- Dillo-dev - lists.mailman3.com

release candidate for dillo-3.0.1 ready for testing

by obeythepenguin＠gmail.com

On Fri, 23 Sep 2011 15:24:05 -0400, Axel Beckert <abe at deuxchevaux.org> wrote: > On Fri, Sep 23, 2011 at 03:17:53PM -0400, Benjamin Johnson wrote: >>> it's pretty hard to please everyone when it comes to the look >>> of the UI - and vertical screen space is a sparse resource. >> >> Surely adding four pixels to the tab height isn't wasting screen >> space? > > I disagree. > > Four pixels are already quite a lot on todays "HD" screens which just > have a screen height of 768 pixels. And quite some low-end netbooks > just have 600 or 480 pixels screen height. There, fource pixels do > matter. > > But OTOH: Why not make it configurable if we disagree? > >> It's hardly a dramatic change, yet it dramatically improves the tab >> appearance and readability: >> http://dillo-win32.sourceforge.net/dillo/screenshots.php?closeup=Z02-dillo3… > > Looks perfect and readable for me. > > Kind regards, Axel You say it's wasting space, yet you agree it helps the readability. Obviously that means it serves its purpose, so I don't consider that wasted space. (If you *really* need to conserve vertical space, take a couple pixels off the File menu and location bar; they can spare it. Or maybe set FL_NORMAL_SIZE a bit smaller, since the 14px default is frankly ridiculous.) ~Benjamin

Sept. 23, 2011

release candidate for dillo-3.0.1 ready for testing

by abe＠deuxchevaux.org

On Fri, Sep 23, 2011 at 03:17:53PM -0400, Benjamin Johnson wrote: >> it's pretty hard to please everyone when it comes to the look >> of the UI - and vertical screen space is a sparse resource. > > Surely adding four pixels to the tab height isn't wasting screen > space? I disagree. Four pixels are already quite a lot on todays "HD" screens which just have a screen height of 768 pixels. And quite some low-end netbooks just have 600 or 480 pixels screen height. There, fource pixels do matter. But OTOH: Why not make it configurable if we disagree? > It's hardly a dramatic change, yet it dramatically improves the tab > appearance and readability: > http://dillo-win32.sourceforge.net/dillo/screenshots.php?closeup=Z02-dillo3… Looks perfect and readable for me. Kind regards, Axel -- /~\ Plain Text Ribbon Campaign | Axel Beckert \ / Say No to HTML in E-Mail and News | abe at deuxchevaux.org (Mail) X See http://www.asciiribbon.org/ | abe at noone.org (Mail+Jabber) / \ I love long mails: http://email.is-not-s.ms/ | http://noone.org/abe/ (Web)

Sept. 23, 2011

release candidate for dillo-3.0.1 ready for testing

by obeythepenguin＠gmail.com

On Fri, 23 Sep 2011 15:06:51 -0400, Johannes Hofmann <Johannes.Hofmann at gmx.de> wrote: > it's pretty hard to please everyone when it comes to the look > of the UI - and vertical screen space is a sparse resource. > > Cheers, > Johannes Surely adding four pixels to the tab height isn't wasting screen space? It's hardly a dramatic change, yet it dramatically improves the tab appearance and readability: http://dillo-win32.sourceforge.net/dillo/screenshots.php?closeup=Z02-dillo3… Here's my patch; the line of interest is in the second section: http://dillo-win32.sourceforge.net/patches/23-tabs.diff ~Benjamin

Sept. 23, 2011

release candidate for dillo-3.0.1 ready for testing

by Johannes.Hofmann＠gmx.de

On Wed, Sep 21, 2011 at 01:49:59PM -0800, Roger wrote: > > On Wed, Sep 21, 2011 at 07:48:31PM +0200, Johannes Hofmann wrote: > >Hi, > > > >due to some important privacy fixes we decided to make a > >bugfix release. > >http://www.dillo.org/download/dillo-3.0.1-rc1.tar.bz2 > >is available for download now and want's to be tested. > >Please give it a try and report back. If everything goes smoothly, > >this will become dillo-3.0.1.tar.bz2 on Friday. > > Compiles & runs on i686 (P3). > > ... the tabs look nice, but could increase the height to match the File button > or no? it's pretty hard to please everyone when it comes to the look of the UI - and vertical screen space is a sparse resource. Cheers, Johannes

Sept. 23, 2011

border space

by itsme_410＠yahoo.com

--- On Fri, 9/23/11, corvid <corvid at lavabit.com> wrote: > From: corvid <corvid at lavabit.com> > Subject: Re: [Dillo-dev] border space > To: dillo-dev at dillo.org > Date: Friday, September 23, 2011, 1:44 AM > Globe Trotter wrote: > > same story with http://www.nytimes.com as well as http://www.huffingtonpost.com, for instance. The last > is pretty much unreadable on dillo as a result. Is this a > result of frames? > > These don't look worse than usual to me. I wonder whether > you > have some unusual settings. > Interesting! My settings are the same as what comes with the stock dillorc.

