Emacs-w3m NEWS -- history of user-visible changes.
Copyright (C) 2007 TSUCHIYA Masatoshi
See the end of the file for license conditions.
Please send bug reports to email@example.com.
If possible, use M-x report-emacs-w3m-bug.
This file is about changes in emacs-w3m version 1.5.
* Required Emacs version changes
** Emacs-w3m no longer works with old Emacs versions.
Now emacs-w3m will hopefully work with Emacs 21.1 or greater, XEmacs
21.4.17 or greater, or XEmacs 21.5-b19 or greater.
*** Version-specific modules are now w3m-ems.el and w3m-xmas.el.
The modules w3m-e21.el, w3m-e23.el, and w3m-fsf.el have been
integrated into w3m-ems.el.
*** We can now use shy group in regular expressions.
* Appearance changes
** Emacs-w3m can now display italic text.
While web pages often use italic text to emphasize words or sentences,
emacs-w3m (and w3m) displayed them as normal text formerly. To enable
this feature, you need to have w3m 0.5.2 or greater installed. The
face used to display italic text is `w3m-italic'.
** Emphasizing of text to display has been much improved.
The face attributes including bold, italic, underline, and strike-thru
can be overlapped in the same text now.
** Emacs-w3m can now display the uri and the title in the hreader-line.
This is controlled by the `w3m-use-header-line-title' variable.
** Emacs-w3m can now use the title of the page as a buffer name.
This is controlled by the `w3m-use-title-buffer-name' variable.
** The PNG icon files in addition to the XPM icons are now available.
Those are the ones used in the tool bar, are the best suited to Emacs
built with GTK. So, emacs-w3m uses them for GTK Emacs by default.
Which type of icons to use is controlled by the variable
* Changes in tabs, sessions, and menu
** A handy menu now pops up by clicking mouse-3 on links.
It doesn't work on tty of course, though.
** Tabs now have useful menu succeeding to Firefox+TabMixPlus' way.
You can pop the menu up by a Mouse-3 click on a tab, and may also see
the key bindings of those menu items. The menu is also available in
the `Tab' menu in the menu bar.
** You can now use the mouse wheel to manipulate tabs.
Rolling the mouse wheel on the tabs line circulates the tabs. In
addition to this, pressing the control key while rolling the mouse
wheel changes the order of the tabs.
** Emacs-w3m can now create a new session in the background.
If this is enabled, the page you instruct emacs-w3m to visit as a new
session will not appear unless you select the buffer. This is disabled
by default. You can set the `w3m-new-session-in-background' variable
to a non-nil value to enable it.
** The w3m menu appears about the middle of the menu bar by default.
If you'd like to place it in the leftmost of the menu bar as before,
set the `w3m-menu-on-forefront' variable to a non-nil value.
* Changes in contents decoding
** Emacs-w3m now uses a superset of the charset the page specifies.
For instance, some European web sites sometimes use the `windows-1252'
charset for encoding contents even if the page is labeled with
`iso-8859-1'. In such a case, emacs-w3m uses `windows-1252', which is
a superset of `iso-8859-1', for decoding contents if it is available.
The variable `w3m-compatible-encoding-alist' holds the table of
charsets and their supersets.
** Emacs-w3m can now follow links containing non-ASCII characters.
Emacs-w3m encodes the urls of links whenever retrieving the contents
using the charset by which the page containing the links has been
encoded. (In shimbun, the charset can be overridden by
the `shimbun-url-coding-system' class variable; see below.)
** You can now instruct emacs-w3m how much it decodes html sources.
Give a numeric prefix to the `w3m-view-source' command, which is bound
to the `\' key. See the doc string of the command for details.
** Emacs-w3m now decodes application/xml pages as text/html.
In the case where the page source looks like xhtml+xml.
** Now emacs-w3m trusts the image type gotten from parsing image data.
Because the type that web sites specify as the Content-Type header is
sometimes bogus. Currently, gif, jpeg, and png types can be identified.
* New Lisp modules
** w3m-fb.el -- frame-local buffers support.
You can now have several emacs-w3m frames for various purposes. For
example, one is for news sites, the other is for search engines. See
(info "(emacs-w3m)Frame Local Buffers").
** w3m-mail.el -- interface to mail-user-agent for sending web pages.
You can now mail a web page that you are viewing. The command is
`M-x w3m-mail'. Note that you have to have the `mail-user-agent' set
** w3m-session.el -- functions to load/save sessions.
You can now save in the file the current session in which you visit
several pages. Moreover the last session you use before quitting
emacs-w3m will be saved automatically if `w3m-session-autosave' is
non-nil (the default). Saved sessions can be loaded afterward.
