code block hashes
Due to changes in the code resolving code block header arguments
hashing of code block results should now re-run a code block when
an argument to the code block has changed.  As a result of this
change *all* code blocks with cached results will be re-run after
upgrading to the latest version.

*** Testing update
Anyone using the org-mode test suite will need to update the jump
repository for test navigation by executing the following from
the root of the org-mode repository.
: git submodule update
Failure to update this repository will cause loading of
org-test.el to throw errors.
** Details
*** Org-babel speed commands
All Org-babel commands (behind the C-c C-v key prefix) are now
available as speed commands when the point is on the first line of a
code block.  This uses the existing Org-mode speed key mechanisms.

Thanks to Jambunathan K for implementation this new feature.

*** Fontify code in code blocks.

Source code in code blocks can now be fontified.  Please customize the
varable =org-src-fontify-natively=. For very large blocks (several
hundreds of lines) there can be delays in editing such fontified
blocks, in which case C-c ' should be used to bring up a dedicated
edit buffer.

Thanks to Dan Davison for this.

*** Language-mode commands are available in the Org-buffer
    The most general machinery for doing this is the macro
    `org-babel-do-in-edit-buffer'. There is also the convenience
    function `org-babel-do-key-sequence-in-edit-buffer' which makes
    use of this macro, and is bound to C-c C-v C-x and C-c C-v x. If
    there is an active region contained within the code block, then
    this is inherited by the edit buffer. Some examples of the sorts
    of usage this permits are

C-c C-v C-x M-;       comment region according to language
C-c C-v C-x C-M-\     indent region according to language

Users can make these more convenient, e.g.

(defun my/org-comment-dwim (&optional arg)
    (interactive "P")
    (or (org-babel-do-key-sequence-in-edit-buffer "\M-;")
        (comment-dwim arg)))

(define-key org-mode-map "\M-;" 'my/org-comment-dwim)

A common instance of this general pattern is built in to Org-mode,
controlled by the variable `org-src-tab-acts-natively': if this
variable is set, then TAB in a code block has the effect that it would
have in the language major mode buffer.

*** Org-babel commands are available in language-mode edit buffer
    Mirroring the language-native commands in Org buffers above, a new
    macro `org-src-do-at-code-block' and convenience function
    `org-src-do-key-sequence-at-code-block' provide the converse. When
    used in a language major-mode edit buffer (i.e. a buffer generated
    by C-c '), `org-src-do-key-sequence-at-code-block' executes a key
    sequence at the code block in the source Org buffer.  The command
    bound to the key sequence in the Org-babel key map is executed
    remotely with point temporarily at the start of the code block in
    the Org buffer.

    The command is not bound to a key by default, to avoid conflicts
    with language major mode bindings. To bind it to C-c @ in all
    language major modes, you could use

  (add-hook 'org-src-mode-hook
            (lambda () (define-key org-src-mode-map "\C-c@"
                    'org-src-do-key-sequence-at-code-block)))

    In that case, for example, C-c @ t issued in code edit buffers
    would tangle the current Org code block, C-c @ e would execute
    the block and C-c @ h would display the other available
    Org-babel commands.

*** Multi-line header arguments to code blocks
Code block header arguments can now span multiple lines using the
new =#+header:= or =#+headers:= lines preceding a code block or
nested in between the name and body of a named code block.
Examples are given below.

- multi-line header arguments on an un-named code block
  : #+headers: :var data1=1
  : #+begin_src emacs-lisp :var data2=2
  :   (message "data1:%S, data2:%S" data1 data2)
  : #+end_src
  :
  : #+results:
  : : data1:1, data2:2

- multi-line header arguments on a named code block
  :   #+source: named-block
  :   #+header: :var data=2
  :   #+begin_src emacs-lisp
  :     (message "data:%S" data)
  :   #+end_src
  :
  :   #+results: named-block
  :   : data:2

*** Unified handling of variable expansion for code blocks
The code used to resolve variable references in code block header
arguments has now been consolidated.  This both simplifies the
code base (especially the language-specific files), and ensures
that the arguments to a code block will not be evaluated multiple
times.  This change should not be externally visible to the
Org-mode user.
*** Improved Caching
Code block caches now notice if the value of a variable argument
to the code block has changed, if this is the case the cache is
invalidated and the code block is re-run.  The following example
can provide intuition for the new behavior.
#+begin_src org :exports code
  ,#+srcname: random
  ,#+begin_src R :cache yes
  ,runif(1)
  ,#+end_src

  ,#+results[a2a72cd647ad44515fab62e144796432793d68e1]: random
  ,: 0.4659510825295

  ,#+srcname: caller
  ,#+begin_src emacs-lisp :var x=random :cache yes
  ,x
  ,#+end_src

  ,#+results[bec9c8724e397d5df3b696502df3ed7892fc4f5f]: caller
  ,: 0.254227238707244
#+end_src

*** Added :headers header argument for LaTeX code blocks
This makes it possible to set LaTeX options which must take place in
the document pre-amble for LaTeX code blocks.  This header argument
accepts either a single string or a list, e.g.

#+begin_src org
  ,#+begin_src latex :headers \usepackage{lmodern} :file name1.pdf
  ,  latex body
  ,#+end_src

  ,#+begin_src latex :headers '("\\usepackage{mathpazo}" "\\usepackage{fullpage}") :file name2.pdf
  ,  latex body
  ,#+end_src
#+end_src

*** New function `org-export-string'
Allows exporting directly from a string to the specified export format.
*** Code block header argument ":noweb tangle"
Only expands <<noweb>> syntax references when tangling, not during
export (weaving).
*** New function `org-babel-switch-to-session-with-code'
C-c C-v z (`org-babel-switch-to-session-with-code') is a variant of
C-c C-v C-z (`org-babel-switch-to-session'): instead of switching to
the session buffer, it splits the window between (a) the session
buffer and (b) a language major-mode edit buffer for the code block in
question. This can be convenient for using language major mode for
interacting with the session buffer.

*** Improvements to R sessions
    R now uses standard ESS code evaluation machinery in the :results
    value case, which avoids unnecessary output to the comint
    buffer. In addition, the R command responsible for writing the
    result to file is hidden from the user.  Finally, the R code edit
    buffer generated by C-c ' is automatically linked to the ESS
    session if the current code block is using :session.

*** Temporary file directory
All babel temporary files are now kept in a single sub-directory in
the /tmp directory and are cleaned up when Emacs exits.