Sept. 23, 2011

border space

by corvid＠lavabit.com

Globe Trotter wrote: > same story with http://www.nytimes.com as well as http://www.huffingtonpost.com, for instance. The last is pretty much unreadable on dillo as a result. Is this a result of frames? These don't look worse than usual to me. I wonder whether you have some unusual settings.

Sept. 23, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by corvid＠lavabit.com

Axel wrote: > there's probably > need for an at least a little bit more fine grained redirect/image > blocking configuration. Here's some code that experiments with letting you list a bunch of filter_auto_requests options in dillorc (same_domain, allow_redirect, allow_image, allow_stylesheet, allow_all), somewhat similar to search_url. Or we could turn it into three boolean preferences. Whatever.

Sept. 23, 2011

border space

by itsme_410＠yahoo.com

--- On Wed, 9/21/11, corvid <corvid at lavabit.com> wrote: > From: corvid <corvid at lavabit.com> > Subject: Re: [Dillo-dev] border space > To: dillo-dev at dillo.org > Date: Wednesday, September 21, 2011, 4:53 PM > higuita wrote: > > ??? Sometimes i find sites where the > right or left border space doesnt > > exist and the text is "touching" the window border, > making it difficult to > > read...??? Example: > > > > http://blog.pfsense.org/ > > http://karidawn.wordpress.com/ > > > > ??? these two are wordpress blogs, so > maybe its also a wordpress theme > > problem, but i think that adding a few pixels to both > borders would help > > a lot... if not hard to do, of course, as there is the > workaround to > > disable CSS to remove the text from the borders > > It might be partly a matter of pages expecting floats to > work, so that there's > something between the text and the window border. same story with http://www.nytimes.com as well as http://www.huffingtonpost.com, for instance. The last is pretty much unreadable on dillo as a result. Is this a result of frames? > > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev at dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev >

Sept. 23, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by abe＠deuxchevaux.org

Hi corvid, On Thu, Sep 22, 2011 at 01:13:55PM +0000, corvid wrote: > Axel wrote: > > One easy way to do so would to display the content of the redirect > > page which usually contains a link to the redirect target. > > That's an interesting idea! I should be sleeping, but Hehe. > how does the attached work for you? Exactly as I expected, i.e. fine, only that way less redirection services than expected include links in the 30x page itself. I just clicked through the first 10 pages of my identi.ca home feed and checked all redirection services I found in use plus a standard Apache redirect (Apache 2.2.29 as on Debianb Lenny): OK: Apache 2.2.9 (tested http://blosxom.net/) NO CONTENTS: is.gd (tested: http://is.gd/dLoEoo) OK: identi.ca/url (tested: http://identi.ca/url/53842451) NO CONTENTS: ub0.cc (tested: http://ub0.cc/xm/97) OK: deb.li (tested: http://deb.li/7s0l) NO LINK: ur1.ca (tested: http://ur1.ca/54dz0) NO CONTENTS: lallus.net (tested: http://lallus.net/852) NO CONTENTS: t.co (tested: http://t.co/viemv5qx) OK: goo.gl (tested: http://goo.gl/dzbq8) OK: feedproxy.google.com (tested: http://feedproxy.google.com/~r/OmMalik/~3/Co-kep24E68/) OK: bit.ly (tested: http://bit.ly/njgxiQ) NO CONTENTS: tinyurl.com (tested: http://tinyurl.com/3fg22d8) So I recommend to add the patch anyway, as it makes things way better even if someone wants to filter these redirects, but there's probably need for an at least a little bit more fine grained redirect/image blocking configuration. Kind regards, Axel -- /~\ Plain Text Ribbon Campaign | Axel Beckert \ / Say No to HTML in E-Mail and News | abe at deuxchevaux.org (Mail) X See http://www.asciiribbon.org/ | abe at noone.org (Mail+Jabber) / \ I love long mails: http://email.is-not-s.ms/ | http://noone.org/abe/ (Web)

Sept. 23, 2011

release candidate for dillo-3.0.1 ready for testing

by abe＠deuxchevaux.org

Hi, On Thu, Sep 22, 2011 at 05:49:44PM +0200, Axel Beckert wrote: > > > http://www.dillo.org/download/dillo-3.0.1-rc1.tar.bz2 > > > is available for download now and want's to be tested. > > > Please give it a try and report back. If everything goes smoothly, > > > this will become dillo-3.0.1.tar.bz2 on Friday. > > > > Works OK here (Debian testing amd64). > > Haven't found time for testing the Debian package build yet, but plan > do so later today, [...] Works fine when being build as package on Debian Unstable, too. Kind regards, Axel -- /~\ Plain Text Ribbon Campaign | Axel Beckert \ / Say No to HTML in E-Mail and News | abe at deuxchevaux.org (Mail) X See http://www.asciiribbon.org/ | abe at noone.org (Mail+Jabber) / \ I love long mails: http://email.is-not-s.ms/ | http://noone.org/abe/ (Web)

