http://blog.gmane.org/gmane.emacs.w3m hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://comments.gmane.org/gmane.emacs.w3m/8793 <pre>$ set http://gmane.org/lists.php $ lynx -dump $&lt; at &gt; | wc 65996 188608 3338251 $ w3m -dump $&lt; at &gt; | wc 26 121 1617 Why does that occur? w3m 0.5.2-10 </pre> jidanni< at >jidanni.org 2010-12-12T01:37:22 http://comments.gmane.org/gmane.emacs.w3m/8792 <pre>On http://blog.dreamhost.com/2010/11/04/mod_pagespeed-now-available/ I put the cursor upon the link embed(http://www.youtube.com/v/8moGR2qf994?fs=1&amp;hl=en_US) and hit M runs the command w3m-view-url-with-external-browser but all does not work smoothly... I end up having to do it by hand. w3m-el-snapshot/unstable uptodate 1.4.400+0.20100725-1 </pre> jidanni< at >jidanni.org 2010-12-10T01:35:43 http://comments.gmane.org/gmane.emacs.w3m/8789 <pre>M runs the command w3m-view-url-with-external-browser, which is... which is great, except that it doesn't give up a chance to edit the URL. We are just asked y or n. How about instead just allow us to edit the URL, and if we feel the "n" feeling, we can just hit ^G. And if on top of a URL, well, should it maybe still ask us to edit the URL?... Or maybe with a C-u argument... </pre> jidanni< at >jidanni.org 2010-12-06T01:23:41 http://comments.gmane.org/gmane.emacs.w3m/8788 <pre>http://www.checkfilename.com/view-details/Google-Earth/RespageIndex/4/sTab/2/ gives error in process sentinel: w3m-filter-unsupport-tags: Stack overflow in regexp matcher in http://jidanni.org/comp/configuration/.emacs-w3m . </pre> jidanni< at >jidanni.org 2010-11-27T03:41:49 http://comments.gmane.org/gmane.emacs.w3m/8784 <pre>X-debbugs-Cc: emacs-w3m&lt; at &gt;namazu.org, dickey&lt; at &gt;invisible-island.net Package: w3m Version: 0.5.2-10 Severity: wishlist It seems w3m and thus emacs-w3m does not treat &lt;th&gt; characters differently than &lt;td&gt;, -o color=1 or not. Lynx does. In Firefox they are rendered just like &lt;strong&gt;, which would be a good idea for lynx and w3m too, that way they could still be distinguished on e.g., black and white printers. </pre> jidanni< at >jidanni.org 2010-11-18T20:54:06 http://comments.gmane.org/gmane.emacs.w3m/8783 <pre>MediaWiki ( put some titles in the box on http://radioscanningtw.jidanni.org/index.php?title=Special:Export ) sends headers Content-disposition: attachment;filename=%E5%8F%B0%E6%8E%83-20101118025928.xml Content-Type: application/xml; charset=utf-8 but emacs-w3m ignores them and plasters the xml all around the buffer as if it was HTML, instead of treating it properly as a 'Download'. Maybe it was the little d in disposition? At least with my http://jidanni.org/comp/configuration/.emacs-w3m . On the other hand Firefox of course properly asks if it should save it in a file. Also, http://bugs.debian.org/cgi-bin/bugreport.cgi?msg=5;filename=googleearth-package-0.5.7.patch;att=1;bug=596423 has Content-Disposition: inline; filename="googleearth-package-0.5.7.patch" Content-Type: text/x-diff Well, emacs-w3m does ask about saving it in a file. But the filename it offers is "bugreport.cgi'!! Of course Firefox naturally gets this one right too. </pre> jidanni< at >jidanni.org 2010-11-18T03:21:29 http://comments.gmane.org/gmane.emacs.w3m/8782 <pre>(info "(emacs-w3m) Creating and killing buffers") doesn't say how to create a fresh empty buffer. All it says is that one can clone only. Confirmed via C-c C-h which shows no equivalent to Firefox or IE's 'create new tab' or 'create new window' commands. No we don't always want to make a whole new copy of an unrelated buffer just to get a new buffer! emacs-w3m-version =&gt; "1.4.400" </pre> jidanni< at >jidanni.org 2010-11-17T12:40:11 http://comments.gmane.org/gmane.emacs.w3m/8774 <pre>Viewing http://www.couchsurfing.org/profile.html?id=5KEB3I0 with images will make emacs-w3m take so long that one ends up needing to killall emacs. Note "This member has chosen to show their profile only to members who are logged in. If you are a member, please log in to view this profile." so one needs to signup first. </pre> jidanni< at >jidanni.org 2010-11-10T20:44:10 http://comments.gmane.org/gmane.emacs.w3m/8773 <pre>Reading this message in gnus/emacs-w3m, no matter if as text (lots of lines mentioning .gifs) or as HTML (so few links to click, all just [img]), the results are bad. $ apt-show-versions $&lt; at &gt; emacs-snapshot/sid uptodate 1:20101030-1 w3m/unstable uptodate 0.5.2-10 w3m-el-snapshot/unstable uptodate 1.4.400+0.20100725-1 </pre> jidanni< at >jidanni.org 2010-11-10T20:15:59 http://comments.gmane.org/gmane.emacs.w3m/8765 <pre>There might be a missing quote on the w3m-condition-case lisp-indent-function per below. I noticed this when I made the same mistake in my own code and did a grep to see if anyone else may have :-). </pre> Kevin Ryde 2010-11-02T23:31:17 http://comments.gmane.org/gmane.emacs.w3m/8759 <pre>Can anybody see the images in http://www.tpsea.org.tw/ properly on a 43 line tall screen? They just keep jumping around when you try to scroll the page. emacs-w3m-version "1.4.400" </pre> jidanni< at >jidanni.org 2010-10-24T17:12:16 http://comments.gmane.org/gmane.emacs.w3m/8755 <pre>For a while now, I wondered why sometimes org-capture would throw an error when I entered new notes. I finally found that in w3m-bookmark.el, there's (defvar bookmark-make-record-function nil) which of course breaks any package using this when w3m-bookmark was loaded first, including org-capture. -David </pre> David Engster 2010-10-21T11:39:11 http://comments.gmane.org/gmane.emacs.w3m/8754 <pre>Hi nnshimbun users, Some shimbun modules generate an article containing two or more html parts if the original web contents consist of many pages, though such ones may not be general. To delimit pages, every succeeding html part has "&amp;#012;" (i.e. `^L') in the beginning of an html body. But you may see those `^L's are missing nowadays when displaying, if you use the most recent Gnus and Emacs. It is because the default value of `mm-text-html-renderer' has changed into `shr', and `libxml-parse-html-region' that `mm-shr' uses simply ignores `^L's. This advice will help: (defadvice gnus-mime-display-mixed (around add-page-delimiter (handles) activate) "Add page delimiters to the beginning of succeeding html parts." (if (eq mm-text-html-renderer 'shr) (progn (gnus-mime-display-part (pop handles)) (while handles (if (equal (mm-handle-media-type (car handles)) "text/html") (insert "\C-l\n")) (gnus-mime-display-part (pop handles)))) ad-do-it)) </pre> Katsumi Yamaoka 2010-10-21T07:54:30 http://comments.gmane.org/gmane.emacs.w3m/8753 <pre>Here's what the HTML part of a facebook message looks like if I save it to a file and browse it: +---------------------------------------------------------------------------------------------------------------------+ | | | +--------------------------------------------------------------------------------------+ | | |facebook | | | | |--------------------------------------------------------------------------------------| | | |+------------------------------------------------------------------------------------+| | | ||幸蒨 sent you a message. || | | ||+----------------------------------+ </pre> jidanni< at >jidanni.org 2010-10-20T14:46:30 http://comments.gmane.org/gmane.emacs.w3m/8752 <pre>Hi, `C-c C-s' is my favorite way to navigate my w3m buffers, but with lots of buffers, using &lt;up&gt; and &lt;down&gt; can be a little bit tedious. I found no obvious way to navigate to buffer n with a number prefix, so I quickly tried this and it seems to work: --8&lt;---------------cut here---------------start-------------&gt;8--- (define-key w3m-select-buffer-mode-map "g" (lambda (arg) (interactive "p") (goto-line arg) (w3m-select-buffer-show-this-line))) --8&lt;---------------cut here---------------end---------------&gt;8--- But I am still puzzled that this does not seem to work out of the box. Did I miss anything in the documentation? Memnon </pre> Memnon Anon 2010-10-15T19:56:20 http://comments.gmane.org/gmane.emacs.w3m/8755 <pre>For a while now, I wondered why sometimes org-capture would throw an error when I entered new notes. I finally found that in w3m-bookmark.el, there's (defvar bookmark-make-record-function nil) which of course breaks any package using this when w3m-bookmark was loaded first, including org-capture. -David </pre> David Engster 2010-10-21T11:39:11 http://comments.gmane.org/gmane.emacs.w3m/8754 <pre>Hi nnshimbun users, Some shimbun modules generate an article containing two or more html parts if the original web contents consist of many pages, though such ones may not be