*** Function for demarcating blocks `org-babel-demarcate-block'
Can be called to wrap the region in a block, or to split the block
around point, bound to (C-c C-v d).

*** Function for marking code block contents `org-babel-mark-block'
    Bound to C-M-h in the babel key map (i.e. C-c C-v C-M-h by
    default).  This can be useful in conjunction with
    `org-babel-do-in-edit-buffer', for example for language-native
    commenting or indenting of the whole block.
*** Lists of anniversaries are now handeled better

When several anniversaries are defined in the bbdb anniversaries
field (separated by semicolon), this is now handled nicely by the
agenda.

Thanks to Łukasz Stelmach for a patch to this effect.

*** Table fields are now aligned better, new <c> cookie.

In HTML export, table fields are now properly aligned in accord
with automatic alignment in org, or as set by the =<r>=, =<l>=, and
=<c>= cookies.  The =<c>= cookie is new and has no effect in
Org, but it does do the right thing in HTML export.  A LaTeX export
implementation will follow, but is currently still missing.

*** Update freemind converter to include body text

The freemind exporter now incorporates body text into the mind
map.

Thanks to Lennard Borgman for this patch.

*** Make footnotes work correctly in message-mode
The footnotes code now searches for =message-signature-separator=
(which is "-- " by default) in order to place footnotes before the
signature.  Thanks to Tassilo Horn for this patch.

*** Improve XEmacs compatibility

Org-mode 7.02 now runs again in 21.4.22 if the new XEmacs base
package is installed.

Thanks to Uwe Bauer, Volker Ziegler, Michael Sperber and others
for a discussion that lead to this nice result.

*** Make it configurable wether agenda jumping prefers the future

When jumping to a date from the agenda using the =j= key, you may
or may not like the property of Org's date reader to prefer the
future when you enter incomplete dates.   This can now be
configured using the variable =org-agenda-jump-prefer-future'.

*** Add publishing functions for ASCII, Latin-1 and UTF-8

There are now publishing functions =org-publish-org-to-ascii=,
=org-publish-org-to-latin1=, and =org-publish-org-to-utf8=.

Thanks to Matthias Danzl for showing how to do this.

*** Indentation and headline insertion after inline tasks

Indentation in inline tasks, and headline insertion after inline
tasks now behave as expected.

*** Encryption in MobileOrg finally works

As soon as MobilOrg 1.5 hits the Apple's AppStore, you can
encrypt your org files on public servers.  Please see the
documentation of MobileOrg and Appendix B of the manual for more
details.

*** MobileOrg: Do not force to insert IDs

If you dislike the property of MobileOrg to insert ID properties
for in all entries being part of an agenda view, you can now turn
this off using the variable
=org-mobile-force-id-on-agenda-items=.  When this variable is set
to =nil=, MobileOrg will use outline paths to identify entries.
Note that this may fail if several entries have identical outline
paths.

*** LaTeX minted package for fontified source code export
Patch by Dan Davison.

A non-nil value of `org-export-latex-minted' means to export source
code using the minted package, which will fontify source code
with color.  If you want to use this, you need to make LaTeX use the
minted package. Add minted to `org-export-latex-packages-alist', for
example using customize, or with something like

  (require 'org-latex)
  (add-to-list 'org-export-latex-packages-alist '("" "minted"))

In addition, it is neccessary to install
pygments (http://pygments.org), and to configure
`org-latex-to-pdf-process' so that the -shell-escape option is
passed to pdflatex.

*** Allow to use texi2dvi or rubber for processing LaTeX to pdf

Please see the variable =org-export-latex-to-pdf-process= for
more information.

Thanks to Olivier Schwander for the rubber part.

*** New STARTUP keywords to turn on inline images

If you want to inline images whenever you visit an Org file, use

: #+STARTUP: inlineimages

*** Support for user-extensible speed commands.

There is a new hook =org-speed-command-hook=.  Thanks to
Jambunathan for a patch to this effect.

*** Add macro to insert property values into exported text

you can use {{{property{NAME}}}} to insert the value of a
property upon export.

Thanks to David Maus for a patch to this effect.

*** LaTeX package fixes

We updated the list of default packages loaded by LaTeX exported
files.

*** Allow "#" and "%" in tags

Tags can now also contain the characters =#= and =%=, in addition
to =@= and letters.

*** Show command names in manual

Andreas Röhler is adding command names to keys in the manual.
This will take a while to complete, but a start has been made.

*** Make backslash escape "-" in property matches

When entering a tags/property query, "-" is a logical operator.
However, "-" is also allowed in property names.  So you can now
write "SOME\-NAME" to work around this issue.

This was a request by Ilya Shlyakhter.

*** Document quick insertion of empty structural elements

Org-mode has a built-in template mechanism for inserting block
templates.  This was undocumented until now.

Thanks to Jambunathan K for the patch.

*** Implement MathJax support

Org-mode now uses MathJax to display math on web pages.  We serve
MathJax from the orgmode.org server, at least for the time being
(thanks Bastien!).  If you are going to use this for pages which
are viewed often, please install MathJax on your own webserver.

To return to the old way of creating images and inserting them
into web pages, you would have to set

: (setq org-export-with-LaTeX-fragments 'dvipng)

or on a per-file basis

: #+OPTIONS: LaTeX:dvipng

*** Agenda: Allow compact two-column display in agenda dispatcher

If you have many custom agenda commands, you can have the display
in the dispatcher use two columns with the following settings

: (setq org-agenda-menu-show-match nil
:       org-agenda-menu-two-column t)

This was a request by John Wiegley.

*** Add org-wikinodes.el as a contributed package

One frequent request has been to be able to use CamelCase words
for automatic cross links in a Wiki created by Org.  THis is now
possible with org-wikinodes.el, which is available in the contrib
directory.  We also have some [[http://orgmode.org/worg/org-contrib/org-wikinodes.php][documentation]] for this feature up
on Worg.

*** Timer/clock enhancements

=org-timer-set-timer= displays a countdown timer in the modeline.
From the agenda, `J' invokes =org-agenda-clock-goto=.

* Version 7.01

** Incompatible Changes

*** Emacs 21 support has been dropped

Do not use Org mode 7.xx with Emacs 21, use [[http://orgmode.org/org-6.36c.zip][version 6.36c]] instead.

*** XEmacs support requires the XEmacs development version

To use Org mode 7.xx with XEmacs, you need to run the developer
version of XEmacs.   I was about to drop XEmacs support entirely,
but Michael Sperber stepped in and made changes to XEmacs that
made it easier to keep the support.  Thanks to Michael for this
last-minute save.  I had hoped to be able to remove
xemacs/noutline.el from release 7 by moving it into XEmacs, but
this is not yet done.

