Re: [Conkeror] C-x-b on Twitter page blocks people
On Thu, Feb 14, 2013 at 10:00:44AM -0600, aditya siram wrote:
Hi all,
I've had this happen 3 times over the past 2 weeks where if I am reading a
conversation on Twitter and I do a C-x-b to switch to the previous buffer,
Twitter interprets the `b` as `Block` and automatically blocks the
conversation
starter. This is somewhat embarrassing and I only knew about it when someone
asked me why I had blocked them.
Does Twitter mode fix this?
Thanks!
-deech
Try using key-kill-mode. I haven't added the mode to Conkeror yet, so
here it is and you can just put it in your rc, and add a test for twitter.
function key_kill_event_kill (event) {
var elem = event.target;
if (elem instanceof Ci.nsIDOMHTMLInputElement ||
elem instanceof Ci.nsIDOMHTMLTextAreaElement)
{
return;
}
event_kill(event);
}
define_page_mode(key-kill-mode,
[],
function enable (buffer) {
buffer.browser.addEventListener(keyup, key_kill_event_kill, true);
buffer.browser.addEventListener(keydown, key_kill_event_kill, true);
},
function disable (buffer) {
buffer.browser.removeEventListener(keyup, key_kill_event_kill,
true);
buffer.browser.removeEventListener(keydown, key_kill_event_kill,
true);
},
$display_name = Key-kill);
page_mode_activate(key_kill_mode);
key_kill_mode.test.push(build_url_regexp($domain = www.smbc-comics));
key_kill_mode.test.push(build_url_regexp($domain = github));
key_kill_mode.test.push(/\/\/.*slashdot\.org\//);
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] add super modifier
On Thu, Feb 14, 2013 at 10:14:10PM +0700, Nguyễn Tuấn Anh wrote: This allows using the Windows-logo key as a modifier. It used to trigger meta modifier, but changed to os modifier in recent xulrunner versions (probably from 12). See http://bugs.conkeror.org/issue410 Thank you. I hadn't noticed the patch that you posted to issue410, but I have now responded there on the bug tracker. This patch does not exactly solve that issue, but once we decide what we want to do about it, this patch with some modifications will. Let's continue the discussion on the tracker now. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] reddit.js: add support for comments
On Wed, Feb 13, 2013 at 04:11:11AM +0100, Joren Van Onder wrote: On Wed, Feb 13, 2013 at 3:38 AM, John J. Foerch jjfoe...@earthlink.net wrote: On Mon, Feb 04, 2013 at 05:01:26PM +0100, Joren Van Onder wrote: Features: - scroll through comments with 'j' and 'k' - scroll through parent comments with 'J' and 'K' - vote on comments with ',' and '.' - automatically load more comments on end of page For a large patch like this, we should update the legalese of reddit.js. Do you want copyright on your patch, or to assign it to someone else? Okay, I'd like copyright on my patch. Will you update it? Or should I do it and send it again? I'll do that. Can you stop by the #conkeror irc channel some time to chat about the patch? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] reddit.js: add support for comments
On Mon, Feb 04, 2013 at 05:01:26PM +0100, Joren Van Onder wrote: Features: - scroll through comments with 'j' and 'k' - scroll through parent comments with 'J' and 'K' - vote on comments with ',' and '.' - automatically load more comments on end of page For a large patch like this, we should update the legalese of reddit.js. Do you want copyright on your patch, or to assign it to someone else? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] jumps not being read
On Tue, Feb 05, 2013 at 12:37:38PM -0800, Silvio Levy wrote: My conkeror rc is the directory /u/levy/.c and it is apparently being read, because most of my customizations work. Also, M-x reinit prints Loaded: /u/levy/.c/*.js. However, the file .c/jumps.js, containing my web jumps, doesn't seem to be read: all lines added in the last several weeks do not work. Yet older jumps do work. Note: the web jumps that don't work are near the top of .c/jumps.js, so the problem is not that the file is being only partially read. It's not being read at all. (The read time of the file doesn't get updated when conkeror is started.) Yet some older jumps do work. Any idea what's going on, or how I might debug it? output of conkeror -v: conkeror.mozdev.org conkeror 1.0pre called with the command exec xulrunner /usr/local/src/conkeror/application.ini $@ Thanks, Silvio Sounds like the startup-cache at work. Try evaluating something with M-: (a simple number will do), and that will clear the startup cache. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] jumps not being read
On Tue, Feb 05, 2013 at 02:15:33PM -0800, Silvio Levy wrote: Sounds like the startup-cache at work. Try evaluating something with M-: (a simple number will do), and that will clear the startup cache. Thanks for the reply. Alas, this is not it: 1. evaluating an expression and running reinit makes no difference 2. quitting conkeror and restarting makes no difference 3. as reported in the previous message, jumps.js is not being read when conkeror starts. From googling startup cache I gather than the startup cache is supposed to last only until the end of the session. So, something else is hiding jumps.js, and using jumps from some other source. Please see log below of a session where I cleaned out my conkerorrc for the purpose. Note that when I restore my usual conkerorrc, other things such as preferences and key bindings work as far as I can tell; only the jumps behave strangely. Thanks, Silvio gar:; ps aux | grep xulr # make sure no processes survive levy 21871 0.0 0.0 4660 816 pts/15 S+ 19:49 0:00 grep xulr gar:; ps aux | grep conk # make sure no processes survive levy 21873 0.0 0.0 4660 816 pts/15 S+ 19:49 0:00 grep conk gar:; ls .c # removed all other files from conkerorrc jumps.js gar:; cat .c/jumps.js# only a single line define_webjump(foo,file:///u/levy/foo.html); gar:; xulrunner /usr/local/src/conkeror/application.ini # no errors or warnings at initialization C-xC-f about:config confirms that conkeror.rcfile is set to /u/levy/.c C-xC-f foo goes to foo.com, ignoring the define_webjump command in jumps.js C-xC-f sub goes to http://www.tvsubtitles.net/, a months-old web jump. The startup-cache does persist across sessions unless explicitly cleared. How old is your conkeror? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] jumps not being read
On Tue, Feb 05, 2013 at 03:13:57PM -0800, Silvio Levy wrote: The startup-cache does persist across sessions unless explicitly cleared. Good to know. But if I understand your first message, evaluating an expression with C-: is supposed to clear it, right? How old is your conkeror? I had caught up with the current version maybe two months ago, and I just did it again now using sudo -s; cd /usr/local/src/conkeror; git fetch; git merge origin No change in behavior. Another clue, though, is that changes in bindings.js are likewise not taking effect (which I hadn't noticed since I seldom make any). So yes, it's probably something to do with the startup-cache. But evaluation with M-: definitely doesn't help. Please tell me: 1. Is there a command to clear the startup-cache explicitly? 2. How can I load a specific file by hand? (googling didn't help; M-x load TAB gives nothing) Silvio The javascript code to clear the startup-cache is as follows: var obs = Cc[@mozilla.org/observer-service;1] .getService(Ci.nsIObserverService); obs.notifyObservers(null, startupcache-invalidate, null); You can put this in a command or a function for easy access. Also, your webjump is called foo? Check the value of webjumps.foo with the M-: prompt, and see whether it shows up. What version of XULRunner was it? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] zombie tabs
On Tue, Feb 05, 2013 at 06:53:42PM -0800, Silvio Levy wrote: Since I have your attention, let me report another problem, which while not critical is annoying enough. I use conkeror in a single window, with multiple tabs. Sometimes, when I close the active tab, its marker remains on the tab bar, though it moves to the left of the newly active tab. The buffer itself is no longer accessible -- either by clicking on the tab bar, or by cycling through the tabs. So there is no way to get rid of the zombie tab except by restarting conkeror. When I load the previous session, the zombie is gone. Certain tasks I do require a lot of tab opening and closing, so I get a string of zombies, and end up restarting conkeror more often than I'd like. I can't figure out why sometimes a zombie is left and sometimes not. Don't know. First I've heard of such a problem. Make use of bugs.conkeror.org pls. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Ghostery - 302 moved
On Fri, Jan 11, 2013 at 12:39:39PM +0100, Trash Folder wrote:
Hi,
I have started to use Conkeror just for a few weeks and i find it really good.
But i miss the extension from Firefox called Ghostery, to block trackers and
cookies.
I have downloaded from Mozilla Addons:
https://addons.mozilla.org/en-US/firefox
/addon/ghostery/
And edited the .xpi (deleting the MANIFESTs and adding the next paragraph to
install.rdf).
!-- Conkeror --
em:targetApplication
Description
em:id{a79fe89b-6662-4ff4-8e88-09950ad4dfde}/em:id
em:minVersion0.1/em:minVersion
em:maxVersion9.9/em:maxVersion
/Description
/em:targetApplication
After the installation and to allow open the settings i have added to the
conkerorrc:
interactive(ghostery,
Open Ghostery settings.,
function (I) {
browser_object_follow(I.buffer, OPEN_NEW_BUFFER, 'chrome://ghostery/
content/options.html');
});
It seems to work (disabling everything in trakers and cookies), but then, when
im surfing... problems:
A simple search on Google, like viewmarks, it doesn't matter what link i try
to access... always the same error:
302 Moved
The document has moved here.
And the link here doesn't work neither. Even if i unselect all the trackers
and cookies from settings and save, still doesn't work. The only way to can
surf normally is disabling Ghostery from extensions menu.
Any help?
Would you document what you have learned so far on the wiki?
http://conkeror.org/Extensions
Signup instructions at:
http://conkeror.org/WikiAccess
Thanks,
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Can Conkeror provide Firefox's Save As: Web Page, Complete function?
On Sat, Dec 29, 2012 at 05:01:54PM -0500, Sean McAfee wrote: The subject line pretty much says it all, but: I'm looking to have Conkeror save a web page in the same way that Firefox does when you click the Save Page As... command and choose Format: Web Page, Complete. This means that a) a directory named after the page but with a _files suffix is created, copies of all of the resources needed by the page are saved there, and resource links on the page are rewritten to point there, and b) the saved page reflects any changes wrought by Javascript since the page was loaded. M-x save-page-complete -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Tracking keypresses independently of chosen keyboard layout
On Fri, Dec 21, 2012 at 10:58:25AM +0600, Artyom Shalkhakov wrote: Hello all, I am having this problem: if I switch my keyboard to Russian, then key some combinations don't work. For instance, the letter 'n' is mapped to letter 'т' in Russian, and when I press M-т I get the message that M-т is undefined. Is it possible to make Conkeror react to 'т' as if it was 'n' if it is pressed with a one of the modifier keys? And likewise for other letters? FYI, I am using Conkeror on Windows, but would like to hear any advice (perhaps for Linux, it may serve as an inspiration) nonetheless. Hello, It would be necessary to make equivalent key bindings for all of the russian letters to the latin counterparts. If you, or someone does this, it could perhaps be included in Conkeror as a module. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] pdf viewer extension
On Wed, Dec 19, 2012 at 09:12:48PM +0100, Stefan Husmann wrote:
Hello,
maybe this is of broader interest: I just tried this firefox extension in
conkeror:
https://addons.mozilla.org/firefox/downloads/file/181716/pdf_viewer-0.7.1.xpi
It opens pdf files directly in conkeror. This is what I added to the
install.rdf:
!-- Conkeror --
em:targetApplication
Description
em:id{a79fe89b-6662-4ff4-8e88-09950ad4dfde}/em:id
em:minVersion0.8/em:minVersion
em:maxVersion2.0/em:maxVersion
/Description
/em:targetApplication
That's it!