Sept. 22, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by corvid＠lavabit.com

Rogut?s wrote: > Benjamin Johnson (2011-09-22 14:47): > > On Thu, 22 Sep 2011 08:18:21 -0400, Axel Beckert > > <abe at deuxchevaux.org> wrote: > > > > >Hi, > > > > > >this bug report came in for dillo 3.0 in Debian Unstable/Testing, see > > >also http://bugs.debian.org/642391 > > > > > >See below for my thoughts. > > > > > > ... snip... > > > > > >Until now I noticed the effects of filter_auto_requests just for > > >images and was quite happy with it. But with images you can click on > > >them and they will be loaded. This does not seem to work for > > >cross-domain redirects. And as annoying as URL shorteners are, I > > >usually do like to follow those redirects. > > > > > >Is it possible to separate the settings for images and redirects? Or > > >is it possible to allow those redirect on a case-by-case base? > > > > > >One easy way to do so would to display the content of the redirect > > >page which usually contains a link to the redirect target. > > > > > > Kind regards, Axel > > > > You know, this may be the first time I've agreed with one of > > Debian's patches. Dillo's default setting breaks many major web > > sites, including Wikipedia, Google Images, and every photo hosting > > site. This hurts Dillo's usability, particularly for new and/or > > less technically knowledgeable users. The average user isn't going > > to stop and think "oh, these images aren't loading because they're > > hosted on a different domain"; he's going to think "Dillo is broken > > because this site doesn't work," and probably switch to a different > > browser. > > > > (Keep in mind, dillo-dev probably represents a small minority of > > users -- those who are both technically knowledgeable and interested > > enough to subscribe and participate. But good software can't be > > designed just for the minority. Firefox isn't popular only for its > > technical merits, but also because it's easy for people to install > > and use.) > > > > Granted, I like Axel's suggestions -- allow more fine-grained > > control, and display redirect links. They're much better than the > > current all-or-nothing approach. I'm just saying while we're at it, > > maybe we should also reconsider Dillo's IMHO overly paranoid > > defaults. A lot more people would probably stick with Dillo if it > > just behaved like they expect it to. > > I, for example, use Firefox with 'permissions.default.image=1', which > makes it show the link to 3rd party images instead of loading them. The > only place where I really missed images was Wikipedia, for which I added > an exception through an add-on. Once upon a time, back before the filter era, I experimented with something more fine-grained that understood adblock syntax http://lists.auriga.wearlab.de/pipermail/dillo-dev/2008-December/005573.html but it wasn't received very positively -- I think due to not being simple enough.

Sept. 22, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by rogutes＠googlemail.com

Benjamin Johnson (2011-09-22 14:47): > On Thu, 22 Sep 2011 08:18:21 -0400, Axel Beckert > <abe at deuxchevaux.org> wrote: > > >Hi, > > > >this bug report came in for dillo 3.0 in Debian Unstable/Testing, see > >also http://bugs.debian.org/642391 > > > >See below for my thoughts. > > > > ... snip... > > > >Until now I noticed the effects of filter_auto_requests just for > >images and was quite happy with it. But with images you can click on > >them and they will be loaded. This does not seem to work for > >cross-domain redirects. And as annoying as URL shorteners are, I > >usually do like to follow those redirects. > > > >Is it possible to separate the settings for images and redirects? Or > >is it possible to allow those redirect on a case-by-case base? > > > >One easy way to do so would to display the content of the redirect > >page which usually contains a link to the redirect target. > > > > Kind regards, Axel > > You know, this may be the first time I've agreed with one of > Debian's patches. Dillo's default setting breaks many major web > sites, including Wikipedia, Google Images, and every photo hosting > site. This hurts Dillo's usability, particularly for new and/or > less technically knowledgeable users. The average user isn't going > to stop and think "oh, these images aren't loading because they're > hosted on a different domain"; he's going to think "Dillo is broken > because this site doesn't work," and probably switch to a different > browser. > > (Keep in mind, dillo-dev probably represents a small minority of > users -- those who are both technically knowledgeable and interested > enough to subscribe and participate. But good software can't be > designed just for the minority. Firefox isn't popular only for its > technical merits, but also because it's easy for people to install > and use.) > > Granted, I like Axel's suggestions -- allow more fine-grained > control, and display redirect links. They're much better than the > current all-or-nothing approach. I'm just saying while we're at it, > maybe we should also reconsider Dillo's IMHO overly paranoid > defaults. A lot more people would probably stick with Dillo if it > just behaved like they expect it to. I, for example, use Firefox with 'permissions.default.image=1', which makes it show the link to 3rd party images instead of loading them. The only place where I really missed images was Wikipedia, for which I added an exception through an add-on. -- -- Rogut?s Sparnuotos