*** Org-babel configuration changes

Babel took the integration into Org-mode as an opportunity to do
some much needed house cleaning.  Most importantly we have
simplified the enabling of language support, and cleared out
unnecessary configuration variables -- which is great unless you
already have a working configuration under the old model.

The most important changes regard the /location/ and /enabling/
of Babel (both core functionality and language specific support).

- Babel :: Babel is now part of the core of Org-mode, so it is
     now loaded along with the rest of Org-mode.  That means that
     there is /no configuration/ required to enable the main
     Babel functionality.  For current users, this means that
     statements like
     #+begin_src emacs-lisp
       (require 'org-babel)
     #+end_src
     or
     #+begin_src emacs-lisp
       (require 'org-babel-init)
     #+end_src
     that may by lying around in your configuration must now be
     removed.
- load path :: Babel (including all language specific files --
     aside from those which are located in the =contrib/=
     directory for reasons of licencing) now lives in the base of
     the Org-mode lisp directory, so /no additional directories/
     need to be added to your load path to use babel.  For Babel
     users this means that statements adding babel-specific
     directories to your load-path should now be removed from
     your config.
- language support :: It is no longer necessary to require
     language specific support on a language-by-language basis.
     Specific language support should now be managed through the
     `org-babel-load-languages' variable.  This variable can be
     customized using the Emacs customization interface, or
     through the addition of something like the following to your
     configuration (note: any language not mentioned will /not/
     be enabled, aside from =emacs-lisp= which is enabled by
     default)
     #+begin_src emacs-lisp
       (org-babel-do-load-languages
        'org-babel-load-languages
        '((R . t)
          (ditaa . t)
          (dot . t)
          (emacs-lisp . t)
          (gnuplot . t)
          (haskell . nil)
          (ocaml . nil)
          (python . t)
          (ruby . t)
          (screen . nil)
          (sh . t)
          (sql . nil)
          (sqlite . t)))
     #+end_src

	   Despite this change it is still possible to add
	   language support through the use of =require=
	   statements, however to conform to Emacs file-name
	   regulations all Babel language files have changed
	   prefix from =org-babel-*= to =ob-*=, so the require
	   lines must also change e.g.
	   #+begin_src emacs-lisp
       (require 'org-babel-R)
	   #+end_src
	   should be changed to
	   #+begin_src emacs-lisp
       (require 'ob-R)
	   #+end_src

We have eliminated the =org-babel-tangle-w-comments= variable as
well as the two main internal lists of languages, namely
- =org-babel-interpreters= and
- =org-babel-tangle-langs=

so any config lines which mention those variables, can/should be
stripped out in their entirety.  This includes any calls to the
=org-babel-add-interpreter= function, whose sole purpose was to
add languages to the =org-babel-interpreters= variable.

With those calls stripped out, we may still in some cases want to
associate a file name extension with certain languages, for
example we want all of our emacs-lisp files to end in a =.el=, we
can do this will the =org-babel-tangle-lang-exts= variable.  In
general you shouldn't need to touch this as it already has
defaults for most common languages, and if a language is not
present in org-babel-tangle-langs, then babel will just use the
language name, so for example a file of =c= code will have a =.c=
extension by default, shell-scripts (identified with =sh=) will
have a =.sh= extension etc...

The configuration of /shebang/ lines now lives in header
arguments.  So the shebang for a single file can be set at the
code block level, e.g.

#+begin_src org
  ,#+begin_src clojure :shebang #!/usr/bin/env clj
  ,  (println "with a shebang line, I can be run as a script!")
  ,#+end_src
#+end_src

Note that whenever a file is tangled which includes a /shebang/
line, Babel will make the file executable, so there is good
reason to only add /shebangs/ at the source-code block level.
However if you're sure that you want all of your code in some
language (say shell scripts) to tangle out with shebang lines,
then you can customize the default header arguments for that
language, e.g.

#+begin_src emacs-lisp
  ;; ensure this variable is defined defined
  (unless (boundp 'org-babel-default-header-args:sh)
    (setq org-babel-default-header-args:sh '()))

  ;; add a default shebang header argument
  (add-to-list 'org-babel-default-header-args:sh
               '(:shebang . "#!/bin/bash"))
#+end_src

The final important change included in this release is the
addition of new security measures into Babel.  These measures are
in place to protect users from the accidental or uninformed
execution of code.  Along these lines /every/ execution of a code
block will now require an explicit confirmation from the user.
These confirmations can be stifled through customization of the
`org-confirm-babel-evaluate' variable, e.g.
#+begin_src emacs-lisp
  ;; I don't want to be prompted on every code block evaluation
  (setq org-confirm-babel-evaluate nil)
#+end_src

In addition, it is now possible to remove code block evaluation
form the =C-c C-c= keybinding.  This can be done by setting the
=org-babel-no-eval-on-ctrl-c-ctrl-c= variable to a non-nil value,
e.g.
#+begin_src emacs-lisp
  ;; I don't want to execute code blocks with C-c C-c
  (setq org-babel-no-eval-on-ctrl-c-ctrl-c t)
#+end_src

An additional keybinding has been added for code block
evaluation, namely =C-c C-v e=.

Whew! that seems like a lot of effort for a /simplification/ of
configuration.

*** New keys for TODO sparse trees

The key =C-c C-v= is now reserved for Org Babel action.  TODO
sparse trees can still be made with =C-c / t= (all not-done
states) and =C-c / T= (specific states).

*** Customizable variable changes for DocBook exporter

To make it more flexible for users to provide DocBook exporter
related commands, we start to use format-spec to format the
commands in this release.  If you use DocBook exporter and use it
to export Org files to PDF and/or FO format, the settings of the
following two customizable variables need to be changed:

- =org-export-docbook-xslt-proc-command=
- =org-export-docbook-xsl-fo-proc-command=

Instead of using =%s= in the format control string for all
arguments, now we use /three/ different format spec characters:

- =%i=: input file argument
- =%o=: output file argument
- =%s=: XSLT stylesheet argument

For example, if you set =org-export-docbook-xslt-proc-command= to

: java com.icl.saxon.StyleSheet -o %s %s /path/to/docbook.xsl

in the past, now you need to change it to

: java com.icl.saxon.StyleSheet -o %o %i %s

and set a new customizable variable called
=org-export-docbook-xslt-stylesheet= to =/path/to/docbook.xsl=.