Kind Regards
Stefan
Have a wiki account? We document this kind of information at:
http://conkeror.org/Extensions
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] refactor-completers branch
Here is the complete breaking-changes list for this branch, so far: - define_webjump: - '$argument' keyword removed. its value is now automatically computed, except in the case of function webjumps that take an optional argument but do not give an alternative. in those cases, give '$alternative = true'. - keyword '$alternative = true' means to use the primary function as the alternative as well. - accepts the keyword '$require_match'. - '$description' keyword renamed to '$doc'. - completers are now instantiated with the js keyword 'new'. - all instances of the '$match_required' keyword have been renamed to '$require_match'. - completers no longer take the keyword '$match_required' (or '$require_match'). - 'C-k' in a read-url prompt now deletes the selected bookmark or history item. - completers are now classes instead of functions. refer to completers.js or ask a qualified representative for how to update your completers. - completers' 'complete' method no longer takes a third argument, 'conservative'. - javascript_completer now works in any scope. - javascript_completer no longer has special handling for the symbols 'buffer' and 'window'. ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] refactor-completers branch
Hi all, I'm currently working on a big refactoring of the completions system. The basic goal is to have a completions system that is easier to read and understand than the current one, and more flexible. Then to write a nice file-path completer, and more cool completer-based features. One little thing that I added already is the ability to delete bookmarks and history items while in the url prompt with C-k, analogous to how you can delete buffers in the switch-to-buffer prompt. Well, this refactoring is a very large, complicated undertaking, and could use the help of anybody out there in the community who wants to test and review. I haven't compiled a complete list of breaking changes that are currently, or will end up being, in the branch, but if you're testing and have things that use completers in your rc, the biggest ones that might affect you are: - completers are now classes, and instantiated with the 'new' keyword - $match_required keywords renamed to $require_match - webjump keyword $description renamed to $doc So, check it out, check my work, and hopefully we can get this thing merged soon. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Reddit mode
On Fri, Dec 07, 2012 at 10:32:31PM -0600, Jay Belanger wrote: Hi Everybody, Is anyone else using Reddit mode with reddit_end_behavior=page; When going through the Reddit entries (with j), this is supposed to go to the next page after the last entry of the current page. Lately, for me, after the last entry, I get the message call interactively: TypeError: node is null If I try again, I get sent to the first entry of the current page. I don't know enough javascript to figure out the problem, but I was wondering if it was a local issue with me or some other problem. Jay Looks like the xpath expression that locates the next link needs to be updated. Will have a patch pushed shortly. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Reddit mode
On Sat, Dec 08, 2012 at 12:11:14AM -0500, John J. Foerch wrote: On Fri, Dec 07, 2012 at 10:32:31PM -0600, Jay Belanger wrote: Hi Everybody, Is anyone else using Reddit mode with reddit_end_behavior=page; When going through the Reddit entries (with j), this is supposed to go to the next page after the last entry of the current page. Lately, for me, after the last entry, I get the message call interactively: TypeError: node is null If I try again, I get sent to the first entry of the current page. I don't know enough javascript to figure out the problem, but I was wondering if it was a local issue with me or some other problem. Jay Looks like the xpath expression that locates the next link needs to be updated. Will have a patch pushed shortly. pushed -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] tabs on the left
On Thu, Dec 06, 2012 at 03:48:35PM +, Georg wrote: With firefox, I use the ``Vertical Tabs'' extension to show tabs on the left of the browser window. I would like to modify tab-bar.js to archive such a behaviour. Changing line 15 to element.setAttribute(orient, vertical); seems to be a good start. What I need now is a way to tell conkeror to put the arrowscrollbox to the left, instead of to the top of the main window. Does anybody know how to do this? Kind regards, Georg There will need to be a container with horizontal orientation around #buffer-container that the tab-bar can be added to. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Conkeror Digest, Vol 96, Issue 1
On Thu, Dec 06, 2012 at 02:23:01PM -0600, Benjamin Slade wrote: From: Axel Beckert a...@deuxchevaux.org [add_delicious_webjumps(username); was] deprecated quite some time now and now have been removed finally. Use define_delicious_webjumps and define_lastfm_webjumps instead. See also http://conkeror.org/BreakingChanges Is there information on how to set it up the new way? I replaced add_delicious_webjumps(username); with define_delicious_webjumps(username);, and---while this doesn't generate any errors---it also doesn't seem to actually give me any delicious webjumps (in contrast to the old add_...). The call form is the same. The old add_* functions were just aliases. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Conkeror Digest, Vol 96, Issue 1
On Thu, Dec 06, 2012 at 04:06:59PM -0600, Benjamin Slade wrote: The call form is the same. The old add_* functions were just aliases. But it didn't automatically generate delicious or adelicious etc. as possible webjumps, like the old add_* function did. Perhaps there was another problem in the rc that prevented it from evaluating. Try calling it with M-: -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Running Conkeror on Windows
On Sun, Nov 11, 2012 at 09:32:44PM +0600, Artyom Shalkhakov wrote:
2012/11/11 John J. Foerch jjfoe...@earthlink.net:
On Sun, Nov 11, 2012 at 04:18:38PM +0600, Artyom Shalkhakov wrote:
Hello all,
I'm running Conkeror on Windows 7 via Emacs' browse-url function with
the following command line:
c:/path/to/xulrunner/xulrunner.exe -app
c:/path/to/conkeror/application.ini url
(The recipe given on the wiki didn't work for me. Could anybody tell
how to read error messages from Conkeror?)
It works, but it seems that xulrunner redirects all command arguments
to Conkeror, including the first two, so Conkeror tries to open the
application.ini file (and fails). For the moment, I've resolved it
with the following addition to modules/command-line.js:
// added the following dummy command switch
// to prevent Conkeror trying to open it's application.ini file
// when it is invoked from within Emacs
command_line_param_handler(app, false, function (expr, ctx) {});
So, now I'm wondering if there is any way to run Conkeror without
having to add this fix.
Also, on the wiki, there is a mention of unsupported external editing
on Windows. Could anybody tell where to start looking to resolve the
issue?
Finally, would like to thank you people for this browser. Great work.
Hello John,
Does the xulrunner-stub method work? That is where you put
xulrunner-stub.exe into the conkeror directory, and name it conkeror.exe.
Thanks, this worked. I also followed a hint from
https://github.com/jungleji/chmsee:
1. copy xulrunner-stub.exe to conkeror.exe (at a DOS prompt):
copy path\to\xulrunner-stub.exe path\to\conkeror\conkeror.exe
2. create a symlink to xulrunner directory:
mklink /D xulrunner path\to\xulrunner
The second step gets rid of the Can't find Mozilla runtime message.
Now I can run conkeror.exe.
What version of xulrunner is it, btw?
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Any way to take screenshots of individual page elements?
On Thu, Oct 25, 2012 at 05:30:01PM -0700, Scott Jaderholm wrote: On Thu, Oct 25 2012,Sean McAfee wrote: Thanks for the tip, and I'll investigate such a solution if necessary, but I was thinking more of getting Conkeror to generate images for elements internally, which would make things much easier if there are many elements to capture, if an element lies outside the current view, if any elements are larger than the screen, etc. Ultimately, I'd like to have a function that takes an array of DOM elements and emits image representations of those elements, Doesn't help with doing this from Conkeror, but something like this appears to be a new feature in Firefox 16 via Shift-F2: screenshot heading.png 0 false h1 https://hacks.mozilla.org/2012/08/new-firefox-command-line-helps-you-develop-faster/ Perhaps this is the inspiration for your question. I thought I'd share in case not. Scott This is excellent news. Just need to look up the source of that command, and then it should be a simple thing to write an equivalent command for Conkeror. -- John ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Any way to take screenshots of individual page elements?
On Sun, Oct 21, 2012 at 03:58:32PM -0700, Sean McAfee wrote: Here's something I've been pondering how to accomplish for a while: Having designated an arbitrary set of elements on the current page, how can I obtain individual images of those elements? If there's no way to do this in Conkeror, then pointers to alternate means (Firefox plugins, or whatever) would also be appreciated. It is possible to get absolute screen coordinates for any given DOM node - node.getBoundingClientRect or somesuch. These coordinates may need to be offset by the x and y left and top coordinates of the window. Once you have that, you can use scrot or ImageMagick import or such to take a screenshot of the rectangle that you have calculated. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Files don't open after download
On Sat, Oct 20, 2012 at 06:26:07PM +0100, Johnny wrote: Hi, Recently, probably linked with my recent re-install of F17, conkeror no longer opens pdf's (or other file types) after downloading them. After download is finshed, conkeror asks to open the file with mupdf as I have configured it, but nothing happens. i am using conkeror 1.0pre on XULrunner 15.0.1. I am puzzled as I haven't change any configs... Thanks for any help! Make sure that conkeror-spawn-helper is installed or has been built in the conkeror directory. Try rebuilding it. It doesn't have to be installed to a PATH directory, but if there is one in the path, be sure to update it. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] GMail focus issues
On Wed, Oct 03, 2012 at 09:16:38AM -0400, Tim Ramsey wrote: Hello Conkeror Community, Recently I have been having troubles in GMail mode in Conkeror. The focus seems to have changed such that the keyboard shortcuts no longer work without manually clicking elements to move the focus (and therefore defeat the purpose of a keyboard shortcut). A typical example is when I respond to a message (keyboard shortcut r) it used to go to the reply screen and the focus would be in the text area ready to compose a new message. Now, when I type 'r' on a message the response screen comes up and I can see the focus go to the text area but then immediately the text area loses focus. To start composing a message I have to click back in the text area so that it once again has the focus. That is only one example of several little annoyances that seem to have popped up overnight. I am on a Windows 7 machine and running the Firefox exe (with the --app switch) version 15.0.1. My conkeror HEAD is at: commit 34fbea2b8833ad35de12b3718ddff6e232efa060 Author: Axel Beckert a...@deuxchevaux.org Date: Sat Sep 1 11:36:29 2012 +0200 Debian package: Add support for xulrunner 17 Any one else experience these? Anyone have any fixes? Thanks, -Tim Ramsey The first thing to try is to see if some small change to gmail_focus_primary_frame (like frame number) resolves it. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] GMail focus issues
On Mon, Oct 15, 2012 at 10:15:46PM +0200, Jean-Baptiste Mestelan wrote: On 15 October 2012 22:00, Jean-Baptiste Mestelan meste...@gmail.com wrote: I should probably not have mentioned the cursor keys: this has to be a different issue altogether. Let's make this clearer: Two issues have been discussed about the GMail mode: 1) the navigation keys (gi, gl, r, c...) work only after a manual mouse focus. Both patches solve this issue. 2) The scrolling keys (C-n, C-p, PgUp, PgDown, Home, End) never work, whichever frame gets the focus. Neither of the patches solve this issue. I could not tell how much these issues are related. I have pushed a patch. Please test it. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] image downloading
I would like to have such a note on the wiki, but do we really know enough about the problem to say anything authoritative about it? Vague documentation can become a problem in its own right, because generalizations are difficult to fact-check later on, and when there may be differences from version to version. bugs.conkeror.org may be the better place until we know more. On Mon, Sep 03, 2012 at 10:22:24AM -0500, Benjamin Slade wrote: Hoorah! I can download images again. Is it perhaps appropriate to put a note on conkeror.org somewhere advising against using the firefox --app way of running conkeror, at least for the moment? Any firefox from 14+ is really buggy with conkeror. John J. Foerch writes: Download the binary from mozilla and extract it to a directory. Launch conkeror with a script like: exec /path/to/xulrunner/xulrunner-bin /path/to/conkeror/application.ini $@ On Sun, Sep 02, 2012 at 10:42:09PM -0500, Benjamin Slade wrote: Then, does anyone know an easy way of installing xulrunner directly on an Ubuntu-based distro? ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Segmentation fault
On Thu, Aug 23, 2012 at 10:32:35AM -0700, Silvio Levy wrote: Hi John Maybe your profile got corrupted. Try a new profile. That helps. It now crashes a bit later - see below. But first: /* SUGGESTION: Yesterday I tried numerous google searches. Most helpful would have been a page under conkeror.org with the words Segmentation fault and the hint about profiles. (The search box on top returns noting under Segmentation fault and clicking Text.) Translation: this problem would be a lot easier to solve if someone else had it before me and documented the solution. Yes. /sarcasm I did try yesterday to move my Conkeror directory out the way, but the profile stuff is in another directory and I didn't think of it. */ Now back to the PROBLEM: While I can now view some pages with conkeror, most pages (anything that's not pure HTML, it looks like) cause this: Segmentation fault (core dumped) No core file seems to be created, and the are no error messages. (See below for a transcript of a conkeror run -- all the messages, apart from the Seg fault, occur at startup time.) This is the same behavior I noticed with firefox yesterday, so the problem is not conkeror per se. Still I'm hoping that you might have some ideas about how to troubleshoot. Silvio gar:; xulrunner /usr/local/src/conkeror/application.ini -P levy TypeError: ar is undefined chrome://conkeror/content/array.js:50 array_find_index((void 0),(function (x) x.name == func.name))@chrome://conkeror/content/array.js:50 replace_hook(content_policy_hook,block_by_host)@file:///usr/local/src/conkeror/components/application.js - file:///u/levy/.c/content-policy.js:9 @file:///usr/local/src/conkeror/components/application.js - file:///u/levy/.c/content-policy.js:46 Segmentation fault (core dumped) I don't know what caused that segfault, but the error before it is probably because you forgot to: require(content-policy); before using content_policy_hook in your rc. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Segmentation fault
On Thu, Aug 23, 2012 at 11:11:43AM -0700, Silvio Levy wrote: Translation: this problem would be a lot easier to solve if someone else had it before me and documented the solution. Not sure why this deserves sarcasm. Is this really the first case of Segmentation fault that has come to the developers' attention? I can document the solution, if you'd like. (BTW you wrote Maybe your profile got corrupted. Try a new profile. It's not obvious how you do this if you can't even get conkeror started. In the event I found /home/levy/.conkeror.mozdev.org/conkeror/profiles.ini and changed StartWithLastProfile=1 to 0. Not sure if that's the best way.) I don't know what caused that segfault, but the error before it is probably because you forgot to: require(content-policy); before using content_policy_hook in your rc. No - it's there (line 1 of my content-policy.js) Silvio Sorry if that came off too harsh. Changing profiles is a very general diagnostic step, not specific to segfaults, and to my knowledge, we have seen no cases where changing profiles solved a segfault. I'm out of ideas for your problem. It sounds like something is corrupted somewhere. Reinstall everything, try different versions of xulrunner (or firefox). Fresh profile. Maybe corruption in some other library that xulrunner/firefox depends on? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] image downloading
On Fri, Aug 17, 2012 at 01:20:08PM -0500, Benjamin Slade wrote:
For a while now (a week or so) I've been getting an error message when i try
to
save an image (i-s):
The first time I try in a session, I get an error:
Console error: [JavaScript Error: [Exception... Component returned failure
code: 0x80520012 (NS_ERROR_FILE_NOT_FOUND) [nsIXPCComponents_Utils.import]
nsresult: 0x80520012 (NS_ERROR_FILE_NOT_FOUND) location: JS frame ::
resource://gre/modules/XPCOMUtils.jsm :: XPCU_moduleLambda :: line 222 data:
no] {file: resource://gre/modules/XPCOMUtils.jsm line: 222}]
The second time I try in a session, I get the error:
call interactively: [Exception ... '[JavaScript Error: DownloadsCommon is
not defined {file: resource:///components/DownloadsUI.js line: 11}]' when
calling method: [nsIDown...