Sept. 22, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by obeythepenguin＠gmail.com

On Thu, 22 Sep 2011 08:18:21 -0400, Axel Beckert <abe at deuxchevaux.org> wrote: > Hi, > > this bug report came in for dillo 3.0 in Debian Unstable/Testing, see > also http://bugs.debian.org/642391 > > See below for my thoughts. > > ... snip... > > Until now I noticed the effects of filter_auto_requests just for > images and was quite happy with it. But with images you can click on > them and they will be loaded. This does not seem to work for > cross-domain redirects. And as annoying as URL shorteners are, I > usually do like to follow those redirects. > > Is it possible to separate the settings for images and redirects? Or > is it possible to allow those redirect on a case-by-case base? > > One easy way to do so would to display the content of the redirect > page which usually contains a link to the redirect target. > > Kind regards, Axel You know, this may be the first time I've agreed with one of Debian's patches. Dillo's default setting breaks many major web sites, including Wikipedia, Google Images, and every photo hosting site. This hurts Dillo's usability, particularly for new and/or less technically knowledgeable users. The average user isn't going to stop and think "oh, these images aren't loading because they're hosted on a different domain"; he's going to think "Dillo is broken because this site doesn't work," and probably switch to a different browser. (Keep in mind, dillo-dev probably represents a small minority of users -- those who are both technically knowledgeable and interested enough to subscribe and participate. But good software can't be designed just for the minority. Firefox isn't popular only for its technical merits, but also because it's easy for people to install and use.) Granted, I like Axel's suggestions -- allow more fine-grained control, and display redirect links. They're much better than the current all-or-nothing approach. I'm just saying while we're at it, maybe we should also reconsider Dillo's IMHO overly paranoid defaults. A lot more people would probably stick with Dillo if it just behaved like they expect it to. ~Benjamin

Sept. 22, 2011

release candidate for dillo-3.0.1 ready for testing

by abe＠deuxchevaux.org

Hi, On Thu, Sep 22, 2011 at 12:46:00PM -0300, Jorge Arellano Cid wrote: > > due to some important privacy fixes we decided to make a > > bugfix release. > > http://www.dillo.org/download/dillo-3.0.1-rc1.tar.bz2 > > is available for download now and want's to be tested. > > Please give it a try and report back. If everything goes smoothly, > > this will become dillo-3.0.1.tar.bz2 on Friday. > > Works OK here (Debian testing amd64). Haven't found time for testing the Debian package build yet, but plan do so later today, including corvid's patch for the redirect filtering issue. Kind regards, Axel -- /~\ Plain Text Ribbon Campaign | Axel Beckert \ / Say No to HTML in E-Mail and News | abe at deuxchevaux.org (Mail) X See http://www.asciiribbon.org/ | abe at noone.org (Mail+Jabber) / \ I love long mails: http://email.is-not-s.ms/ | http://noone.org/abe/ (Web)

Sept. 22, 2011

release candidate for dillo-3.0.1 ready for testing

by jcid＠dillo.org

On Wed, Sep 21, 2011 at 07:48:31PM +0200, Johannes Hofmann wrote: > Hi, > > due to some important privacy fixes we decided to make a > bugfix release. > http://www.dillo.org/download/dillo-3.0.1-rc1.tar.bz2 > is available for download now and want's to be tested. > Please give it a try and report back. If everything goes smoothly, > this will become dillo-3.0.1.tar.bz2 on Friday. Works OK here (Debian testing amd64). -- Cheers Jorge.-

Sept. 22, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by corvid＠lavabit.com

Axel wrote: > One easy way to do so would to display the content of the redirect > page which usually contains a link to the redirect target. That's an interesting idea! I should be sleeping, but how does the attached work for you?

Sept. 22, 2011

Fwd: Bug#642391: dillo: sane default for http redirects [origin: debian-bts@unpopularminds.org]