Please check the documentation of these two variables for more
details and other examples.

Along with the introduction of variable
=org-export-docbook-xslt-stylesheet=, we also added a new
in-buffer setting called =#+XSLT:=.  You can use this setting to
specify the XSLT stylesheet that you want to use on a per-file
basis.  This setting overrides
=org-export-docbook-xslt-stylesheet=.

** Details

*** Org Babel is now part of the Org core
See [[#ob-configuration-changes][Org-babel configuration changes]] for instructions on how to
update your babel configuration.

The most significant result of this change is that Babel now has
documentation!  It is part of Org-mode's documentation, see
Chapter 14 [[http://orgmode.org/manual/Working-with-source-code.html#Working-with-source-code][Working With Source Code]].  The Babel keybindings
are now listed in the refcard, and can be viewed from any
Org-mode buffer by pressing =C-c C-v h=.  In addition this
integration has included a number of bug fixes, and a significant
amount of internal code cleanup.

*** The default capture system for Org mode is now called org-capture

This replaces the earlier system org-remember.  The manual only
describes org-capture, but for people who prefer to continue to
use org-remember, we keep a static copy of the former manual
section [[http://orgmode.org/org-remember.pdf][chapter about remember]].

The new system has a technically cleaner implementation and more
possibilities for capturing different types of data.  See
[[http://thread.gmane.org/gmane.emacs.orgmode/26441/focus%3D26441][Carsten's announcement]] for more details.

To switch over to the new system:

1. Run

   : M-x org-capture-import-remember-templates RET

   to get a translated version of your remember templates into the
   new variable =org-capture-templates=.  This will "mostly" work,
   but maybe not for all cases.  At least it will give you a good
   place to modify your templates.  After running this command,
   enter the customize buffer for this variable with

   : M-x customize-variable RET org-capture-templates RET

   and convince yourself that everything is OK.  Then save the
   customization.

2. Bind the command =org-capture= to a key, similar to what you did
   with org-remember:

   : (define-key global-map "\C-cc" 'org-capture)

   If your fingers prefer =C-c r=, you can also use this key once
   you have decided to move over completely to the new
   implementation.  During a test time, there is nothing wrong
   with using both system in parallel.

*** Implement pretty display of entities, sub-, and superscripts.

The command =C-c C-x \= toggles the display of Org's special
entities like =\alpha= as pretty unicode characters.  Also, sub
and superscripts are displayed in a pretty way (raised/lower
display, in a smaller font).  If you want to exclude sub- and
superscripts, see the variable
=org-pretty-entities-include-sub-superscripts=.

Thanks to Eric Schulte and Ulf Stegeman for making this possible.

*** Help system for finding entities

The new command =M-x org-entities-help= creates a structured
buffer that lists all entities available in Org.  Thanks to Ulf
Stegeman for adding the necessary structure to the internal
entity list.

*** New module to create Gantt charts

Christian Egli's /org-taskjuggler.el/ module is now part of Org.
He also wrote a [[http://orgmode.org/worg/org-tutorials/org-taskjuggler.php][tutorial]] for it.

*** Refile targets can now be cached

You can turn on caching of refile targets by setting the variable
=org-refile-use-cache=.  This should speed up refiling if you
have many eligible targets in many files.  If you need to update
the cache because Org misses a newly created entry or still
offers a deleted one, press =C-0 C-c C-w=.

*** Enhanced functionality of the clock resolver

Here are the new options for the clock resolver:

: i/q/C-g  Ignore this question; the same as keeping all the idle time.
:
: k/K      Keep X minutes of the idle time (default is all).  If this
:          amount is less than the default, you will be clocked out
:          that many minutes after the time that idling began, and then
:          clocked back in at the present time.
: g/G      Indicate that you \"got back\" X minutes ago.  This is quite
:          different from 'k': it clocks you out from the beginning of
:          the idle period and clock you back in X minutes ago.
: s/S      Subtract the idle time from the current clock.  This is the
:          same as keeping 0 minutes.
: C        Cancel the open timer altogether.  It will be as though you
:          never clocked in.
: j/J      Jump to the current clock, to make manual adjustments.

For all these options, using uppercase makes your final state
to be CLOCKED OUT.  Thanks to John Wiegley for making these
changes.

*** A property value of "nil" now means to unset a property

This can be useful in particular with property inheritance, if
some upper level has the property, and some grandchild of it
would like to have the default settings (i.e. not overruled by a
property) back.

Thanks to Robert Goldman and Bernt Hansen for suggesting this
change.

*** The problem with comment syntax has finally been fixed

Thanks to Leo who has been on a year-long quest to get this fixed
and finally found the right way to do it.

*** Make it possible to protect hidden subtrees from being killed by =C-k=

This was a request by Scott Otterson.
See the new variable =org-ctrl-k-protect-subtree=.

*** New module org-mac-link-grabber.el

This module allows to grab links to all kinds of applications on
a mac.  It is available in the contrib directory.

Thanks to Anthony Lander for this contribution.

*** LaTeX export: Implement table* environment for wide tables

Thanks to Chris Gray for a patch to this effect.

*** When cloning entries, remove or renew ID property

Thanks to David Maus for this change.

* Version 6.36

** Details
*** Inline display of linked images

Images can now be displayed inline.  The key C-c C-x C-v does toggle the
display of such images.  Note that only image links that have no
description part will be inlined.

*** Implement offsets for ordered lists

If you want to start an ordered plain list with a number different from 1,
you can now do it like this:

: 1. [@start:12] will star a lit a number 12

*** Extensions to storing and opening links to Wanderlust messages

- Remove filter conditions for messages in a filter folder

  If customization variable `org-wl-link-remove-filter' is non-nil, filter
  conditions are stripped of the folder name.

- Create web links for messages in a Shimbun folder

  If customization variable `org-wl-shimbun-prefer-web-links' is non-nil,
  calling `org-store-link' on a Shimbun message creates a web link to the
  messages source, indicated in the Xref: header field.

- Create web links for messages in a nntp folder

  If customization variable `org-wl-nntp-prefer-web-links' is non-nil,
  calling `org-store-link' on a nntp message creates a web link either to
  gmane.org if the group can be read trough gmane or to googlegroups
  otherwise. In both cases the message-id is used as reference.

- Open links in namazu search folder

  If `org-wl-open' is called with one prefix, WL opens a namazu search
  folder for message's message-id using `org-wl-namazu-default-index' as
  search index.  If this variable is nil or `org-wl-open' is called with
  two prefixes Org asks for the search index to use.