(screen cuts off remainder)
in the commandline I see the errors:
Console error: [JavaScript Error: DownloadsCommon is not defined {file:
resource:///components/DownloadsStartup.js line: 134}]
Category: XPConnect JavaScript
Category: xbl javascript
Console error: [JavaScript Warning: Use of attributes' nodeValue attribute is
deprecated. Use value instead. {file: http://pluck.nationalgeographic.com/
ver1.0/Content/ua/scripts/pluckApps.js line: 763}]
Category: DOM Core
download state change: http://images.nationalgeographic.com/wpf/media-live/
photos/000/580/cache/hippo-pod-africa_58004_990x742.jpg: -1, 5, 3533
NS_ERROR_FILE_NOT_FOUND: Component returned failure code: 0x80520012
(NS_ERROR_FILE_NOT_FOUND) [nsIXPCComponents_Utils.import]
XPCU_moduleLambda()@resource://gre/modules/XPCOMUtils.jsm:222
null()@resource://gre/modules/XPCOMUtils.jsm:176
null()@null:0
null()@resource:///components/DownloadsUI.js:111
null()@null:0
save_uri()@chrome://conkeror/content/save.js:78
null()@chrome://conkeror/content/commands.js:486
_do_call()@chrome://conkeror/content/coroutine.js:277
null()@chrome://conkeror/content/coroutine.js:408
exit_minibuffer()@chrome://conkeror/content/minibuffer-read.js:455
null()@chrome://conkeror/content/minibuffer-read.js:459
call_interactively()@chrome://conkeror/content/interactive.js:115
_do_call()@chrome://conkeror/content/coroutine.js:277
co_call()@chrome://conkeror/content/coroutine.js:425
input_handle_keypress()@chrome://conkeror/content/input.js:302
null()@null:0
What version of XULRunner?
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] image downloading
On Fri, Aug 17, 2012 at 01:38:11PM -0500, Benjamin Slade wrote: 17 15. I'm guessing they're not supported? (17 stills seems to work better than 15. 14 didn't work at all.) So the error that you reported appears for both 15 and 17? What do you mean by 14 doesn't work at all? It fails silently? Did the appearance of this problem coincide with a XULRunner upgrade? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] CPU usage on CSS animations
On Wed, Aug 08, 2012 at 11:41:08PM +0200, Vincent Bernat wrote: ❦ 8 août 2012 22:00 CEST, Vincent Bernat ber...@luffy.cx : Please, have a look to: http://www.red-team-design.com/wp-content/uploads/2011/11/css3-slick-progress-bars-demo.html With Conkeror, I get a 20% CPU use of this page when the buffer is displayed and 10% when it is not. With Firefox, I get a 20% CPU use but 0% when the page is in another tab. I think Conkeror is missing something to tell XUL that the buffer is not displayed so that it does not play useless animations. Maybe this is related with requestAnimationFrame(). I am not familiar enough with this. Those kind of animated progress bars are becoming popular and if you have a few tabs with them, you'll get a lot of wasted CPU use. I am using xulrunner 15b3 (from mozilla.debian.net). OK, I wasn't able to sleep with this unresolved. :) So I have digged a bit and I have found that we can tell the browser object if a buffer is active or not. I have came up with this patch. I know _nothing_ about XUL and Gecko, so this is just an idea. Also, maybe there are other places where this could be applied (when Conkeror window is iconified? Or not visible?). Tell me if you want me to file a proper bug for this. Your patch is pushed. Thank you, this was a subtle bug, but a big improvement. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] CPU usage on CSS animations
On Thu, Aug 09, 2012 at 10:11:16AM +0200, Riccardo Murri wrote: Hello, I'm unsure if it's the same issue that other are discussing here, but everytime I visit this page: http://www.oci.uzh.ch/group.pages/baldridge/index.php Conkeror goes up to 100% CPU and starts eating memory like mad. If I do not kill it within a few seconds, the computer starts thrashing and the only solution is hard reboot... This does not happen with any other browser. (I'm using Ubuntu 12.04, with Conkeror installed from the oneiric package.) Cheers, Riccardo Doesn't sound like the same problem. Not sure what could cause that. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Patch for google-search-results page mode
On Tue, Jun 05, 2012 at 08:46:25AM +0200, Konrad Hinsen wrote:
Recently I switched to https transfer for most sites (using HTTPSEveryWhere),
including Google searches. This broke the page mode for Google search results
because the returned URLs start with encrypted. With the attached patch,
the search result page mode applies to those URLs as well.
Konrad.
diff --git a/modules/page-modes/google-search-results.js
b/modules/page-modes/google-search-results.js
index 307611b..126485b 100644
--- a/modules/page-modes/google-search-results.js
+++ b/modules/page-modes/google-search-results.js
@@ -62,7 +62,7 @@ var google_search_results_modality = {
define_page_mode(google-search-results-mode,
-build_url_regexp($domain = google,
+build_url_regexp($domain = /(?:encrypted.)?google/,
$allow_www = true,
$path = /search\?|cse\?/,
$tlds = [com, com.au, co.uk, de, dk, es,
Pushed, but didn't test. Made a trivial change to the regexp. Let me
know if I broke it :)
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Remove Info.plist
On Fri, Feb 24, 2012 at 06:02:20AM +, Philip Hudson wrote: Solved: xulrunner-bin --install-app does not reference our Info.plist file at all. Instead, it gets the values it puts into the deployed Info.plist from our application.ini. The deployed Info.plist file is identical whether our Info.plist file is present or not. So, unless anyone knows of any other reason for the existence of our Info.plist file, I guess it should be deleted. I got distracted from this thread and am just now coming back to it. It sounds to me like deleting Info.plist from conkeror's root directory is the right thing to do, but we also need to decide what to do about advising people of how to use/modify Info.plist. For example, should a copy of it be provided in contrib/, or a patch file that people can use? I don't have a mac, so I will be looking for advice from people who know that platform, and will do what you/they advise. Want to make the installation procedure as straight-forward as possible for users. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] TypeError: can't access dead object
On Tue, Aug 14, 2012 at 02:17:19PM +0200, Vincent Bernat wrote: Hi! I am often stuck in Conkeror with this error: TypeError: can't access dead object This is very easy to reproduce: on any page, type f, then click any link. The minibuffer will stay in the follow mode but I cannot exit this mode anymore. Even C-g does not work. The only way to exit is to close Conkeror. Pretty annoying. Xulrunner 15.0b3. Thanks! Agreed, pretty annoying. Conkeror's input system does not currently support the mouse, so don't mix mouse and keyboard. Expect mouse support in 2.0. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] TypeError: can't access dead object
On Tue, Aug 14, 2012 at 07:53:15PM +0200, Vincent Bernat wrote: ❦ 14 août 2012 19:30 CEST, John J. Foerch jjfoe...@earthlink.net : I am often stuck in Conkeror with this error: TypeError: can't access dead object This is very easy to reproduce: on any page, type f, then click any link. The minibuffer will stay in the follow mode but I cannot exit this mode anymore. Even C-g does not work. The only way to exit is to close Conkeror. Pretty annoying. Xulrunner 15.0b3. Thanks! Agreed, pretty annoying. Conkeror's input system does not currently support the mouse, so don't mix mouse and keyboard. Expect mouse support in 2.0. It also happens when using just the keyboard but it is more difficult to reproduce. Here is an example: 1. Go to www.google.com. 2. Enter any search. 3. f + 21 + Enter + f Hitting f right after Enter (before the new page starts rendering) will do the trick. Thank you. Never seen that one, but that is definitely something I would want to fix for the 1.0 release. Seems you have to get the timing just right; I haven't reproduced it yet. Go ahead an do the report on bugs.conkeror.org where it won't get lost. Thanks -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Twitter page mode
On Fri, Aug 10, 2012 at 06:53:35PM +0200, Vincent Bernat wrote:
Hi!
I am loading Twitter page mode with the following snippet:
#v+
require(twitter);
undefine_key(twitter_keymap, f);
#v-
While the `f` key can now be used for its original purpose without
prefixing it with `C-z`, Twitter still manages to receive this
keypress. I have looked at the code of Twitter and it registers an event
handler on `document.body` for `keydown`, `keypress` and `keyup`.
What would be the best way to not allowing it to receive the `f` key?
Which event does Conkeror will intercept? `keypress`?
--
printk(What? oldfid != cii-c_fid. Call 911.\n);
2.4.3 linux/fs/coda/cnode.c
Conkeror's input system only deals with keypress events, so dealing with
websites that listen to keyup and keydown is a long-standing annoyance.