by abe＠deuxchevaux.org

Hi, this bug report came in for dillo 3.0 in Debian Unstable/Testing, see also http://bugs.debian.org/642391 See below for my thoughts. ----- Forwarded message from The Don Lachlan <debian-bts at unpopularminds.org> ----- From: The Don Lachlan <debian-bts at unpopularminds.org> Date: Thu, 22 Sep 2011 02:07:15 -0400 Subject: Bug#642391: dillo: sane default for http redirects To: Debian Bug Tracking System <submit at bugs.debian.org> User-Agent: Mutt/1.5.21 (2010-09-15) Package: dillo Version: 3.0-1 Severity: important Dear Maintainer, By default, Dillo 3.0 does not follow HTTP redirects between second level domains (foo.com -> bar.com) which breaks URL shorteners and many web sites. This is controlled by the configuration option "filter_auto_requests" and defaults to "same_domain". I've attached two patches. The first modifies the source to "allow_all" by default and the commented entry in the rc file; the second modifies the rc file to "allow_all" by default. The first patch is preferable so that 1) the source code defaults match the shipped rc file 2) users who have an existing ~/.dillo/dillorc (and ignore the global dillorc) will receive the proper default of "allow_all" if they lack a "filter_auto_requests" entry -Lachlan (this is a resend, it left my MTA but never reached the bug logs) -- System Information: [...] --- dillo-3.0/dillorc 2011-09-21 15:38:16.664618360 -0400 +++ dillo-3.0/dillorc 2011-09-21 18:21:26.000000000 -0400 @@ -31,7 +31,7 @@ # allow_all # same_domain : Permit www.example.org to load an image from img.example.org, # but not from the unrelated ad.doubleclick.net. -#filter_auto_requests=same_domain +#filter_auto_requests=allow_all # Change the buffering scheme for drawing # 0 no double buffering - useful for debugging --- dillo-3.0/src/prefs.c 2011-09-21 18:52:17.000000000 -0400 +++ dillo-3.0/src/prefs.c 2011-09-21 18:52:54.000000000 -0400 @@ -40,7 +40,7 @@ prefs.buffered_drawing = 1; prefs.contrast_visited_color = TRUE; prefs.enterpress_forces_submit = FALSE; - prefs.filter_auto_requests = PREFS_FILTER_SAME_DOMAIN; + prefs.filter_auto_requests = PREFS_FILTER_ALLOW_ALL; prefs.focus_new_tab = TRUE; prefs.font_cursive = dStrdup(PREFS_FONT_CURSIVE); prefs.font_factor = 1.0; --- dillo/dillorc 2011-09-21 15:38:16.664618360 -0400 +++ dillo/dillorc 2011-09-21 15:38:42.000000000 -0400 @@ -31,7 +31,7 @@ # allow_all # same_domain : Permit www.example.org to load an image from img.example.org, # but not from the unrelated ad.doubleclick.net. -#filter_auto_requests=same_domain +filter_auto_requests=allow_all # Change the buffering scheme for drawing # 0 no double buffering - useful for debugging ----- End forwarded message ----- Until now I noticed the effects of filter_auto_requests just for images and was quite happy with it. But with images you can click on them and they will be loaded. This does not seem to work for cross-domain redirects. And as annoying as URL shorteners are, I usually do like to follow those redirects. Is it possible to separate the settings for images and redirects? Or is it possible to allow those redirect on a case-by-case base? One easy way to do so would to display the content of the redirect page which usually contains a link to the redirect target. Kind regards, Axel -- /~\ Plain Text Ribbon Campaign | Axel Beckert \ / Say No to HTML in E-Mail and News | abe at deuxchevaux.org (Mail) X See http://www.asciiribbon.org/ | abe at noone.org (Mail+Jabber) / \ I love long mails: http://email.is-not-s.ms/ | http://noone.org/abe/ (Web)

Sept. 22, 2011

I hear there's some SSL unpleasantness going on...

by corvid＠lavabit.com

I wrote: > http://www.theregister.co.uk/2011/09/19/beast_exploits_paypal_ssl/ more: http://www.mail-archive.com/openssl-dev at openssl.org/msg29718.html [Not totally on-topic for dillo, I know, but...]

Sept. 22, 2011

I hear there's some SSL unpleasantness going on...

by corvid＠lavabit.com

http://www.theregister.co.uk/2011/09/19/beast_exploits_paypal_ssl/

Sept. 22, 2011

fltk linkage

by corvid＠lavabit.com

Do we still have a reason for telling people to stick with static linkage when building fltk?

Sept. 22, 2011

release candidate for dillo-3.0.1 ready for testing

by rogerx.oss＠gmail.com

> On Wed, Sep 21, 2011 at 07:48:31PM +0200, Johannes Hofmann wrote: >Hi, > >due to some important privacy fixes we decided to make a >bugfix release. >http://www.dillo.org/download/dillo-3.0.1-rc1.tar.bz2 >is available for download now and want's to be tested. >Please give it a try and report back. If everything goes smoothly, >this will become dillo-3.0.1.tar.bz2 on Friday. Compiles & runs on i686 (P3). ... the tabs look nice, but could increase the height to match the File button or no? -- Roger http://rogerx.freeshell.org/

Sept. 21, 2011

border space

by corvid＠lavabit.com

higuita wrote: > Sometimes i find sites where the right or left border space doesnt > exist and the text is "touching" the window border, making it difficult to > read... Example: > > http://blog.pfsense.org/ > http://karidawn.wordpress.com/ > > these two are wordpress blogs, so maybe its also a wordpress theme > problem, but i think that adding a few pixels to both borders would help > a lot... if not hard to do, of course, as there is the workaround to > disable CSS to remove the text from the borders It might be partly a matter of pages expecting floats to work, so that there's something between the text and the window border.