Thanks to David Maus for these changes.

*** Org-babel: code block body expansion for table and preview

In org-babel, code is "expanded" prior to evaluation. I.e. the code that is
actually evaluated comprises the code block contents, augmented with the
extra code which assigns the referenced data to variables. It is now
possible to preview expanded contents, and also to expand code during
during tangling. This expansion takes into account all header arguments,
and variables.

A new key-binding C-c M-b p bound to `org-babel-expand-src-block' can be
used from inside of a source code block to preview its expanded contents
(which can be very useful for debugging).  tangling

The expanded body can now be tangled, this includes variable values which
may be the results of other source-code blocks, or stored in headline
properties or tables. One possible use for this is to allow those using
org-babel for their emacs initialization to store values (e.g. usernames,
passwords, etc…) in headline properties or in tables.

Org-babel now supports three new header arguments, and new default behavior
for handling horizontal lines in tables (hlines), column names, and
rownames across all languages.

* Version 6.35

** Incompatible Changes

*** Changes to the intended use of =org-export-latex-classes=

So far this variable has been used to specify the complete header of the
LaTeX document, including all the =\usepackage= calls necessary for the
document.  This setup makes it difficult to maintain the list of packages
that Org itself would like to call, for example for the special symbol
support it needs.  Each time I have to add a package, I have to ask people
to revise the configuration of this variable.  In this release, I have
tried to fix this.

First of all, you can *opt out of this change* in the following way: You
can say: /I want to have full control over headers, and I will take
responsibility to include the packages Org needs/.  If that is what you
want, add this to your configuration and skip the rest of this section
(except maybe for the description of the =[EXTRA]= place holder):

#+begin_src emacs-lisp
  (setq org-export-latex-default-packages-alist nil
        org-export-latex-packages-alist nil)
#+end_src

/Continue to read here if you want to go along with the modified
setup./

There are now two variables that should be used to list the LaTeX
packages that need to be included in all classes.  The header
definition in =org-export-latex-classes= should then not contain
the corresponding =\usepackage= calls (see below).

The two new variables are:

1. =org-export-latex-default-packages-alist= :: This is the
     variable where Org-mode itself puts the packages it needs.
     Normally you should not change this variable.  The only
     reason to change it anyway is when one of these packages
     causes a conflict with another package you want to use.
     Then you can remove that packages and hope that you are not
     using Org-mode functionality that needs it.

2. =org-export-latex-packages-alist= :: This is the variable
     where you can put the packages that you'd like to use across
     all classes.  For example, I am putting =amsmath= and =tikz=
     here, because I always want to have them.

The sequence how these customizations will show up in the LaTeX
document are:
1. Header from =org-export-latex-classes=
2. =org-export-latex-default-packages-alist=
3. =org-export-latex-packages-alist=
4. Buffer-specific things set with =#+LaTeX_HEADER:=

If you want more control about which segment is placed where, or
if you want, for a specific class, have full control over the
header and exclude some of the automatic building blocks, you can
put the following macro-like place holders into the header:

#+begin_example
[DEFAULT-PACKAGES]      \usepackage statements for default packages
[NO-DEFAULT-PACKAGES]   do not include any of the default packages
[PACKAGES]              \usepackage statements for packages
[NO-PACKAGES]           do not include the packages
[EXTRA]                 the stuff from #+LaTeX_HEADER
[NO-EXTRA]              do not include #+LaTeX_HEADER stuff
#+end_example

If you have currently customized =org-export-latex-classes=, you
should revise that customization and remove any package calls that
are covered by =org-export-latex-default-packages-alist=.  This
applies to the following packages:

- inputenc
- fontenc
- fixltx2e
- graphicx
- longtable
- float
- wrapfig
- soul
- t1enc
- textcomp
- marvosym
- wasysym
- latexsym
- amssymb
- hyperref

If one of these packages creates a conflict with another package
you are using, you can remove it from
=org-export-latex-default-packages-alist=.  But then you risk
that some of the advertised export features of Org will not work
properly.

You can also consider moving packages that you use in all classes
to =org-export-latex-packages-alist=.  If necessary, put the
place holders so that the packages get loaded in the right
sequence.  As said above, for backward compatibility, if you omit
the place holders, all the variables will dump their content at
the end of the header.

Damn, this has become more complex than I wanted it to be.  I
hope that in practice, this will not be complicated at all.

*** The constant =org-html-entities= is obsolete

Its content is now part of the new constant =org-entities=, which
is defined in the file org-entities.el.  =org-html-entities= was
an internal variable, but it is possible that some users did
write code using it - this is why I am mentioning it here.

** Editing Convenience and Appearance

*** New faces for title, date, author and email address lines.

The keywords in these lines are now dimmed out, and the title is
displayed in a larger font, and a special font is also used for
author, date, and email information.  This is implemented by the
following new faces:

org-document-title
org-document-info
org-document-info-keyword

In addition, the variable =org-hidden-keywords= can be used to
make the corresponding keywords disappear.

Thanks to Dan Davison for this feature.

*** Simpler way to specify faces for tags and todo keywords

The variables =org-todo-keyword-faces=, =org-tag-faces=, and
=org-priority-faces= now accept simple color names as
specifications.  The colors will be used as either foreground or
background color for the corresponding keyword.  See also the
variable =org-faces-easy-properties=, which governs which face
property is affected by this setting.

This is really a great simplification for setting keyword faces.
The change is based on an idea and patch by Ryan Thompson.

*** <N> in tables now means fixed width, not maximum width

Requested by Michael Brand.

*** Better level cycling function

=TAB= in an empty headline cycles the level of that headline
through likely states.  Ryan Thompson implemented an improved
version of this function, which does not depend upon when exactly
this command is used.  Thanks to Ryan for this improvement.

*** Adaptive filling

For paragraph text, =org-adaptive-fill-function= did not handle the
base case of regular text which needed to be filled.  This is now
fixed.  Among other things, it allows email-style ">" comments
to be filled correctly.

Thanks to Dan Hackney for this patch.

*** `org-reveal' (=C-c C-r=) also decrypts encrypted entries (org-crypt.el)

Thanks to Richard Riley for triggering this change.

*** Better automatic letter selection for TODO keywords

When all first letters of keywords have been used, Org now assigns
more meaningful characters based on the keywords.

Thanks to Mikael Fornius for this patch.

** Export

*** Much better handling of entities for LaTeX export

Special entities like =\therefore= and =\alpha= now know if
they need to be in LaTeX math mode and are formatted accordingly.

