Changes in version 0.8
**********************

* Comment types.

Octave mode can now distinguish between various types of comments,
similar to Emacs Lisp:

`#'	To be aligned to the same column on the right of the source
	code.

`##'	To be aligned to the same level of indentation as the code.

`###'	To start at the left margin.

(Without this distinction, things like

	code			# This is an inline comment which extends 
				# across on line

were not uniquely identifiable, because the second line of the in-line
comment could not be distinguished from a full-line comment.)

* Inferior Octave support.

It is now possible to run Octave from within Emacs, either by directly
entering commands at the prompt in a buffer in Inferior Octave mode, or
by interacting with Octave from within a file with Octave code.  This is
useful in particular for debugging Octave code.

For more details, see the Texinfo documentation.

* New function `octave-insert-defun' to reduce typing effort.


Changes in version 0.7
**********************

This version is a rather complete rewrite of 0.6.

In general, it was attempted to use standard Emacs features and
functions as much as possible.

* Font Lock support was improved and enhanced.

Function declarations are now fontified correctly.  Builtin keywords
(reserved words and text functions) and variables are now fontified as
well.

* Partial Imenu support was added.

Function declarations can now be indexed.

* A completion mechanism for builtin keywords and variables was added.

Completion can be performed by pressing M-TAB (octave-complete-symbol)
after typing the initial characters of the keyword.

In a future release, completion (as well as imenu and font locking) may
also include user defined variables.

* Bug reporting.

The function octave-submit-bug-report was added.

* Commands for dealing with blocks.

The functions for dealing with begin-else-end blocks were rewritten from
scratch.  They are now based on the function octave-scan-blocks, which
works similar to the standard Emacs scan-lists function.  One can now go
forward and backward across balanced blocks, and go up and down block
levels.  It is no longer required that extended `end' keywords are used;
in fact, an error message will be issued in the case of nonmatching
block keywords.  Matching is achieved through one general function, and
now also works for else keywords.

The block motion commands are

	octave-forward-block
	octave-backward-block
	octave-down-block
	octave-up-block
	octave-backward-up-block

One can also mark the `current' block (the innermost block containing
point) using octave-mark-block.

* Commands for dealing with functions.

Moving across functions, as well as marking and indenting them now works
as for defuns in Lisp mode.

* Filling.

The code for auto-filling was rewritten, and octave-fill-paragraph was
added.  The code for filling is far from perfection yet.  In future
releases, a function for filling the region will be added.

The problem of getting the right fill function after toggling Auto-Fill
mode was solved by an advice to auto-fill-mode.

* Abbrevs.

The mechanism of listing all abbrevs using `? or ` was retained.  All
other abbrev code was dropped---why should Octave mode only use its own
abbrevs?

* Comments.

As the comment syntax and comment-start are specified correctly, one can
use the standard comment-region.  The function octave-uncomment-region
is still provided, but now based on comment-region.

The special treatment of commenting regions was removed, but may easily
be added again.

* Paragraphs.

Empty lines and form feeds are now recognized as separating paragraphs
of Octave code, so one can now move across them and fill them.

* Indentation.

Indentation should work without problems now, assuming `sane' coding.
The problems in 0.6 which came from assuming that only one else or end
keyword would occur in one line should now have disappeared.

The user-level variables for customizing indentation were renamed as
follows:

	0.7				0.6

	octave-block-offset		octave-stmt-indent
	octave-continuation-offset	octave-continuation-indent

Of course, it is still easy to write code in a way that makes Octave
mode get the indentations wrong.  For example,

	if (something) \

will result in having octave-continuation-offset added although this is
wrong.  Or,

	printf ("This is some stupid %s \
	for getting indentation wrong.",
		"text")

is what you deserve anyway :-)  Octave mode currently assumes that
strings do not extend across lines ...

* Other commands for moving.

The functions octave-previous-statement and octave-next-statement were
removed, partially because their names do not coincide with the usage of
`statement' in the Octave manual.  The functions now provided are

	octave-previous-code-line
	octave-next-code-line
	octave-beginning-of-line
	octave-end-of-line

The first two look for the previous or next `code' line, i.e., they skip
across all empty or comment lines.  The latter two also go understand
continuation lines, and move to their beginning and end, respectively.

The effect of the former octave-previous-statement can now be achieved
upon following octave-previous-code-line by octave-beginning-of-line.

* Special insertions:  LFD, SPC and semicolon.

These characters are now `electric', doing a little extra work.  All
three expand abbrevs if abbrev mode is on and blink matching blocks if
octave-blink-matching-blocks is t.

In future versions, SPC might also do auto-newlining after certain
keywords (or e.g., a continuation character).

* User-level customization.

The variables for customizing Octave mode are as follows.

** As in 0.6.

	octave-auto-newline
	octave-comment-column
	octave-comment-start
	octave-continuation-string

** Different from 0.6.

	0.7				0.6

	octave-blink-matching-block	octave-blink-matching-blocks
	octave-block-offset		octave-statement-indent
	octave-continuation-offset	octave-continuation-indent
	octave-inhibit-startup-message	octave-startup-message

The first three have only been renamed.  (The first in analogy to the
standard blink-matching-paren, the others because they are really extra
offsets and not the absolute indentations.)

Controlling startup messages now works as in Emacs itself.

** New in 0.7.

	octave-fill-column

** Removed from 0.7.

	octave-comment-indent-style
	octave-comment-indent-char
	octave-comment-region

* Help

A mechanism for looking up entries in the indices of ALL info files with
documentation for Octave (as specified by octave-help-files) was added
(octave-help).  If multiple matches are found, one can cycle through the
matches.

* Other changes.

octave-comment-hook was removed.  One can instead use the standard Emacs
indent-for-comment.