Sept. 21, 2011

border space

by higuita7＠yahoo.co.uk

Hi Sometimes i find sites where the right or left border space doesnt exist and the text is "touching" the window border, making it difficult to read... Example: http://blog.pfsense.org/ http://karidawn.wordpress.com/ these two are wordpress blogs, so maybe its also a wordpress theme problem, but i think that adding a few pixels to both borders would help a lot... if not hard to do, of course, as there is the workaround to disable CSS to remove the text from the borders Thanks higuita -- Naturally the common people don't want war... but after all it is the leaders of a country who determine the policy, and it is always a simple matter to drag the people along, whether it is a democracy, or a fascist dictatorship, or a parliament, or a communist dictatorship. Voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked, and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same in every country. -- Hermann Goering, Nazi and war criminal, 1883-1946

Sept. 21, 2011

release candidate for dillo-3.0.1 ready for testing

by Johannes.Hofmann＠gmx.de

Hi, due to some important privacy fixes we decided to make a bugfix release. http://www.dillo.org/download/dillo-3.0.1-rc1.tar.bz2 is available for download now and want's to be tested. Please give it a try and report back. If everything goes smoothly, this will become dillo-3.0.1.tar.bz2 on Friday. Cheers, Johannes

Sept. 21, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by onepoint＠starurchin.org

Johannes Hofmann wrote: > On Mon, Sep 19, 2011 at 07:15:58PM +0100, Jeremy Henty wrote: > > > > Diego wrote: > > > > > Please try to load the page, go to another one (bm: for example) > > > and go back to original page. > > > > > > ?do you see diference? > > > > Yes! The layout problem disappears when I return to the original > > page. How can that happen? > > The timing is different when it comes from cache. OK, but I am still puzzled as to how this could make a difference. > I hope to get this also fixed soon. It seems that it is related to > the <pre> rendering issue. I wondered if that might be so, since the symptoms (extremely wide paragraphs) were so similar to my other report. Looking forward to testing the fix. Regards, Jeremy Henty

Sept. 19, 2011

Floating objects

by Johannes.Hofmann＠gmx.de

On Mon, Sep 19, 2011 at 09:04:21AM -0300, Jorge Arellano Cid wrote: > Hi there, > > Well, time to focus on floating objects > (at least for me) > > @Johannes: what's the status and info we have from Sebastian > on his prototype so far? > > I vaguely remember there were some issues with the design, and > on whether to try to implement it inside the textblock or elsewhere. I don't know about these considerations, but I got some patches from Sebastian. I have them integrated in http://flpsed.org/hgweb/dillo_floats/ You can look at test/dw_float_test.cc to see how it should work. I also hooked it up to the CSS system, but real pages fail in funny ways. The obvious limitations are: * FloatSide only handles one float properly. * Rendereing and redrawing has obvious issues. * Clicks are not forwarded properly to floats. Cheers, Johannes

Sept. 19, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by Johannes.Hofmann＠gmx.de

On Mon, Sep 19, 2011 at 07:15:58PM +0100, Jeremy Henty wrote: > > Diego wrote: > > > Please try to load the page, go to another one (bm: for example) and > > go back to original page. > > > > ?do you see diference? > > Yes! The layout problem disappears when I return to the original > page. How can that happen? The timing is different when it comes from cache. I hope to get this also fixed soon. It seems that it is related to the <pre> rendering issue. Cheers, Johannes Hofamnn

Sept. 19, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by onepoint＠starurchin.org

Diego wrote: > Please try to load the page, go to another one (bm: for example) and > go back to original page. > > ?do you see diference? Yes! The layout problem disappears when I return to the original page. How can that happen? Thanks for the suggestion, Jeremy Henty

Sept. 19, 2011

Floating objects

by jcid＠dillo.org

Hi there, Well, time to focus on floating objects (at least for me) @Johannes: what's the status and info we have from Sebastian on his prototype so far? I vaguely remember there were some issues with the design, and on whether to try to implement it inside the textblock or elsewhere. -- Cheers Jorge.-

Sept. 19, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by jcid＠dillo.org

On Mon, Sep 19, 2011 at 10:18:14AM +0200, Diego . wrote: > I have see a similar problem with a page > (http://www.natami.net/knowledge.php) When load from net latout is > bad, but good from dillo cache. > > Please try to load the page, go to another one (bm: for example) and > go back to original page. > > ?do you see diference? Yes, I can reproduce it. This is an old bug present at least in 3.x and 2.x branches. -- Cheers Jorge.-