The useful commands are `M-S' for saving and `M-s' for selecting saved
ones. Those are available also in the `Session' submenu in the `w3m'
* Search engines changes
** Many search engines have been added.
See (info "(emacs-w3m)Using Search Engines").
* Changes in filters
** Google searches are prevented from getting corrupted and tidied up.
** Displaying shortened uris for Amazon.
** Support direct access to mixi's diaries being displayed externally.
** Support accessing to the online dictionary Eijiro.
** vm-w3m.el has been transferred to the new VM team.
It is now in the VM 8.x package. See http://www.nongnu.org/viewmail/.
** Emacs-w3m now allows arbitrary string as a uri.
This feature succeeds to Google's `I'm Feeling Lucky', i.e., the string
is searched for using Google. The `w3m-enable-google-feeling-lucky'
variable controls this feature (enabled by default).
** `w3m-goto-article-function' is now a user option.
** Emacs-w3m now visits the thread page in Gmane semi-automatically.
Gmane (http://www.gmane.org/) archives articles distributed in a lot
of the mailing lists and offers them through nntp and http. If you
execute the `w3m' command on the References header or the Message-ID
header of an article, emacs-w3m tries to visit the thread page to
which the article belongs.
** Run XEmacs 21.5 safely.
We gave up asynchronous operation when XEmacs 21.5 accesses many pages
at a time, in order to prevent it from freezing.
** Prevent the byte compiler from issuing useless warnings.
We will continue to do this thoroughly. It makes it easy to find real
errors when compiling emacs-w3m.
** The installation directory name is allowed to have spaces and/or tabs.
** We've done many other improvements and bug fixes in this release.
(Some of them might cause new bugs, though.)
* Shimbun changes
** New shimbun class variable `shimbun-prefer-text-plain'.
It controls whether a shimbun module generates text/plain articles or
html articles. While the sb-asahi.el module (for example) generates
text/plain articles by default, you can tell it to generate html
articles by setting the `shimbun-asahi-prefer-text-plain' variable to
nil. Oppositely, setting `shimbun-mainichi-prefer-text-plain' to non-
nil leads the sb-mainichi.el module to generate text/plain articles
while it generates html articles by default. With this feature,
sb-asahi-html.el and sb-yomiuri-html.el have been made semi-obsolete.
Note that all the shimbun modules don't allow for it.
** Shimbun now supports redirection of urls when fetching articles.
Urls that some web sites offer in the index pages are not the ones that
point to the article pages directly. The real url for the article is
in the page to which such a url points, and it is also likely to
require the client to wait for a while (often for displaying ads).
Shimbun now examines it and fetches the real article contents with no
** Shimbun can now convert wide non-ASCII characters into Hankaku.
English words and numbers displayed with wide (a.k.a. Zenkaku) non-
ASCII characters are illegible. Although it is not available in all
the shimbun modules, it will be worth trying setting
the `shimbun-japanese-hankaku' variable to non-nil. It not only
converts wide non-ASCII characters into the normal ones but also
performs Wakachi-Gaki (which means separating alphanumeric words and
Japanese sentences with space characters). If you want to do it per
shimbun server, use `shimbun-SERVER-japanese-hankaku' instead.
** New shimbun class `shimbun-newspaper'.
This is used for adding a caution on the copyright to non-Japanese
shimbun articles. For Japanese, use `shimbun-japanese-newspaper'.
** New shimbun class variable `shimbun-url-coding-system'.
This overrides the charset used for encoding the urls of links which
contain non-ASCII letters (by default, the urls of links will be
encoded by the charset by which the page containing the links has been
encoded). Use the `shimbun-SERVER-url-coding-system' variable per
** New shimbun class variable `shimbun-retry-fetching'.
When fetching contents of a url fails, shimbun retries it up to that
number of times if it is a positive number. Use the variable
`shimbun-SERVER-retry-fetching' per server.
** New shimbun modules have been added.
sb-aljazeera.el sb-debugmagazin-de.el sb-gendai-net.el sb-nytimes.el
sb-ohmynews-jp.el sb-perlentaucher-de.el sb-slashdot.el
** Some shimbun modules are being well-maintained but others aren't.
*** New variable `nnshimbun-default-group-level' for Gnus.
The value of this variable determines the initial group level of
a shimbun group that is newly created. The default value is nil,
which means that of `gnus-level-default-subscribed' is used.
*** New command `gnus-group-make-shimbun-groups' for Gnus.
This makes all the shimbun groups a server provides.
Emacs-w3m is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.
Emacs-w3m is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with emacs-w3m; see the file COPYING. If not, write to the
Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
Boston, MA 02110-1301, USA.
paragraph-separate: "[ ]*$"