There is an experimental solution, but it is known to have problems on OS X:
function input_setup_extra_key_events (window) {
window.addEventListener(keydown, input_handle_keydown, true);
window.addEventListener(keyup, input_handle_keyup, true);
}
add_hook(window_initialize_hook, input_setup_extra_key_events);
Or here is another snippet of code that can be used on a per-site
basis.. modify it for the event(s) you want to intercept and the sites you
want it to work on:
define_page_mode(keyup-kill-mode,
[],
function enable (buffer) {
buffer.browser.addEventListener(keyup, event_kill, true);
},
function disable (buffer) {
buffer.browser.removeEventListener(keyup, event_kill, true);
},
$display_name = Keyup-kill);
page_mode_activate(keyup_kill_mode);
keyup_kill_mode.test.push(build_url_regexp($domain = www.smbc-comics));
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] CPU usage on CSS animations
On Wed, Aug 08, 2012 at 11:41:08PM +0200, Vincent Bernat wrote:
❦ 8 août 2012 22:00 CEST, Vincent Bernat ber...@luffy.cx :
Please, have a look to:
http://www.red-team-design.com/wp-content/uploads/2011/11/css3-slick-progress-bars-demo.html
With Conkeror, I get a 20% CPU use of this page when the buffer is
displayed and 10% when it is not.
With Firefox, I get a 20% CPU use but 0% when the page is in another
tab.
I think Conkeror is missing something to tell XUL that the buffer is not
displayed so that it does not play useless animations. Maybe this is
related with requestAnimationFrame(). I am not familiar enough with
this.
Those kind of animated progress bars are becoming popular and if you
have a few tabs with them, you'll get a lot of wasted CPU use.
I am using xulrunner 15b3 (from mozilla.debian.net).
OK, I wasn't able to sleep with this unresolved. :) So I have digged a
bit and I have found that we can tell the browser object if a buffer is
active or not. I have came up with this patch. I know _nothing_ about
XUL and Gecko, so this is just an idea.
Also, maybe there are other places where this could be applied (when
Conkeror window is iconified? Or not visible?).
Tell me if you want me to file a proper bug for this.
From 7804cfa8318e4efb43addbc0c9cdb1d02bec461d Mon Sep 17 00:00:00 2001
From: Vincent Bernat ber...@luffy.cx
Date: Wed, 8 Aug 2012 23:33:50 +0200
Subject: [PATCH] buffers: make a buffer inactive when switching away from it
When a buffer is not visible, it can still eats CPU due to animations
and Javascript still running. Gecko has some mechanisms to mitigate
this (don't play CSS animations, run timers less often, ...) to avoid
eating CPU for something not displayed but we must tell the browser
object if it is active or not.
---
modules/buffer.js |2 ++
1 file changed, 2 insertions(+)
diff --git a/modules/buffer.js b/modules/buffer.js
index 06f4a56..a362a8f 100644
--- a/modules/buffer.js
+++ b/modules/buffer.js
@@ -449,6 +449,7 @@ buffer_container.prototype = {
// Save focus state
old_value.saved_focused_frame = old_value.focused_frame;
old_value.saved_focused_element = old_value.focused_element;
+old_value.browser.docShell.isActive = false;
old_value.browser.setAttribute(type, content);
},
@@ -467,6 +468,7 @@ buffer_container.prototype = {
this.window.focus();
// Restore focus state
+buffer.browser.docShell.isActive = true;
buffer.browser.focus();
if (buffer.saved_focused_element)
set_focus_no_scroll(this.window, buffer.saved_focused_element);
--
1.7.10.4
That's very interesting. Will have time to investigate it next week.
Meantime, if anybody else wants to research this, provide citations and
arguments - is this the right thing to do? I'll be very interested.
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Installing on Ubuntu 12.04
On Tue, Aug 07, 2012 at 06:57:53PM -0400, Paul Burkander wrote: Hi, I'm really eager to try conkeror, but I'm having a heck of a time installing it on Ubuntu 12.04. I've installed xulrunner 15.0b3 from https://ftp.mozilla.org/pub/mozilla.org/xulrunner/releases/15.0b3/runtimes/, and it seems to work. If I type xulrunner from a command prompt it comes right up. I followed the guide here, http://ubuntulandforever.blogspot.com/2009/06/conkeror-is-keyboard-oriented-highly.html, and got conkeror from here: http://repo.or.cz/w/conkeror.git?a=snapshot;h=master;sf=tgz. When I try to run conkeror, it says xulrunner not found How can it be installed on Ubuntu 12.04? Thanks! The run-conkeror script in contrib is deprecated for XULRunner versions higher than 2.0. See the instructions on conkeror.org instead. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Installing on Ubuntu 12.04
On Tue, Aug 07, 2012 at 08:25:54PM -0400, John J. Foerch wrote: Paul Burkander p...@burkander.com wrote: Also, I checked about:config, and the variable conkeror.rcfile is set to /home/paul/.conkerorrc, which is where my file is. On Tue, Aug 7, 2012 at 7:29 PM, Paul Burkander p...@burkander.com wrote: great - thanks seems to have worked. However, it doesn't seem to read my .conkerorrc file. In that file, I have just two line: define_webjump(gm http://mail.google.com/;); homepage=www.msu.edu; But when I start up conkeror, it does not go to that home page, and the jump doesn't seem to work. Any ideas? Thanks There is a syntax error in your call to define_webjump. You need a comma after gm. And there is another syntax error on the line where you set 'homepage'. Extra quotation mark. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Installing on Ubuntu 12.04
If there is an error, it will be reported in the terminal. Do your testing by restarting conkeror, not with M-x reinit. There is a mozilla feature called the startup-cache that can royally screw with conkeror's js loading mechanism. Other troubleshooting tips here: http://conkeror.org/ConkerorRC#Debugging On Tue, Aug 07, 2012 at 08:36:08PM -0400, Paul Burkander wrote: Thanks. The extra quotation mark was to close the quoting of my .conkerorrc. Still, no matter what I write in it there's no indication whatsoever that it's been implemented. I've corrected the syntax error to no avail. On Tue, Aug 7, 2012 at 8:31 PM, John J. Foerch jjfoe...@earthlink.net wrote: On Tue, Aug 07, 2012 at 08:25:54PM -0400, John J. Foerch wrote: Paul Burkander p...@burkander.com wrote: Also, I checked about:config, and the variable conkeror.rcfile is set to /home/paul/.conkerorrc, which is where my file is. On Tue, Aug 7, 2012 at 7:29 PM, Paul Burkander p...@burkander.com wrote: great - thanks seems to have worked. However, it doesn't seem to read my .conkerorrc file. In that file, I have just two line: define_webjump(gm http://mail.google.com/;); homepage=www.msu.edu; But when I start up conkeror, it does not go to that home page, and the jump doesn't seem to work. Any ideas? Thanks There is a syntax error in your call to define_webjump. You need a comma after gm. And there is another syntax error on the line where you set 'homepage'. Extra quotation mark. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Installing on Ubuntu 12.04
Glad it's working now. yw
On Tue, Aug 07, 2012 at 09:28:13PM -0400, Paul Burkander wrote:
I had an error in it! Now that it's fixed it seems to be working!
Honestly though, I tried several different .conkerorrc's - I thought
they were error free. I guess I was making dumb mistakes.
Thanks for your help!
On Tue, Aug 7, 2012 at 9:15 PM, John J. Foerch jjfoe...@earthlink.net wrote:
I need to see the contents of your rc to know what is indicated by the
error message.
On Tue, Aug 07, 2012 at 09:09:59PM -0400, Paul Burkander wrote:
Thanks. I'll look at the debugging site. This is the error I got
from terminal:
SyntaxError: missing ; before statement
file:///usr/share/conkeror/components/application.js -
file:///home/paul/.conkerorrc:1
load1@file:///usr/share/conkeror/components/application.js:108
@file:///usr/share/conkeror/components/application.js:137
load_rc@chrome://conkeror/content/rc.js:60
handle_command_line@chrome://conkeror/content/command-line.js:179
@file:///usr/share/conkeror/components/command-line.js:23
Console error: [JavaScript Error: [Exception... Component returned
failure code: 0x80520012 (NS_ERROR_FILE_NOT_FOUND)
[nsIXPCComponents_Utils.import] nsresult: 0x80520012
(NS_ERROR_FILE_NOT_FOUND) location: JS frame ::
resource://gre/modules/XPCOMUtils.jsm :: XPCU_moduleLambda :: line
256 data: no] {file: resource://gre/modules/XPCOMUtils.jsm line:
256}]
Category: XPConnect JavaScript
Console error: [JavaScript Warning: function onKeyPress does not
always return a value {file:
chrome://global/content/bindings/autocomplete.xml line: 439 column:
12 source: return false;
}]
Category: xbl javascript
Console error: [JavaScript Warning: function onKeyPress does not
always return a value {file:
chrome://global/content/bindings/autocomplete.xml line: 503 column:
10 source: return true;
}]
Category: xbl javascript
Console error: [JavaScript Error: The character encoding of the HTML
document was not declared. The document will render with garbled text
in some browser configurations if the document contains characters
from outside the US-ASCII range. The character encoding of the page
must to be declared in the document or in the transfer protocol.
{file: chrome://conkeror-help/content/help.html line: 0}]
Category: HTML parser
Console error: [JavaScript Error: DownloadsCommon is not defined
{file: resource:///components/DownloadsStartup.js line: 134}]
Category: XPConnect JavaScript
On Tue, Aug 7, 2012 at 8:54 PM, John J. Foerch jjfoe...@earthlink.net
wrote:
If there is an error, it will be reported in the terminal.
Do your testing by restarting conkeror, not with M-x reinit. There is a
mozilla feature called the startup-cache that can royally screw with
conkeror's js loading mechanism.
Other troubleshooting tips here:
http://conkeror.org/ConkerorRC#Debugging
On Tue, Aug 07, 2012 at 08:36:08PM -0400, Paul Burkander wrote:
Thanks. The extra quotation mark was to close the quoting of my
.conkerorrc.
Still, no matter what I write in it there's no indication whatsoever
that it's been implemented. I've corrected the syntax error to no
avail.
On Tue, Aug 7, 2012 at 8:31 PM, John J. Foerch jjfoe...@earthlink.net
wrote:
On Tue, Aug 07, 2012 at 08:25:54PM -0400, John J. Foerch wrote:
Paul Burkander p...@burkander.com wrote:
Also, I checked about:config, and the variable conkeror.rcfile
is set
to /home/paul/.conkerorrc, which is where my file is.
On Tue, Aug 7, 2012 at 7:29 PM, Paul Burkander
p...@burkander.com wrote:
great - thanks seems to have worked. However, it doesn't
seem to read
my .conkerorrc file. In that file, I have just two line:
define_webjump(gm http://mail.google.com/;);
homepage=www.msu.edu;
But when I start up conkeror, it does not go to that home
page, and
the jump doesn't seem to work.
Any ideas?
Thanks
There is a syntax error in your call to define_webjump. You need a
comma
after gm.
And there is another syntax error on the line where you set
'homepage'.
Extra quotation mark.