Sept. 19, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by jcid＠dillo.org

On Sun, Sep 18, 2011 at 10:59:55PM +0100, Jeremy Henty wrote: > > This page[1] renders badly in Dillo - the text wraps to a width much > wider than the browser window. But when I save the page locally it > renders fine! How is this possible? The only CSS is a small inline > <style>, which rules out the obvious guess of a style sheet failing to > load. Fixing all the relative URLs to absolute ones has no effect. > > I'm mystified. Any ideas? > > Regards, > > Jeremy Henty > > [1] http://thbecker.net/articles/rvalue_references/section_08.html After trying remote, local, and a slow local stream, it doesn't happen here... BTW, beware there's a TABLE width=680 in the page. Also, there're two image references that will not be found when local, (and the alt text used instead). If you want to chase it, please try to reproduce it locally first, then reduce the testcase and mist may get thiner. HTH. -- Cheers Jorge.-

Sept. 19, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by darkspirit5000＠gmail.com

I have see a similar problem with a page (http://www.natami.net/knowledge.php) When load from net latout is bad, but good from dillo cache. Please try to load the page, go to another one (bm: for example) and go back to original page. ?do you see diference? Diego. 2011/9/18, Jeremy Henty <onepoint at starurchin.org>: > > This page[1] renders badly in Dillo - the text wraps to a width much > wider than the browser window. But when I save the page locally it > renders fine! How is this possible? The only CSS is a small inline > <style>, which rules out the obvious guess of a style sheet failing to > load. Fixing all the relative URLs to absolute ones has no effect. > > I'm mystified. Any ideas? > > Regards, > > Jeremy Henty > > [1] http://thbecker.net/articles/rvalue_references/section_08.html > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev at dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev >

Sept. 19, 2011

Line breaking bug.

by johannes.hofmann＠gmx.de

Hi Jeremy, On Mon, Sep 19, 2011 at 12:20:49AM +0100, Jeremy Henty wrote: > > In the attached example the first paragraph is only broken before the > last word. The bug depends on the text containing the <pre> element > and on it being inside a <div> inside a <table>. Real live example: > > http://www.devx.com/cplus/10MinuteSolution/34577/0/page/4 I'm working on this. Should be fixed soon. Cheers, Johannes

Sept. 19, 2011

Line breaking bug.

by onepoint＠starurchin.org

In the attached example the first paragraph is only broken before the last word. The bug depends on the text containing the <pre> element and on it being inside a <div> inside a <table>. Real live example: http://www.devx.com/cplus/10MinuteSolution/34577/0/page/4 Regards, Jeremy Henty

Sept. 18, 2011

Page layout bug: disappears mysteriously when the source is saved locally.

by onepoint＠starurchin.org

This page[1] renders badly in Dillo - the text wraps to a width much wider than the browser window. But when I save the page locally it renders fine! How is this possible? The only CSS is a small inline <style>, which rules out the obvious guess of a style sheet failing to load. Fixing all the relative URLs to absolute ones has no effect. I'm mystified. Any ideas? Regards, Jeremy Henty [1] http://thbecker.net/articles/rvalue_references/section_08.html

Sept. 18, 2011

Close dillo with multiple tabs open

by rogutes＠googlemail.com

Benjamin Johnson (2011-09-18 13:33): <...> > Even if you're just viewing static HTML pages, a browser session is > not *itself* a static thing: it is the unique result of a long chain > of user interactions. If you close a text editor, all you lose is a > single file, which is very easy to re-open. If you close a browser, > you lose not only the open pages, but also their history, and for > dynamic pages, the interactions and form posts that generated them > -- in short, the pages' context. And that's much more difficult to > recover, even in the most sophisticated modern browsers. Add the scrolling position to this list, which applies to both, long text files and long HTML pages :) -- -- Rogut?s Sparnuotos

Sept. 18, 2011

Close dillo with multiple tabs open

by obeythepenguin＠gmail.com

On Sun, 18 Sep 2011 13:05:31 -0400, Roger <rogerx.oss at gmail.com> wrote: > You answered this one yourself. The default option for VIM on quit, is > to only > ask if it should quit when the file in the buffer has been modified. > Other > then this, if the file buffer is unmodified, no prompt before quitting is > issued. > > As such, should a browser do the same; only when forms or HTML > composer/editor > has a current editing session, or associated email composition has been > edited, > should a browser ask when to quit? The only thing likely editable > within Dillo > are bookmarks and some forms. > > When you think about it, all data being viewed within the browser is > static to > it's original location. The only information lost is the original > entered URL, > which usually can be found via Google, hence, no information is at > threat of > being lost. I beg to differ; it's no longer 1994, and a good bit of the web -- in fact, I'd dare say most of it -- is no longer static content. In the absence of anything to indicate otherwise, it's better to assume *all* information displayed in a browser could potentially be lost. A couple weeks ago, one of my professors had to take an online exam for certification. Her printer was out, so she had to send it through the department printer. Long story short, somehow the print job got lost in transmission. She'd already closed the browser, and it didn't save her session, so she had to take the entire 4+ hour exam again. Bear in mind, the page she was trying to print -- and subsequently closed out of -- didn't have any HTML forms. Basically it was just an image and some text to say she had successfully completed the exam. Granted, it was *generated* from form data, but it was not itself a "current editing session". Even if you're just viewing static HTML pages, a browser session is not *itself* a static thing: it is the unique result of a long chain of user interactions. If you close a text editor, all you lose is a single file, which is very easy to re-open. If you close a browser, you lose not only the open pages, but also their history, and for dynamic pages, the interactions and form posts that generated them -- in short, the pages' context. And that's much more difficult to recover, even in the most sophisticated modern browsers. In short, the *data* may be static, but the larger browser session is not. Don't confuse the two. ~Benjamin