Thanks to Ulf Stegemann for the tedious work to make this
possible.

*** LaTeX export: Set coding system automatically

The coding system of the LaTeX class will now be set to the value
corresponding to the buffer's file coding system.  This happens
if your setup sets up the file to have a line
=\usepackage[AUTO]{inputenc}= (the default setup does this).

*** New exporters to Latin-1 and UTF-8

While Ulf Stegemann was going through the entities list to
improve the LaTeX export, he had the great idea to provide
representations for many of the entities in Latin-1, and for all
of them in UTF-8.  This means that we can now export files rich
in special symbols to Latin-1 and to UTF-8 files.  These new
exporters can be reached with the commands =C-c C-e n= and =C-c
C-e u=, respectively.

When there is no representation for a given symbol in the
targeted coding system, you can choose to keep the TeX-macro-like
representation, or to get an "explanatory" representation.  For
example, =\simeq= could be represented as "[approx. equal to]".
Please use the variable =org-entities-ascii-explanatory= to state
your preference.

*** Full label/reference support in HTML, Docbook, and LaTeX backends

=#+LABEL= definitions for tables and figures are now fully
implemented in the LaTeX, Docbook, and HTML interfaces.
=\ref{xxx}= is expanded to a valid link in all backends.

*** BEAMER export: Title of the outline frame is now customizable

The new option =org-outline-frame-title= allows to set the
title for outline frames in Beamer presentations.

Patch by Łukasz Stelmach.

*** BEAMER export: fragile frames are better recognized

A =lstlisting= environment now also triggers the fragile option in
a beamer frame, just like =verbatim= environments do.

Thanks to Eric Schulte for this patch.

*** BEAMER export: Protect <...> macro arguments

Macros for the BEAMER package can have arguments in angular
brackets.  These are now protected just like normal arguments.

Requested by Bill Jackson.

*** HTML export: Add class to outline containers using property

The =HTML_CONTAINER_CLASS= property can now be used to add a
class name to the outline container of a node in HTML export.

*** New option =org-export-email-info= to turn off export of the email address

Default is actually off now.

*** Throw an error when creating an image from a LaTeX snippet fails

This behavior can be configured with the new option variable
=org-format-latex-signal-error=.

** Index generation

Org-mode can now produce a 2-level subject index spanning an
entire publishing project.  Write index entries in your files as

#+begin_src org
,* What is org-mode?
#+index: Org-mode
#+index: Definitions!Org-mode
#+end_src

where the first line will produce an index entry /Org-mode/,
while the second line will create /Definitions/ with a sub-item
/Org-mode/.  Three-level entries are not supported.

To produce the index, set

#+begin_src emacs-lisp
:makeindex t
#+end_src

in the project definition in =org-publish-project-alist=.  You
may have to force re-export of all files to get the index by
using a =C-u= prefix to the publishing command:

#+begin_example
C-u M-x org-publish-all
#+end_example

Whenever an Org file is published in this project, a new file
with the extension "orgx" will be written.  It contains the index
entries and corresponding jump target names.  When all project
files are published, Org will produce a new file "theindex.inc"
containing the index as a to-level tree.  This file can be
included into any project file using

#+begin_src org
  ,#+include: "theindex.inc"
#+end_src

Org-mode will also create a file "theindex.org" with this include
statement, and you can build a more complex structure (for
example style definitions, top and home links, etc) around this
statement.  When this file already exists, it will not be
overwritten by Org.

Thanks to Stefan Vollmar for initiating and driving this feature.

*** TODO Still need to do the LaTeX portion

** MobileOrg

*** Encrypting stage files for MobileOrg

Since the use of (often pubic) servers is needed for MobileOrg,
it is now possible to encrypt the files to be staged for
MobileOrg.  Version 1.2 of MobileOrg will be needed for this
feature, and Richard Moreland will show instructions on his
website once that is available.  Basically, on the Org-side this
will require the following settings:

#+begin_src emacs-lisp
  (setq org-mobile-use-encryption t
        org-mobile-encryption-password "My_MobileOrg_Password")
#+end_src

So the password will be visible in your local setup, but since
the encryption is only for the public server, this seems
acceptable.

** Agenda

*** Specify entry types as an option

Custom Agenda commands can now limit the sets of entry types
considered for this command by binding =org-agenda-entry-types=
temporarily in the options section of the command.  This can lead
to significant speedups, because instead of laboriously finding
entries and then rejecting them, a whole search cycle is skipped.
For more information see the new section in
[[http://orgmode.org/worg/org-tutorials/org-custom-agenda-commands.php#sec-5][Matt Lundin's agenda custom command tutorial]].

Thanks to Matt Lundin for this feature.

*** Speed up multiple calls to org-diary by only doing buffer prep once

Also a patch by Matt Lundin.

*** Show and hide deadlines in the agenda

You can now hide all deadline entries in the agenda by pressing
=!=.

Thanks to John Wiegley for this feature.

*** Agenda: Allow to suppress deadline warnings for entries also scheduled

The the docstring of the variable
=org-agenda-skip-deadline-prewarning-if-scheduled=.

*** Expand file names in org-agenda-files (external file case)

If you are using a file to manage the list of agenda files, the
names in this file can now contain environment variables and "~"
to write them more compactly and portable.

Thanks to Mikael Fornius for a patch to this effect.

*** Agenda: Allow TODO conditions in the skip functions

The agenda skip function has now special support for skipping
based on the TODO state.  Here are just two examples, see the
manual for more information.

#+begin_src emacs-lisp
(org-agenda-skip-entry-if 'todo '(\"TODO\" \"WAITING\"))
(org-agenda-skip-entry-if 'nottodo 'done)
#+end_src

Thanks to Łukasz Stelmach for this patch.

*** Extracting the time-of-day when adding diary entries

The time of day can now be extracted from new diary entries made
from the agenda with (for example) =i d=.  When
=org-agenda-insert-diary-extract-time= is set, this is done, and
the time is moved into the time stamp.

Thanks to Stephen Eglen for this feature.

*** The customization group org-font-lock has been renamed

The new name is `org-appearance'.

Thanks to Dan Davison for a patch to this effect.

*** The TODO list: Allow skipping scheduled or deadlined entries

Skipping TODO entries in the global TODO list based on whether
they are scheduled or have a deadline can now be controlled in
more detail.  Please see the docstrings of
=org-agenda-todo-ignore-scheduled= and
=org-agenda-todo-ignore-deadline=.

Thanks to Łukasz Stelmach for patches to this effect.