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Webjump: Lookup in imdb
On Fri, Jul 27, 2012 at 08:47:49AM +0100, Johnny wrote: John J. Foerch jjfoe...@earthlink.net writes: On Fri, Jul 20, 2012 at 11:14:14PM +0100, Johnny wrote: hi all, I have just started to look into webjumps, and wonder if someone has a clever way to e.g. look up a movie title in imdb? Something like select a link or text, then directly search for it in imdb? Basically, I /borrowed/ the idea from conkerorwiki to use simply : define_webjump(imdb, http://imdb.com/find?q=%s;); It would be great to have the search string inserted by default! cheers! Maybe this: http://conkeror.org/Tips#SelectionSearches Thanks John, this is excellent! Only one problem, IMDB does not recognise the escape symbols (%20) and searches for the literally. Is there any way to avoid this? (Wikipedia interprets these as expected though) Regards You could use a function webjump instead. They probably expect spaces to be encoded as plusses. This is really an imdb problem; conkeror is doing the right thing. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Webjump: Lookup in imdb
On Fri, Jul 20, 2012 at 11:14:14PM +0100, Johnny wrote: hi all, I have just started to look into webjumps, and wonder if someone has a clever way to e.g. look up a movie title in imdb? Something like select a link or text, then directly search for it in imdb? Basically, I /borrowed/ the idea from conkerorwiki to use simply : define_webjump(imdb, http://imdb.com/find?q=%s;); It would be great to have the search string inserted by default! cheers! Maybe this: http://conkeror.org/Tips#SelectionSearches -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Display of hints
On Wed, May 30, 2012 at 08:13:02PM +0300, Nicholas Ochiel wrote: In a previous thread, a chop-mode for hints had been discussed. What I'm wondering is if it might be possible to narrow the list of displayed hints as input is received (the way the HoK plugin for Keysnail does it in firefox https://github.com/mooz/keysnail/blob/master/plugins/hok.ks.js) i.e. if the following hints are displayed: 100,200,300 then typing 2 causes 100 and 300 to disappear completely leaving only hints whose numbers begin with 2, in this case 200. When the page is crowded with hints, such a feature is particularly useful. P.S. If this had been discussed previously, do forgive me. Nice idea. Could maybe be added as an option. Suggestions for variable name, anybody? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Yanking range of nodes
On Fri, May 25, 2012 at 08:58:48PM +0400, Evgeniy Dolzhenko wrote: Hi guys, Wondering, is there any mode/hack which will let user to enter numeric range of nodes to yank text from? I.e. having the page like [1] Paragraph 1 [2] Paragraph 2 I can use y 1 - 2 and get both paragraphs into kill buffer. Not yet. I'm aiming to have this general capability for all of the browser object commands for 2.0. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] adblockplus.js fixes
On Mon, May 21, 2012 at 04:33:58PM -0400, Ivy Foster wrote: Hello, I've updated adblockplus.js. adblockplus_settings has been updated to work with the current adblock, hypothetically, using content/ui/settings.xul instead of a now-defunct function from modules/Utils.jsm. However, due to [a longstanding bug][1], it still doesn't work on Linux. Other platforms are untested. However, the preferences are still accessible via the extensions dialog. [1]: https://bugzilla.mozilla.org/show_bug.cgi?id=156333 Also, since said extensions dialog can't seem to open the Adblock filters dialog, I've added a function, adblockplus_filter_settings, and a corresponding interactive function, adblockplus-filter-settings. Enjoy! Ivy thank you -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Conkeror and XULRunner 14?
On Fri, May 18, 2012 at 01:33:30AM +0200, Axel Beckert wrote: Hi, I just tried to use Conkeror with XULRunner 14 from mozilla.debian.net's xulrunner-14.0 package (iceweasel-aurora repo, version 14.0~a2+20120511042006-1~bpo60+1). $ xulrunner-14.0 /usr/share/conkeror/application.ini Cannot find XULRunner runtime directory. Exiting. $ echo $? 1 $ Anyone tried Conkeror with XULRunner 14, too, preferably from some other source? Does the same error message appear? (Or asked the other way around: Is this a general issue or an issue with mozilla.debian.net's xulrunner-14.0 package?) Kind regards, Axel Maybe try with some simpler xulrunner apps, to narrow down exactly what changed. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] suppressing disk mode
On Sun, May 06, 2012 at 12:53:29AM -0700, Silvio Levy wrote: * Sometimes it's convenient to paste text into conkeror with middle mouse (since that's how I do it elsewhere). But middle mouse instead draws a little white disk with arrows, and puts conkeror in a scrolling mode that I no use for. How can I suppress this behavior and make middle mouse call paste-x-primary-selection? The preference that enables and disables the scrolling mode is general.autoScroll. I have never heard of this feature interfering with pasting into text fields. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Conkeror reloading everything
On Sun, Apr 29, 2012 at 11:28:15PM +0200, Thorsten Jolitz wrote: When I boot my machine and start conkeror, it loads everything I had open before, the windows and all its buffers. With stumpwm, one conkeror window is shown (probably the last one I had open yesterday), e.g. with the 4 buffers I had open, but in the background, other conkeror windows are opened too, with many buffers. Worst of all, the html5 videos on youtube I watched before all start to play automatically at conkeror startup and repeat themselves several times. This is a bit overkill for me. How can I change this behaviour, or at least, how can I kill all open buffers with one command? Because just shutting down conkeror with C-x C-c apparently is not enough to start with a clean conkeror next day, the buffers that were open when exiting are reloaded at next start-up. http://conkeror.org/Sessions -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Shell commands
On Fri, Apr 27, 2012 at 02:01:16PM +0200, Thorsten Jolitz wrote: Hi List, I have some problems with the shell commands in conkeror. When pressing x, then selecting a link to a pdf, and accepting the proposed default (in my case evince) everything is fine - the document is opened. But when I do the same thing with a link to a html website and the shellcommand wget, it seems that nothing happens - no error message, but no downloaded html file in ~/ either (wget works on the bash prompt). It should download to whatever your cwd is, but I recommend running the command in a terminal so that you can see its output. When I use X I'm queried for an url, but I'm not really sure what the semantics of this command are. Should the url be a shellscript? I tried several things, but always got error messages. It is prompting for an url to run the command on. To have it prompt for a link instead, use the sequence 'n X'. BTW: Is there something like the Emacs 'Messages' buffer in Conkeror, where one can look up the messages that popped up in the minibuffer? No, not yet. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Hard-coded viewers (was: Shell commands)
On Fri, Apr 27, 2012 at 04:00:28PM +0200, Axel Beckert wrote: Hi, my reply likely doesn't answer Thorsten's question, but reminds me that I noticed something else: On Fri, Apr 27, 2012 at 02:01:16PM +0200, Thorsten Jolitz wrote: When pressing x, then selecting a link to a pdf, and accepting the proposed default (in my case evince) everything is fine Not only in your case: It's hard-coded in modules/mime.js, even if evince is not installed at all. Definitely something I'll try to fix for the Debian package soon. (Would be cool if there would be a non-Debian-specific fix, too. :-) 'Hard-coded' would be the wrong word; it's just the default value of a configurable variable. Yes, it would be a good idea to customize defaults such as these in OS packages of Conkeror. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Remote external editing
On Fri, Apr 27, 2012 at 03:36:39PM +0200, Thorsten Jolitz wrote: Hi List, I guess it would be possible with Conkeror to attach to a remote gnu screen or tmux process that runs an editor, using the appropriate screen or tmux command with the remote url as shell command. Then (an exotic idea, I admit) wouldn't it be possible to leave authentication, session management and all the interactive stuff to the screen session and show only static html, if the terminal window could be integrated/attached to the Conkeror window in a way the user doesn't notice that there are two different application at work. For example, if the minibuffer could switch to the remote terminal session and the white space in the minibuffer is then actually a file on the remote server opened in the Zile editor, input areas and textfields in the html code could be replaced by input/output in the minibuffer. If that would be technically possible - would it be a good idea? Even when the remote Zile would be run as a underpriviledged user with only minimal rights, wouldn't it be a security issue to let a user attach to such a screen session? Maybe all this is only science-fiction, but I'm at least curios if it could be done. Hmm. Probably a pipe dream, but I suppose the way to do it would be through the plugin API. Sounds hard. Perhaps customizing your window manager would be a workable alternative; if the window manager reserved a certain space for your editor, it would have the same overall effect, and also would apply to all of your programs, not just your web browser. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Shell commands
On Fri, Apr 27, 2012 at 11:01:15AM -0400, John J. Foerch wrote: On Fri, Apr 27, 2012 at 02:01:16PM +0200, Thorsten Jolitz wrote: Hi List, I have some problems with the shell commands in conkeror. When pressing x, then selecting a link to a pdf, and accepting the proposed default (in my case evince) everything is fine - the document is opened. But when I do the same thing with a link to a html website and the shellcommand wget, it seems that nothing happens - no error message, but no downloaded html file in ~/ either (wget works on the bash prompt). It should download to whatever your cwd is, but I recommend running the command in a terminal so that you can see its output. I'm sorry, I misspoke. That command downloads the file to a temporary directory, not your cwd, and deletes it when the command is finished. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Shell commands
On Fri, Apr 27, 2012 at 05:53:39PM +0200, Thorsten Jolitz wrote: John J. Foerch jjfoe...@earthlink.net writes: But when I do the same thing with a link to a html website and the shellcommand wget, it seems that nothing happens - no error message, but no downloaded html file in ~/ either (wget works on the bash prompt). I'm sorry, I misspoke. That command downloads the file to a temporary directory, not your cwd, and deletes it when the command is finished. Hmm ... maybe a misunderstanding, but how do I use it if it is immediatley deleted? The temporary file is deleted after the program exits. This is designed for viewers; it's the wrong tool for using an alternative downloader. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Multiple Conkeror instances
On Sun, Apr 15, 2012 at 12:49:34AM +0200, Thorsten wrote: John J. Foerch jjfoe...@earthlink.net writes: On Fri, Apr 13, 2012 at 12:21:21PM +0200, Thorsten wrote: Hi List, I can open several independent 'conkerors' in stumpwm, but when I kill one of them, all of them got killed. Is that the expected behaviour? Are you sure they are separate instances, running with separate profiles, as opposed to separate windows in a single instance with a single profile? Not really, I can only say that I open a new xterm in stumpwm and type conkeror in it. Doing that twice, I have two conkerors runnung, that are apparently independent from each other (different buffers etc). But when I shut down one, the other gets killed too. Its not really a real issue, more a curiosity. You're opening multiple windows in a single instance of conkeror. In the current version of conkeor, each window has its own set of buffers. Here is some information on running multiple instances of conkeror at the same time: http://conkeror.org/Profiles -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Enable hints-minibuffer-annotation-mode by default.
On Fri, Apr 06, 2012 at 11:49:52AM +0930, David Kettler wrote: The consensus seems to be that this mode should be enabled by default; or at least there has been no dissent expressed. Great, please go ahead. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] [RFC] Open multiple links in hints or find-url interaction.