general. To delimit pages, every succeeding html part has "&amp;#012;" (i.e. `^L') in the beginning of an html body. But you may see those `^L's are missing nowadays when displaying, if you use the most recent Gnus and Emacs. It is because the default value of `mm-text-html-renderer' has changed into `shr', and `libxml-parse-html-region' that `mm-shr' uses simply ignores `^L's. This advice will help: (defadvice gnus-mime-display-mixed (around add-page-delimiter (handles) activate) "Add page delimiters to the beginning of succeeding html parts." (if (eq mm-text-html-renderer 'shr) (progn (gnus-mime-display-part (pop handles)) (while handles (if (equal (mm-handle-media-type (car handles)) "text/html") (insert "\C-l\n")) (gnus-mime-display-part (pop handles)))) ad-do-it)) </pre> Katsumi Yamaoka 2010-10-21T07:54:30 http://comments.gmane.org/gmane.emacs.w3m/8753 <pre>Here's what the HTML part of a facebook message looks like if I save it to a file and browse it: +---------------------------------------------------------------------------------------------------------------------+ | | | +--------------------------------------------------------------------------------------+ | | |facebook | | | | |--------------------------------------------------------------------------------------| | | |+------------------------------------------------------------------------------------+| | | ||幸蒨 sent you a message. || | | ||+----------------------------------+ </pre> jidanni< at >jidanni.org 2010-10-20T14:46:30 http://comments.gmane.org/gmane.emacs.w3m/8752 <pre>Hi, `C-c C-s' is my favorite way to navigate my w3m buffers, but with lots of buffers, using &lt;up&gt; and &lt;down&gt; can be a little bit tedious. I found no obvious way to navigate to buffer n with a number prefix, so I quickly tried this and it seems to work: --8&lt;---------------cut here---------------start-------------&gt;8--- (define-key w3m-select-buffer-mode-map "g" (lambda (arg) (interactive "p") (goto-line arg) (w3m-select-buffer-show-this-line))) --8&lt;---------------cut here---------------end---------------&gt;8--- But I am still puzzled that this does not seem to work out of the box. Did I miss anything in the documentation? Memnon </pre> Memnon Anon 2010-10-15T19:56:20 http://comments.gmane.org/gmane.emacs.w3m/8749 <pre>Hi, I found that a function `w3m-tab-previous-buffer' causes an error "Wrong type argument: number-or-marker-p, nil" if it is called with no arguments non-interactively like "(w3m-tab-previous-buffer)". Currently the code is as follows: (defun w3m-tab-previous-buffer (&amp;optional n event) "Turn N pages of emacs-w3m buffers behind." (interactive (list (prefix-numeric-value current-prefix-arg) last-command-event)) (w3m-tab-next-buffer (- n) event)) Here, "(- n)" cannot be calculated if N is omitted. To avoid the error, it should be modified as: (defun w3m-tab-previous-buffer (&amp;optional n event) "Turn N pages of emacs-w3m buffers behind." (interactive (list (prefix-numeric-value current-prefix-arg) last-command-event)) (w3m-tab-next-buffer (- (or n 1)) event)) Thanks. IRIE Shinsuke </pre> IRIE Shinsuke 2010-10-13T22:39:42 http://comments.gmane.org/gmane.emacs.w3m/8737 <pre>With the most recent nognus (44e4398abd1f2d11b3f32d64c54a6366104cf485), all nnshimbun groups in a server are fetched at startup, regardless of whether one has subscribed to them or not. For instance, I subscribe to a single group in sb-nytimes, but, at startup, the nognus fetches and generates nov databases for all the rss feeds for all groups defined in the module (quite a few). I asked about this on the gnus development list and I have been advised that the "new" gnus calls nnshimbun-request-scan without a group parameter: http://permalink.gmane.org/gmane.emacs.gnus.general/72202 Note: a previous fix for this problem in gnus-start.el was reverted because it seems to have caused problems for other backends: http://permalink.gmane.org/gmane.emacs.gnus.general/72195 Looking at nnshimbun-request-scan, I see that if the group parameter is nil, nnshimbun calls shimbun-groups to find *all* groups defined in the server rather than only subscribed groups. Would it be possible for nnshimbun-request-scan to gene</pre> Matt Lundin 2010-09-30T00:41:20 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.emacs.w3m