** Hyperlinks

*** Make =org-store-link= point to directory in a dired buffer

When, in a dired buffer, the cursor is not in a line listing a
file, `org-store-link' will store a link to the directory.

Patch by Stephen Eglen.

*** Allow regexps in =org-file-apps= to capture link parameters

The way extension regexps in =org-file-apps= are handled has
changed.  Instead of matching against the file name, the regexps
are now matched against the whole link, and you can use grouping
to extract link parameters which you can then use in a command
string to be executed.

For example, to allow linking to PDF files using the syntax
=file:/doc.pdf::<page number>=, you can add the following entry to
org-file-apps:

#+begin_example
Extension: \.pdf::\([0-9]+\)\'
Command:   evince "%s" -p %1
#+end_example

Thanks to Jan Böcker for a patch to this effect.

** Clocking

*** Show clock overruns in mode line

When clocking an item with a planned effort, overrunning the
planned time is now made visible in the mode line, for example
using the new face =org-mode-line-clock-overrun=, or by adding an
extra string given by =org-task-overrun-text=.

Thanks to Richard Riley for a patch to this effect.

** Tables

*** Repair the broken support for table.el tables again.

Tables created with the table.el package now finally work again
in Org-mode.  While you cannot edit the table directly in the
buffer, you can use  =C-c '= to edit it nicely in a temporary
buffer.

Export of these tables to HTML seem to work without problems.
Export to LaTeX is imperfect.  If fails if the table contains
special characters that will be replaced by the exporter before
formatting the table.  The replacement operation changes the
length of some lines, breaking the alignment of the table fields.
Unfortunately this is not easy to fix.  It is also not an option
to not do these replacements.  The table.el LaTeX exporter will
for example not escape "&" in table fields, causing the exported
tables to be broken.

** Misc

*** New logging support for refiling

Whenever you refile an item, a time stamp and even a note can be
added to this entry.  For details, see the new option
=org-log-refile=.

Thanks to Charles Cave for this idea.

*** New helper functions in org-table.el

There are new functions to access and write to a specific table
field.  This is for hackers, and maybe for the org-babel people.

#+begin_example
org-table-get
org-table-put
org-table-current-line
org-table-goto-line
#+end_example

*** Tables: Field coordinates for formulas, and improved docs

Calc and Emacs-Lisp formulas for tables can access the current
field coordinates with =@#= and =$#= for row and column,
respectively.  These can be useful in some formulas.  For
example, to sequentially number the fields in a column, use
~=@#~ as column equation.

One application is to copy a column from a different table.  See
the manual for details.

Thanks to Michael Brand for this feature.

*** Archiving: Allow to reverse order in target node

The new option =org-archive-reversed-order= allows to have
archived entries inserted in a last-on-top fashion in the target
node.

Requested by Tom.

*** Better documentation on calc accuracy in tables

Thanks to Michael Brand for this fix.

*** Clock reports can now include the running, incomplete clock

If you have a clock running, and the entry being clocked falls
into the scope when creating a clock table, the time so far spent
can be added to the total.  This behavior depends on the setting
of =org-clock-report-include-clocking-task=.  The default is
=nil=.

Thanks to Bernt Hansen for this useful addition.

*** American-style dates are now understood by =org-read-date=

So when you are prompted for a date, you can now answer like this

#+begin_example
2/5/3         --> 2003-02-05
2/5           --> <CURRENT-YEAR>-02-05
#+end_example

*** org-timer.el now allows just one timer

There is now only a single free timer supported by org-timer.el.
Thanks to Bastien for cleaning this up, after a bug report in
this area by Frédéric Couchet.

*** Remember: Allow to file as sibling of current clock

=C-3 C-c C-c= will file the remember entry as a sibling of the
last filed entry.

Patch by Łukasz Stelmach.

*** Org-reveal: Double prefix arg shows the entire subtree of the parent

This can help to get out of an inconsistent state produced for
example by viewing from the agenda.

This was a request by Matt Lundin.

*** Add org-secretary.el by Juan Reyero to the contrib directory

org-secretary.el is a possible setup for group work using
Org-mode.

Thanks to Juan Reyero for this contribution.

** Babel

Eric and Dan have compiled the following list of changes in and
around org-babel.

- Added support for Matlab and Octave.
- Added support for C and C++ code blocks.
- Added support for the Oz programming language.
  Thanks to Torsten Anders for this contribution
- Can now force literal interpretation of table cell contents
  with extra "$" in table formula.
  Thanks to Maurizio Vitale for this suggestion.
- Variable references which look like lisp forms are now
  evaluated.
- No longer adding extension during tangling when filename is
  provided.
  Thanks to Martin G. Skjæveland and Nicolas Girard for prompting this.
- Added `org-babel-execute-hook' which runs after code block
  execution.
- Working directories and remote execution

  This introduces a new header argument :dir. For the duration of
  source block execution, default-directory is set to the value
  of this header argument. Consequences include:

  - external interpreter processes run in that directory
  - new session processes run in that directory (but existing
    ones are unaffected)
  - relative paths for file output are relative to that directory

  The name of a directory on a remote machine may be specified
  with tramp syntax (/user@host:path), in which case the
  interpreter executable will be sought in tramp-remote-path, and
  if found will execute on the remote machine in the specified
  remote directory.
- Tramp syntax can be used to tangle to remote files.
  Thanks to Maurizio Vitale and Rémi Vanicat.
- org-R removed from contrib.
- gnuplot can now return it's string output -- when session is
  set to "none".
- Now including source code block arguments w/source name on
  export.
- Now able to reference file links as results.
- Allow pdf/png generation directly from latex source blocks
  with :file header argument.

* Version 6.34

** Incompatible changes

*** Tags in org-agenda-auto-exclude-function must be lower case.

When defining an =org-agenda-auto-exclude-function=, you need to
be aware that tag that is being passed into the function is
always lower case - even if it was defined in upper case
originally.

** Details

*** Support for creating BEAMER presentations from Org-mode documents

Org-mode documents or subtrees can now be converted directly in
to BEAMER presentation.  Turning a tree into a simple
presentations is straight forward, and there is also quite some
support to make richer presentations as well.  See the [[http://orgmode.org/manual/Beamer-class-export.html#Beamer-class-export][BEAMER
section]] in the manual for more details.

Thanks to everyone who has contributed to the discussion about
BEAMER support and how it should work.  This was a great example
for how this community can achieve a much better result than any
individual could.