On Wed, Apr 04, 2012 at 09:47:35PM +0930, David Kettler wrote: This patch is not intended for application; the implementation is ugly and only lightly tested. But it does illustrate a facility I have found useful. There are two situations where I find myself wanting to open multiple new buffers in conkeror and where there is enough state in the current interaction to make it inconvenient to follow a single URL and then repeat the interaction. When using M-x follow-new-buffer-background (bound to a key) sometimes I have narrowed down to a few links using the hints interaction and want to load them all. Similarly, when using C-u g sometimes when I have narrowed down to a few links from my browser history using the minibuffer completion, I want to load them all. In both cases this patch allows the currently selected URL to be followed with C-RET, returning to the same state of the minibuffer interaction, ready to load another. A proper implementation of this idea will need a larger restructure of the minibuffer interaction and a better understanding of the coroutine interface than I currently possess. See also issue365: Open multiple links in conkeror My plan is to extend the browser-object system with browser-object-lists, and allow components like hinting to return a list of nodes instead of just a single one. Hinting would be extended with facilities that allow fine-grained control at narrowing down on the list of nodes to return. Planning it for 2.0. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] German Umlaute are gone in Textfields
On Wed, Apr 04, 2012 at 07:11:39PM +0200, Thorsten wrote: Hi List, I just cloned the excellent .conkerorrc from John Foerch, but now all the German Umlaute and other chars on my German keyboard are gone when I try e.g. to write a comment in Disqus or fill out any other textform. It worked out well before, but I can't find the culprit in the many .js files involved. I changed font.js to ,--- | session_pref(browser.display.use_document_fonts, 0); `--- but that did not help. Any hints about where to look would be welcome. PS version 1.0opre, xulrunner 11.0, linux Maybe input.js. Let me know. Note, I make no guarantees about the working state of the version of my rc that I keep online. I put a lot of code in it just for the sake of examples for others, assuming people will copy paste the parts they want, but some of it is commented out, or modified in the copy that I actually run. input.js is such an example. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] German Umlaute are gone in Textfields
On Wed, Apr 04, 2012 at 08:24:30PM +0200, Thorsten wrote: Now, C-k (cut-to-end-of-line) is not working any more, although C-h k shows it has the right binding. It's bound to backward-line in my keys.js. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Facebook renders as a mostly blank page on Conkeror
On Wed, Apr 04, 2012 at 12:15:42PM -0700, Sean McAfee wrote: A day or so ago I found that most of Facebook's pages were being rendered as a mostly blank page on Conkeror. Example: http://seanmcafee.name/conkeror-facebook.png The site still renders properly on Firefox. I just did a git-pull of the Conkeror source, and it made no difference. Anyone have any ideas how this could be happening? Do you get the same results if you set your user-agent string to pretend to be firefox? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] ReferenceError: alert is not defined on M-:
On Sat, Mar 24, 2012 at 10:43:03AM +0100, Carlos López Camey wrote: Hello, Doing M-: RET alert(msg); results in an error message ReferenceError: alert is not defined is this expected? I was trying to set up MozRepl as explained on the wiki, the server starts but trying to alert something results in the same error. I'm attaching my .conkerorrc Thanks. M-: evaluates in Conkeror's application scope, not window scope or content scope. Contrast with the following: get_recent_conkeror_window().alert(hello world); -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] ReferenceError: alert is not defined on M-:
On Sat, Mar 24, 2012 at 03:40:18PM +0100, Carlos López Camey wrote: M-: evaluates in Conkeror's application scope, not window scope or content scope. Contrast with the following: get_recent_conkeror_window().alert(hello world); Thanks for your answer, it works now. Is there any _generated_ documentation for all the functions on this scope? No, just the wiki and comments in the code. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] hints-minibuffer-annotation-mode
Hi all, We still have the question before us whether or not hints-minibuffer-annotation-mode should be turned on by default. From my general survey here and in the irc channel, most people seemed favorable to the idea, and I think it would be a good idea too. Shall we proceed? I would also like to see, as a default or an option, for urls in the annotation to be shown as relative urls when possible. That would make the annotation easier to read quickly. David or anybody, willing to write a patch? Thank you. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Navigating messy Web pages
On Fri, Mar 16, 2012 at 10:53:06AM -0700, Vinh Nguyen wrote: On Fri, Mar 16, 2012 at 9:27 AM, John J. Foerch jjfoe...@earthlink.net wrote: This is not a frame, but a scrollable div. To focus a frame, use the ';' key. To focus a scrollabe div, the trick I usually use is to isearch into the div, and then it will be focused. Any suggestions for gmail? I can never navigate with the keyboard with their new user interface. ; doesn't focus because it appears to be only 1 frame. Your isearch method doesn't work either. Thanks. Another trick is to focus a link inside of a scrollable div or frame. The key sequence is 'n ;' -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Documentation for global modes.
On Wed, Mar 14, 2012 at 08:04:07PM +1030, David Kettler wrote: Merged patch with some changes. Thanks. Relatedly, the change to keymap.js suppresses a warning on describe-command with modern XULRunner. This change is not related, and should be in a separate patch. So it should; I only noticed the problem once I had added the doc changes, but it is independent of those changes. It looks like the bindings that cause the warning (because they have no command member) come from prefix keys in a multiple key binding. I would have done so, except that it is not specific enough to say modern XULRunner. Of the versions I have available, 1.9.1 does not complain but 9.0, 10.0 and 11.0 do. But I don't know when that warning started to be issued, Debian doesn't offer me other versions to try and I don't think its worth installing older versions to answer this question. Would wording like the following suffice? Suppress a warning in describe-command XULRunner 9.0 (and perhaps earlier) complains when an attempt is made to access an undefined property of an object. Binding objects do not always have a command property so, check that it is defined before accessing it. Regards, David I'm using 8.0 and I don't get the warning, so it must have started in 9.0. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] RFC: Remove minibuffer annotation during minibuffer recursion
On Wed, Mar 14, 2012 at 06:25:20PM +1030, David Kettler wrote: When the minibuffer is called recursively, the minibuffer annotation must be hidden or updated appropriately. For instance, previously if C-h c was pressed while the annotation was shown during a hints interaction, the annotation would confusingly remain visible while the minibuffer prompted for a key. Now the annotation will be hidden until the user returns to the hints interaction. --- Sorry that I didn't notice this earlier. It may be better to manage this in minibuffer_input_state so that it works transparently for any future uses of the annotation. However, I haven't got that to work and it also felt like crossing an abstraction boundary. Okay, go ahead and push this. A separate problem I noticed while looking into this, but is independent of the annotation. While in a hints interaction press C-h f C-g; you are returned to the hints interaction, but the minibuffer displays Quit and does not return to the hints prompt. --- I might have fixed that in my work on 2.0, but if you want to track it, add an issue on bugs.conkeror.org. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH 0/2] enable annotation with modes
On Sat, Feb 25, 2012 at 05:35:03PM +1030, David Kettler wrote: This pair of commits implements John's suggestion to enable this feature using modes. It's a nice approach. Further discussion on naming of the functions might be better in the review thread. I like the names you've chosen for the modes. They make it obvious what their purpose is. The minibuffer-annotation system is of course a general system by design, so let's say down the road we add another feature that uses it - for example, ido-style inline completions. This poses a problem, because if this hypothetical feature were enabled, and a person then disabled hints-minibuffer-annotation-mode, minibuffer-annotation-mode should *not* be disabled, because it is still needed for the inline completions system. Any thoughts on how to deal with this eventuality? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Help remedy my JS ignorance
On Fri, Feb 24, 2012 at 07:51:21AM +, Philip Hudson wrote:
// FIXME This sets up unbounded recursion; need another way to store and
// reference the original function
// var old_external_editor_make_base_filename =
//(old_external_editor_make_base_filename ||
// external_editor_make_base_filename);
// function external_editor_make_base_filename (elem, top_doc) {
//return conkeror-edit-+old_external_editor_make_base_filename(elem,
top_doc);
// }
I see no problem with this. It should be safe for repeated evaluation, as
given.
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Remove Info.plist
On Fri, Feb 24, 2012 at 09:22:17PM +0100, Konrad Hinsen wrote: Philip Hudson writes: I read Konrad as saying his patch made our Info.plist look right for what we all imagined was its purpose, namely to allow xulrunner to generate an Info.plist to deploy, and that with the patch our I don't install through XULRunner at all, since that doesn't work for me. The Info.plist I sent is the one I use in my Conkeror application bundle. Konrad. Please explain your installation method. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Remove Info.plist
On Sat, Feb 25, 2012 at 08:26:33AM +0100, Konrad Hinsen wrote: John J. Foerch writes: I don't install through XULRunner at all, since that doesn't work for me. The Info.plist I sent is the one I use in my Conkeror application bundle. Please explain your installation method. It's on the Wiki: http://conkeror.org/InstallationOSX Look for Installation based on Firefox. Basically, I take most of an existing Firefox installation, but replace a few files and copy the Conkeror source code into the application bundle. Konrad. ah, thank you -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCHv3 0/2] url panel replacement
On Tue, Feb 21, 2012 at 06:47:59PM +1030, David Kettler wrote: I'm thinking it would actually be two modes, because there has to be one mode to provide the general mechanism of minibuffer annotation, and a second one specific to hinting. The general one would enable the xbl binding, and the hinting-specific one would would do what hints_information does now. The hinting-specific one would automatically enable the general one as needed. Sound right? I don't quite understand. There should certainly be a hints-specific mode defined by define_global_mode. There should also be an underlying mode of some kind that controls the annotation; the hints thing is the only current user, but others may arise. But I think the general annotation mode shouldn't be defined by define_global_mode; shouldn't that be reserved for user facing features? Regards, David. Yes. Sorry for the misunderstanding; I was using the word 'mode' in a general sense - something that behaves like a mode - it need not literally be a global-mode. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] inline-url-panel branch review
On Tue, Feb 21, 2012 at 06:29:52PM +1030, David Kettler wrote: Great, can we move the annotation-system's css into the default theme? Okay. Shall I push commits onto the branch, ready to be squashed before being rebased onto master? Regards, David Yes, please. Still give a normal commit message though.. I won't necessarily squash when merging. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] inline-url-panel branch review
On Mon, Feb 20, 2012 at 10:19:26PM +1030, David Kettler wrote:
Hi John,
I can't resist responding to some of these points.
Good! :D
Do all of the components of this feature have the best names for what
they do? The user variable 'hints_information' is the one I have the
most doubt about.
I wanted to keep it generic, since the URL panel name is no longer
accurate, but hints_information is probably generic to the point of
meaninglessness. Perhaps something like link properties mode. But the
primary purpose is still to show the URL, so perhaps a name based on
that is still fine; hints-url-mode.
I was thinking about whether the word annotation could also be used in
this context, but I'm unsure.
Can the annotation be styled by Conkeror's theme system?
Is the position of the annotation good, or too close to the minibuffer
input? Is this customizable through the theme system?
Putting these rules in a theme works fine:
#minibuffer-input [anonid=annotation] {
color: blue;
}
#minibuffer-input .textbox-input-box {
width: 10em;
}
#minibuffer-input [anonid=strut] {
padding-right: 2em !important;
}
Great, can we move the annotation-system's css into the default theme?
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Mac default-browser and 'open' support
On Mon, Feb 20, 2012 at 10:14:34AM +0100, Konrad Hinsen wrote: John J. Foerch writes: Offer accepted. Send the file in any way that is convenient, as an attachment on this list, or even through a pastebin on irc. Here it comes. I have deleted CFBundleSignature, after verifying that is indeed not required. In fact, it matters only for applications that create documents in a specific format. Conkeror only opens documents, and only standard formats, so there is no point in having a bundle signature. If it had one, it would in principle have to be registered with Apple in order to be unique. With the attached Info.plist, Conkeror can be made the system default browser and opens correctly Web URLs (http, https) and local file URLs (file). Konrad. Thank you. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Mac default-browser and 'open' support
On Mon, Feb 20, 2012 at 11:56:38AM +0100, Konrad Hinsen wrote: Given the recent exchanges, I should add that this Info.plist works if used directly in an application bundle, after replacing $CONKEROR_SHORT_VERSION$ by the version number. I assumed that that's what XULRunner does when installing, but apparently that is not true. That $CONKEROR_SHORT_VERSION$ thing was a leftover from our versioning system of several years ago. I've removed it now, and put 1.0. I'm not really sure what this key is for, and whether the value I put for it is right, or how it is different than the regular version, which is 1.0pre. In the docs on these properties, the format specified for version numbers doesn't have any mention of development phases like pre, a, and b, and they talk about localizing version numbers... wat?? John J. Foerch writes: What I suggest is the following: you (or any OS X user) write a minimal test case XULRunner application to investigate this issue. (It's not that hard.) If there turns out to be bugs in Mozilla, bring it to their attention. Keep us informed. There is a simple example application available as an attachment to the XULRunner tutorial at: https://developer.mozilla.org/en/getting_started_with_xulrunner There's also a set of instructions for packaging that example app for distribution for various platforms: http://joliclic.free.fr/blog/index.php?post/2011/05/20/Distribute-your-XULRunner-app-1-Preamble Unfortunately I can't help with testing this, because I can't get XULRunner to work. Did you by any chance try running that example app on your machine, and it failed? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCHv3 0/2] url panel replacement
On Mon, Feb 20, 2012 at 10:01:45PM +1030, David Kettler wrote: Even if it were on by default and we used this method of disabling it, it would mean less of a footprint on the GUI for people who didn't use it, and that could be a good thing. If I am thinking this through correctly, it would mean that instead of a user variable 'hints_information', we would have a global mode that enabled the binding. Modes are stylistically nice. Better to use a stylesheet, alone or in a theme, than to use the element.style method on every minibuffer-input, I think, because it's a simple thing to register/unregister a stylesheet. Agreed. I particularly like the idea of using a global mode; it's analogous to overlink_mode for mouse selection. Regards, David I'm thinking it would actually be two modes, because there has to be one mode to provide the general mechanism of minibuffer annotation, and a second one specific to hinting. The general one would enable the xbl binding, and the hinting-specific one would would do what hints_information does now. The hinting-specific one would automatically enable the general one as needed. Sound right? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Mac default-browser and 'open' support
On Mon, Feb 20, 2012 at 04:37:30PM +, Phil Hudson wrote: On 20 Feb, 2012, at 10:56 am, Konrad Hinsen wrote: Which version of XULRunner do you use, and on which version of MacOS? % /Library/Frameworks/XUL.framework/xulrunner-bin --version Mozilla XULRunner 1.9.2 - 20100222070821 % uname -a Darwin MBP 9.8.0 Darwin Kernel Version 9.8.0: Wed Jul 15 16:55:01 PDT 2009; root:xnu-1228.15.4~1/RELEASE_I386 i386 which means OS X 10.5.8 I've found that calling MacPorts' version of xulrunner (/opt/local/bin/xulrunner) does not work for '--install app' on this setup. This sounds like something that needs to be isolated and documented. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Mac default-browser and 'open' support
On Sun, Feb 19, 2012 at 09:55:28AM +0100, Konrad Hinsen wrote: John J. Foerch writes: file, ahead of the basic identification of the program. Now, it's pretty clear to me which part of this patch provides the feature in question, but as I have no OS X machine available for testing, I must leave it to someone else to finish the patch. I can provide a correct and complete Info.plist, but I don't know how to submit patches. Konrad. Offer accepted. Send the file in any way that is convenient, as an attachment on this list, or even through a pastebin on irc. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Mac default-browser and 'open' support
On Sun, Feb 19, 2012 at 03:41:57PM +, Philip Hudson wrote: On 19 Feb, 2012, at 3:23 pm, John J. Foerch wrote: If 'xulrunner-bin --install-app' ignores our Info.plist and writes its own, what good does patching our Info.plist do at all? None that I know; that's why I didn't include that patch to start with. Indeed, I don't know quite what relation if any our Info.plist, patched or unpatched, has with the generated Info.plist; it *seems* to be neither ignored completely, nor faithfully followed. I think this will require some investigation of xulrunner's code to resolve. I conjecture that running conkeror from the checkout directory in the OS-independent, standard way on OS X (using the distributed shell script, rather than using the generated conkeror.app) *might* benefit from a patched git Info.plist, but I have no data on that yet. That's why I say it may be necessary (but is definitely not sufficient) to patch git Info.plist. Hope that's clear. My Mac is at your disposal for further investigation. Two tasks arising from this, then: - Find out wtf xulrunner is doing with Info.plist files when it runs --install-app - Find out whether patching our Info.plist makes any difference to remoting with 'open' after launching Conkeror on OS X using the standard shell script from a git checkout (rather than the app that xulrunner generates in /Applications). I'll get around to these slowly; if anyone else is keen, please don't wait for me. Are you referring to contrib/run-conkeror when you say the standard script? If so, it is standard no longer. The mechanism it uses to locate XULRunner was removed from XULRunner in version 5, so it is deprecated. What I suggest is the following: you (or any OS X user) write a minimal test case XULRunner application to investigate this issue. (It's not that hard.) If there turns out to be bugs in Mozilla, bring it to their attention. Keep us informed. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] duplicate buffer with history
On Sat, Feb 18, 2012 at 12:10:13PM -0500, Rodrigo Amestica wrote: Hi, from conkerorwiki I know that to duplicate a buffer I should use follow-new-buffer. But the duplicated buffer does not duplicates history and, therefore, I cannot navigate back in it. Is there a way for doing so? thanks, Rodrigo Yes, it can be done. You get the sessionHistory object to be copied like this: buffer.web_navigation.sessionHistory; Iterate it with its SHistoryEnumerator to collect the urls in the history, then make a new buffer and call the buffer's 'load' method for each url. https://developer.mozilla.org/en/XPCOM_Interface_Reference/nsISHistory -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Java plugin
On Sat, Feb 18, 2012 at 04:04:46AM -0800, Silvio Levy wrote: Dear Conkerors, On the page http://onlinestatbook.com/stat_sim/normal_approx/index.html I get Your browser either does not support JAVA or has JAVA disabled. But the plugin is there, and is seen and reported enabled by conkeror. See output of about:plugins below. (It doens't seem to be a matter of versions, because the plugin works for the same site under firefox and chrome - although on firefox the path reported in about:plugins is /usr/lib/mozilla/plugins/libnpjp2.so - which is a symlink to the path reported by conkeror below.) Any ideas? Silvio I don't use Java myself, but somebody else tried the site with icedtea7 and said it worked. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCHv3 0/2] url panel replacement
On Sat, Feb 18, 2012 at 07:31:51PM +1030, David Kettler wrote: Thanks for your comments, John. I'm a little hesitant to suggest this, because I think this feature should always be on; browsers should always make the URL visible when following a link. That said, a way to reduce the risk of introducing this feature is to enable the binding dynamically (see the reference below). That way (if it defaults off) there is no change to the minibuffer unless it's enabled. https://developer.mozilla.org/en/XBL/XBL_1.0_Reference/Binding_Attachment_and_Detachment#Attachment_using_element.style_property Regards, David Even if it were on by default and we used this method of disabling it, it would mean less of a footprint on the GUI for people who didn't use it, and that could be a good thing. If I am thinking this through correctly, it would mean that instead of a user variable 'hints_information', we would have a global mode that enabled the binding. Modes are stylistically nice. Better to use a stylesheet, alone or in a theme, than to use the element.style method on every minibuffer-input, I think, because it's a simple thing to register/unregister a stylesheet. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCH] Mac default-browser and 'open' support
On Sat, Feb 18, 2012 at 02:06:21PM -0500, John J. Foerch wrote: On Sat, Feb 18, 2012 at 05:12:02PM +, Philip Hudson wrote: Inserts the necessary plist XML tags to enable setting Conkeror as the Mac OS X default-browser and for using the darwin command-line 'open' command for URLs (without the '-a conkeror'). The code is simply copied from the wiki, not mine. Apologies if this is already in a queue somewhere. This patch does not apply cleanly, and further, does not target the named feature specifically enough, evidence the duplication of key/value pairs that already exist in Info.plist, unfilled information (CFBundleURLName), and the apparent random placement of the new hunk near the top of the file, ahead of the basic identification of the program. Now, it's pretty clear to me which part of this patch provides the feature in question, but as I have no OS X machine available for testing, I must leave it to someone else to finish the patch. There is a comment in our current Info.plist next to CFBundleSignature where I noted that the value is supposed to be a 4 letter code to represent the program, but I never understood why this was required, or whether it had to be unique in some special namespace.. if our OS X developers could investigate this point, I would appreciate it, and I would also prefer the code CNKR instead of CONK (or conk as is there now). This is because I have made a point over the last few years of never abbreviating the name of the program, but if we are forced to for a specific context, my preference is CNKR. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] inline-url-panel branch review
Hi all, As you likely know, there is a new branch in our git repository called inline-url-panel, which contains a feature by David Kettler that replaces the hints-url-panel with a nicer looking and nicer working inline display in the minibuffer area. Since this is a fairly substantial change to Conkeror's gui and may have long term design effects, I am hoping that anybody able to test the branch before merge will do so, and participate in review. Now I'll get the discussion started with some questions. Do all of the components of this feature have the best names for what they do? The user variable 'hints_information' is the one I have the most doubt about. Can the annotation be styled by Conkeror's theme system? Is the position of the annotation good, or too close to the minibuffer input? Is this customizable through the theme system? Does the way the annotation changes position as you use the hint system feel natural? For instance, when you type a number, the prompt changes, shifting both input and annotation to the right, but when you type text, the annotation does not move until the input string has exceded a certain width. Does this design feel natural or does it seem like the annotation should move to the right as you type text? Should the hint annotation be enabled by default? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] client redirects
On Sat, Feb 18, 2012 at 01:45:52AM +1030, David Kettler wrote:
What if the predicate was a keyword argument; then the caller has the
option to place the parameters in order:
define_client_redirect(foo, $predicate=/foo/, function(){})
Keywords come after positional arguments.
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] client redirects
On Thu, Feb 16, 2012 at 10:13:59AM +0100, Tom Willemsen wrote: Hey guys, I hope you don't mind me sharing my question. I wouldn't have posted if I wasn't looking for critique.. Wouldn't it be clearer/handier/better/idunno if the first argument is always a regexp and the second is either a string which can (I assume) use the first regex's groups (like `string.replace(/re(g)ex/string with \1roup/)' or something), or a function which gets each matched group as a parameter (an array probably) which it can check? Constraining the matcher to regexp-only is too limiting. The uri object passed to the transform function is already parsed into its components, so specific parts of an url can be easily matched against without having to deal with the complexities inherent in the limited language of regexps. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] breaking change: buffer_loaded_hook
Hi all, buffer_loaded_hook now only fires for the top-level document load, not the loads of any subframes. This was the original intent of the hook, and all existing uses of the hook that I know of assume this behavior. If anybody was using this hook to watch subframe loads, let me know so we can decide on the best way to support that. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] client redirects
Hi all,
There is a technique that can be employed to have Conkeror do an automatic
redirect when you visit a given url pattern. I use this with Google
Images to skip their intermediary page and jump straight to the image, as
described here:
http://conkeror.org/Tips#ClientRedirect
I had been meaning to abstract this technique into a general-purpose
module, and finally got around around to it. Attached is the first
implementation. By this implementation, configuring a client redirect is
now much more concise:
require(client-redirect);
define_client_redirect(google-images,
function (uri) {
return /(images|www)\.google\.com$/.test(uri.host)
uri.filePath == /imgres
regexp_exec(/imgurl=([^]+)/, uri.query, 1);
});
The general syntax is: define_client_redirect(name, from, to);
The argument 'from' may be either a regexp or a function. If it is a
regexp, a 'to' procedure must be given to transform the result of the
match into a new url. If it is a function, the function should return the
new url itself.
What do people think of this design? Seem good, or is there anything I
missed?
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] Conkeror without XULRunner?