*** Hyperlinks

**** Add Paul Sexton's org-ctags.el

Targets like =<<my target>>= can now be found by Emacs' etag
functionality, and Org-mode links can be used to to link to
etags, also in non-Org-mode files.  For details, see the file
/org-ctags.el/.

This feature uses a new hook =org-open-link-functions= which will
call function to do something special with text links.

Thanks to Paul Sexton for this contribution.

**** Add Jan Böcker's org-docview.el

This new module allows links to various file types using docview,
where Emacs displays images of document pages.  Docview link
types can point to a specific page in a document, for example to
page 131 of the Org-mode manual:

: [[docview:~/.elisp/org/doc/org.pdf::131][Org-Mode Manual]]

Thanks to Jan Böcker for this contribution.

**** New link types that force special ways of opening the file

- =file+sys:/path/to/file=  will use the system to open the file,
  like double-clicking would.
- file+emacs:/path/to/file will force opening the linked file
  with Emacs.

This was a request by John Wiegley.

**** Open all links in a node

When using =C-c C-o= on a headline to get a list of links in the
entry, pressing =RET= will open *all* links.  This allows
something like projects to be defined, with a number of files
that have to be opened by different applications.

This was a request by John Wiegley.

*** Agenda Views

**** Improve the logic of the search view.

The logic of search views is changed a bit.  See the docstring of
the function =or-search-view=.

These changes resulted from a discussion with Matt Lundin.

**** New face for entries from the Emacs diary

Entries that enter the Agenda through the Emacs diary now get the
face =org-agenda-diary=.

This was a request by Thierry Volpiatto.

**** New function `org-diary-class' to schedule classes with skipped weeks.

This was a request by Daniel Martins.

**** Empty matcher means prompt in agenda custom commands

When an agenda custom command has an empty string as MATCH
element, so far this would lead to a meaningless search using an
empty matcher.  Now an empty (or white) string will be
interpreted just like a nil matcher, i.e. the user will be
prompted for the match.

**** Agenda: Selectively remove some tags from agenda display

If you use tags very extensively, you might want to exclude some
from being displayed in the agenda, in order to keep the display
compact.  See the new option =org-agenda-hide-tags-regexp= for
details.

This was largely a patch by Martin Pohlack.

*** Export

**** Direct export of only the current subtree

Pressing =1= after =C-c C-e= and before the key that selects the
export backend, only the current subtree will be exported,
exactly as it you had selected it first with =C-c @=.  So for
example, =C-c C-e 1 b= will export the current subtree to HTML
and open the result in the browser.

**** Direct export of enclosing node

Pressing =SPC= after =C-c C-e= and before the key that selects
the export backend, the enclosing subree that is set up for
subtree export will be exported, exactly as it you had selected
it first with =C-c @=.  So for example, =C-c C-e SPC d= will find
the enclosing node with a LaTeX_CLASS property or an
EXPORT_FILE_NAME property and export that.

**** Caching export images

Images that are created for example using LaTeX or ditaa for
inclusion into exported files are now cached.  This works by
adding a hash to the image name, that reflects the source code
and all relevant settings.  So as long as the hash does not
change, the image does not have to be made again.  His can lead
to a substantial reduction in export/publishing times.

Thanks to Eric Schulte for a patch to this effect.

**** Preserving line breaks for export no longer works

ASCII export always preserves them - no other export format
does.  We had attempted to use =\obeylines= for this in LaTeX,
but that does create too many problems.

**** New symbols =\EUR= and =\checkmark=

=\EUR= symbols from Marvosym package, and =\checkmark= are now
supported symbols in Org-mode, i.e. they will be exported
properly to the various backends.

**** Allow LaTeX_CLASS_OPTIONS to set options, also from a property

You can set the options to the =\documentclass= command on a
per-file basis, using

: #+LaTeX_CLASS_OPTIONS: [11pt]

or on a per-tree basis using the corresponding property.  The
defined string will replace the default options entirely.

**** The encoding of LaTeX files is now handled property

Org now makes sure that the encoding used by the file created
through the export mechanism is reflected correctly in the

: \usepackage[CODINGSYSTEM]{inputenc}

command.  So as long as the =org-export-latex-classes= definition
contains an =\usepackage[utf8]{inputenc}= statement, that
statement will be modified so that the correct option is used.

If you wan to use special encodings, for example =utf8x= instead
of =utf8=, see the variable =org-export-latex-inputenc-alist=.

This was a request by Francesco Pizzolante.

*** Property API enhancements

**** Make a new special property BLOCKED, indicating if entry is blocked

A new special property BLOCKED returns "t" when the entry is
blocked from switching the TODO state to a DONE state.

This was a request by John Wiegley.

**** New hooks for external support for allowed property values

It is now possible to hook into Org in order to provide the
allowed values for any property with a lisp function.  See the
docstring of the variable =org-property-allowed-value-functions=

**** Allow unrestricted completion on properties

When listing the allowed values for a property, for example with
a =:name_ALL:= property, completion on these values enforces that
one of the values will be chosen.  Now, if you add ":ETC" to the
list of allowed values, it will be interpreted as a switch, and
the completion will be non-restrictive, so you can also choose to
type a new value.

*** Changes to Org-babel

- The documentation for Org-babel has been drastically improved
  and is available on Worg at
  http://orgmode.org/worg/org-contrib/babel/
- Source-code block names are now exported to HTML and LaTeX
- Org-babel functions are now bound to keys behind a common key
  prefix (see
  http://orgmode.org/worg/org-contrib/babel/reference.php#sec-5)
- Results are now foldable with TAB
- Header argument values can now be lisp forms
- Readable aliases for #+srcname: and #+resname:
- Sha1 hash based caching of results in buffer
- Can now index into variable values
- org-babel-clojure now supports multiple named sessions

*** Miscellaneous changes

**** Make =C-c r C= customize remember templates

=C-c r C= is now a shortcut for

:  M-x customize-variable RET org-remember-templates RET

This was a proposal by Adam Spiers.

**** Use John Gruber's regular expression for URL's

We now use a better regexp to spot plain links in text.  This
regexp is adopted from [[http://daringfireball.net/2009/11/liberal_regex_for_matching_urls][John Gruber's blogpost]].

Thanks to William Henney for the pointer.

**** Implement tag completion of all tags in all agenda files

The new option =org-complete-tags-always-offer-all-agenda-tags=
makes Org complete all tags from all agenda files if non-nil.
Usually, setting it locally to t in org-remember buffers is the
most useful application of this new feature.

Thanks to Tassilo Horn for a patch to this effect.