On Tue, Feb 14, 2012 at 10:54:11AM +0100, Konrad Hinsen wrote: John J. Foerch writes: Can you document your findings on our wiki? If you haven't been approved for write access yet, just let me know your login name. Some relevant pages: http://conkeror.org/CCOSX http://conkeror.org/InstallationOSX http://conkeror.org/UpstreamBugs#Command-lineremotingdoesnotworkonOSX http://conkeror.org/DefaultBrowser I have added a description of my Firefox-based installation to the InstallationOSX page. Feedback from anyone trying this is appreciated. There's still a bug in menu handling, which will have to be fixed in the Conkeror source code. I'll look into this when I find some time. What exactly is the Conkeror Committee on OS X? Is there a contact? We get a lot of questions about OS X in the irc channel, but we have no regular OS X users to answer the questions. I hoped it might spur participation to declare the existence of CCOSX. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] possibly_valid_url change
On Tue, Feb 14, 2012 at 07:40:02PM +1030, David Kettler wrote: - Proposed behavior: inner whitespace before first / means NOT AN URL Looks good to me. This retains the current behaviour that an unadorned word is url-like. I defined it this way in the first place because I sometimes visit the default url on local machines by just giving the machine name. But I probably more frequently want to search for a single word, so I get annoyed that I have to specifically put in the search engine webjump. The wiki offers a version of possibly_valid_url that requires a url to have certain punctuation; I wouldn't object to a further change to the default along those lines. Regards, David Since default-webjump is an optional feature, it makes most sense for the default possibly_valid_url to treat barewords as url-like. People who use default-webjump can add the alternative version of possibly_valid_url to their rc, which is more convenient for them. That alternative version will be updated to follow suit with the change proposed here, if the change is adopted. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] zoom stuff
Hi all, I've pushed two new features related to zooming that raise certain questions about defaults. The first feature is the command image-zoom-to-fit, bound to 'C-x %'. It toggles zoom-to-fit mode of an image document, when the image is larger than the viewport. Previously the only way to use this mozilla feature was to click on the image. Somebody told me that most of the major browsers use zoom-to-fit on large images by default. Now that we have a way to toggle the mode with the keyboard, the choice is before us whether Conkeror should have zoom-to-fit as a default. (The pref browser.enable_automatic_image_resizing controls this.) The second feature is zoom_widget, a new mode-line widget that displays the zoom status of the current buffer. For practical purposes, image-zoom-to-fit is considered a form of zooming, and is reported by this widget. The widget itself may still need some tweaking, as making sure mode-line widgets are updated at the appropriate times is always a tricky thing. However, having this mode-line widget raises the question of whether it should be enabled by default. Thoughts? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
[Conkeror] 0 as zoom-reset
Hi all, I noticed the other day that Emacs uses '0' as zoom-reset. This would be a useful convention to follow in Conkeror, because currently, Conkeror uses '=' as zoom-reset, and '+' as zoom-in, meaning that on most keyboards, one must hold shift to zoom-in. If we adopted the Emacs convention of using '0' as zoom-reset, we could make '=' another zoom-in key. A related issue is that for long, I have wanted to bind the unmodified numeric keys to be numeric prefix keys. The fact that '-' is a zoom key has blocked this, and if '0' were also a zoom key, it complicates it further. Yet now I think we can do both, by adopting the following convention: '0' and '-' would be zoom keys. '1' through '9' would be numeric prefix keys, that, when entered, would install an overlay keymap that contains bindings for '0' and '-' to work as numeric prefix keys. The tradeoffs are as follows: - to give a negative numeric prefix, you would either need to use 'C-u' as now, or type the number like: 1 - - the zoom bindings would not be callable with a numeric prefix. none of the zoom commands currently accept a numeric prefix, but if they were ever modified to do so, a workaround would need to be designed. Also, how do people feel about the current bindings where unmodified '+' and '-' mean text-zoom, and 'C-+' and 'C--' mean full-zoom? Is it good the way it is, or should it be the other way around? Or what would people feel about going all-out, and adopting the Emacs way of zooming in its entirety, meaning that to use zoom keys, you first give the C-x prefix a single time? -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] page-modes and browser objects
On Mon, Feb 13, 2012 at 10:35:38PM +1030, David Kettler wrote:
google-search-results-mode and other page-modes set the default browser
object for various commands to make it convenient to select the most
useful links. However, they do not restore the default when the page
mode is disabled. A sufficient fix is to include a line like the
following in the page mode's disable method:
buffer.default_browser_object_classes = {};
But page modes now operate more like emacs minor modes. If in the
future we start having multiple page modes activated, each with their
own idea of the default browser objects, then the fix above (and the
current representation of the defaults) will not be sufficient. But it
may not be worthwhile accomodating that possibility; it's not clear that
it has any utility.
In any case, I suggest that for now we install fixes as above.
Affected page modes are:
% grep -l default_browser_object_classes *
google-images.js
google-search-results.js
reddit.js
% grep -l media_setup_local_object_classes *
dailymotion.js
google-video.js
youtube.js
Regards, David
P.S. It looks like media_scrapers also don't get cleaned up properly.
Thanks, fixed good enough for now. Buffer-defaults for browser objects
was not a well designed feature, and I'll be returning to it later.
Media-scrapers is not currently part of the page-mode system.
--
John Foerch
___
Conkeror mailing list
Conkeror@mozdev.org
https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCHv2 0/2] url panel replacement
On Mon, Feb 13, 2012 at 10:09:56PM +1030, David Kettler wrote: Hi John, * Approach #2 - Instead of having a separate xul element for the info label, add the info text directly to the minibuffer input. ... The other thing that looks awkward is dealing with the initial fixed width of the input part of the textbox, before it starts expanding. Inserting spaces between the input and info portions would achieve this, but the width will vary according to the widths of the other characters, so a specific pixel width can't be set, leading to the info position wobbling. Perhaps there's some unicode space character for which we could set the width. What do you mean about the widths of spaces varying depending on the widths of other characters? Surely not.. No; I'm referring to the width of the input portion of the textbox, not that of the individual characters. (Probably superfluous detail follows.) We want to provide some initial space in the input field, say 5ems, to display the input text, with the info displayed to the right (after some separation, say 1em). This initial width is needed to make it clear that this is an input text box; otherwise the info text appearing at the left of the box is confusing. It also looks better; in the usual short input case the info position is fixed. Only once this initial space is full does the input area start growing and the info moves to the right. To accomplish this with a single textbox implies inserting space characters between the input and info parts of the text. Because the individual characters have different fixed widths and the various space characters have fixed widths, there is no general solution to the problem of how many spaces to insert to give a certain pixel width. One can get within a half space width of the desired width, but the exact position will vary according to the characters in the input; i.e. to the left of the inserted spaces. By this method, the info text would always appear at a fixed offset to the right of the input text. Sorry, thought that was obvious. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCHv2 0/2] url panel replacement
On Mon, Feb 13, 2012 at 10:09:56PM +1030, David Kettler wrote: What do you think about the XBL idea? I like it a lot. I read your might be favourable comment as implying that this was merely sweeping the problem under the rug. But I agree with your reframing of the problem below, where a proper abstraction manages the issue. It probably won't be possible to entirely forget the complexity in future development, but this should contain it. All approaches have complexity, but it's a matter of where that complexity is that makes the difference; whether it is local complexity to a single module (hints.js), leaky complexity that infects other modules (minibuffer.js), or complexity that can be wrapped up in a tidy abstraction and forgotten. I think we might be able to use XBL to wrap up the entire stackstrut method in a self-contained file, introducing a new XUL element minibuffer-input/ to be used in place of the textbox. This element would handle the stackstrut mechanism internally and transparently. It would probably just need to provide a method to switch it between normal mode and stackstrut mode. Sounds good. How do you feel about the name annotated-textbox for the XBL binding; with minibuffer-input as the element that binds to it (if I've got my terms straight)? Sounds fine. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] smooth scroll add-on
On Mon, Feb 13, 2012 at 06:45:13PM -0500, Rodrigo Amestica wrote: Hi, I installed this add-on and enabled the keyboard (Up/Down) through its preferences interface. However the scrolling does not work smoothly, it still looks the normal 'stepped' motion. Is it possible the conkeror functions bound to Up/Down are somehow overriding the add-on functionality? Anybody out there having any experience with this add-on? thanks, Rodrigo ps: as per a previous posting I read today I'm skipping the custom xulrunner for firefox directly like this: /usr/lib/firefox-10.0/firefox -app ~/local/conkeror/application.ini plus conkeror from git (repo.or.cz). conkeror-version reports: Conkeror 1.0pre (XULRunner 1.9.2.26, Linux) All this under ubuntu 10.04. Try setting the pref general.smoothScroll to true. There should not be any need for an extension to do this, as smooth scrolling is an internal feature of gecko. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] smooth scroll add-on
On Mon, Feb 13, 2012 at 08:53:22PM -0500, Rodrigo Amestica wrote: On Mon, 13 Feb 2012 20:33:09 -0500, John J. Foerch jjfoe...@earthlink.net wrote: Try setting the pref general.smoothScroll to true. There should not be any need for an extension to do this, as smooth scrolling is an internal feature of gecko. not as smooth as the add-on but an improvement. Many thanks John. Give the add-on a try in ff to fully understand the difference. thanks, Rodrigo Then the general answer to your question is yes, there would be a conflict. Any extension that sets up key event handlers will not be compatible with Conkeror's input system. Those even handlers would have to be written as Conkeror commands and Conkeror bindings to work. -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
Re: [Conkeror] [PATCHv2 0/2] url panel replacement
On Sun, Feb 12, 2012 at 07:28:01PM +1030, David Kettler wrote:
Hi John,
Thanks for your detailed responses. I see this as an important feature
for conkeror; I think that is why I'm willing to put up with more
complexity in its implementation than it would need if the platform was
more reasonable. It looks like any implementation is more complex than
seems reasonable.
As an aside, I got the idea for the deckstrut method from
http://www.blackfishsoftware.com/node/47
Another aside; I wonder what reference materials you use for XUL, etc.
A couple of your suggestions can't be derived from the mozilla
documentation.
* Approach #2
- Instead of having a separate xul element for the info label, add the
info text directly to the minibuffer input.
...
Depending on how hacky it turns out to be to keep the the insertion point
out of the info area and to keep the box scrolled appropriately, this
could actually be a very elegant solution to the problem.
The other thing that looks awkward is dealing with the initial fixed
width of the input part of the textbox, before it starts expanding.
Inserting spaces between the input and info portions would achieve this,
but the width will vary according to the widths of the other characters,
so a specific pixel width can't be set, leading to the info position
wobbling. Perhaps there's some unicode space character for which we
could set the width.
What do you mean about the widths of spaces varying depending on the
widths of other characters? Surely not..
I wonder where you learnt about SELECTION_URLSECONDARY; I can't find any
documentation for it. There are other selection types though, which may
also be usable.
I was thinking about this problem in the most basic terms.. all we really
want to do is have some auxiliary text that appears to be in the
minibuffer input, but differentiated by style. I remembered that Firefox
has two different styles of text in its url bar to set off the domain from
the rest of the url, so I went looking in the source to see how they did
it. It doesn't seem to be documented, and maybe that means it only exists
in Firefox, not sure.
BTW; this feature is very usable with just the first patch. For short
inputs the expansion is not required and surely a long input is unusual.
It is still works, although the input scrolls rather than grows.
I don't consider this feature viable unless we can get the
auto-positioning of the info working. Without that, the UI just feels
incomplete and amateurish. It looks broken when input text is long and
there is nothing in the info field.
Okay. It's straightforward to hide the info display when it's empty,
which would avoid that specific problem. But it's still not great.
...
I think of this as the stackstrut method --- let's use the word strut
instead of shadow. The complexity that this method adds is both up front,
in what you mentioned, and hidden, in that all future development that
involves the minibuffer will force the developers to stop and take account
of this non-obvious mechanism. It makes the minibuffer itself harder to
understand, all for the sake of dealing with a corner case of an optional
feature in a single module's minibuffer use. I *might* be more favorable
to it if the mechanism could be abstracted away into XBL and work
transparently.
Well, let's try the other methods before revisiting this. But I think
they'll all add complexity that will affect future development. BTW, I
think that the hints interaction is such a fundamental part of conkeror
usage that it rates higher than a single module's minibuffer use. And
I'd like this info display to be useful to all users. I won't debate
corner case though.
But it turns out there is another way to get the pixel width of the text
inside of an input box:
* Approach #4
- Add this method to minibuffer.prototype:
input_text_width: function () {
var field = this.input_element.inputField;
//XXX: if we wanted to be really thorough, we could clone all
// ranges, but let's be practical...
var start = field.selectionStart;
var end = field.selectionEnd;
field.select();
var sel =
field.QueryInterface(Ci.nsIDOMNSEditableElement).editor.selection;
var w = sel.getRangeAt(0).getBoundingClientRect().width;
field.selectionStart = start;
field.selectionEnd = end;
return w;
}
Wow, thanks very much. I had a good look for ways to get the text width
from the textbox; this is the first solution I have seen. I was
concerned that that this would affect the X primary selection, but it
doesn't seem to. I wonder if there's any other state that this affects.
To my understanding, side-effects could only arise if there was a net
change to the selection after completion of the event in which the code
ran. (That is what
Re: [Conkeror] Logo
On Fri, Feb 10, 2012 at 09:30:56AM +0100, Konrad Hinsen wrote: John J. Foerch writes: Closest thing to an official logo is a preliminary design in contrib/art. Looks nice, but it's not a very good choice for an icon, which needs to be square. Konrad. Doublechecking that I did in fact use the word preliminary... yep. ;) -- John Foerch ___ Conkeror mailing list Conkeror@mozdev.org https://www.mozdev.org/mailman/listinfo/conkeror