Sept. 18, 2011

Close dillo with multiple tabs open

by rogerx.oss＠gmail.com

> On Sat, Sep 17, 2011 at 06:00:35PM -0400, Benjamin Johnson wrote: >On Sat, 17 Sep 2011 17:08:52 -0400, Roger <rogerx.oss at gmail.com> wrote: > >> However, how would you like a prompt on your vehicle every time you >> pressed the >> brakes? Do you really want to stop? Or, how about every time you start >> your >> vehicle and you're within an emergency situation and get, "Do you really >> want >> to start your car?" With real world applications, security of preventing >> accidental quitting is application response to users intentions. > >I don't know much about cars, but I do know it's pretty hard to >accidentally start one. And computers are, in terms of physical >interactions, much more difficult to use than most real-world objects (and >less predictable, as well). That's why usability is so tricky to get >right, and why most programs' interfaces tend to follow established >standards and conventions. > >By the way, it's worth noting that vim prompts you to prevent accidental >quitting, BSD vi prompts you to prevent accidental quitting, even freaking >*ed* prompts you to prevent accidental quitting. I find it pretty >annoying sometimes, too, but there have been many times when that prompt >has saved me from far greater frustration. You answered this one yourself. The default option for VIM on quit, is to only ask if it should quit when the file in the buffer has been modified. Other then this, if the file buffer is unmodified, no prompt before quitting is issued. As such, should a browser do the same; only when forms or HTML composer/editor has a current editing session, or associated email composition has been edited, should a browser ask when to quit? The only thing likely editable within Dillo are bookmarks and some forms. When you think about it, all data being viewed within the browser is static to it's original location. The only information lost is the original entered URL, which usually can be found via Google, hence, no information is at threat of being lost. -- Roger http://rogerx.freeshell.org/

Sept. 18, 2011

how to open links etc

by itsme_410＠yahoo.com

--- On Sun, 9/18/11, Benjamin Johnson <obeythepenguin at gmail.com> wrote: > From: Benjamin Johnson <obeythepenguin at gmail.com> > Subject: Re: [Dillo-dev] how to open links etc > To: dillo-dev at dillo.org > Date: Sunday, September 18, 2011, 10:05 AM > On Sun, 18 Sep 2011 00:48:37 -0400, > Walter Dnes <waltdnes at waltdnes.org> > wrote: > > >???Can we have 3 options? > > 1) The "out-of-the-box" default is to ask what to do > > 2) Save to disk > > 3) Specify a helper application.? Allow to > specify parameters as required. > > That's exactly what I was thinking.? Good; that > probably means neither of us is crazy.? (Or else we're > *both* crazy, but at least neither of us is crazy alone!) > > >???And for number 3; puhleeeese don't > "pull a Firefox" by de-referencing > > symlinks.? For instance... > > > > $ ll /usr/bin/abiword > > lrwxrwxrwx 1 root root 11 Sep? 3 02:53 > /usr/bin/abiword -> abiword-2.8 > > > >???When you tell Firefox to use abiword > for *.doc files, it de-references > > the symlink and uses /usr/bin/abiword-2.8.? So a > few weeks from now, an > > update upgrades me to abiword-2.9 and Firefox whines > about not being > > able to find /usr/bin/abiword-2.8 when I click on a > link to a *.doc > > file.? See also https://bugzilla.mozilla.org/show_bug.cgi?id=176486 > > Warning, some strong language from annoyed > users.? See especially > > comments 46, 47, and 48. > > What's the point of dereferencing symlinks, anyway?? > Symlinks in /usr/bin are usually there for a reason, and > dereferencing them would actually take *more* code than > not. > > I'm tempted to fire up my Unix machine and just start > coding now... and yes, I do keep an OpenBSD install and > unpatched mainline sources handy, in case anyone thinks I > only code for the one platform :-) Great!! Let me encourage you, Benjamin!! This will, of course, be a great help! Best wishes. T

Sept. 18, 2011

how to open links etc

by obeythepenguin＠gmail.com

On Sun, 18 Sep 2011 00:48:37 -0400, Walter Dnes <waltdnes at waltdnes.org> wrote: > Can we have 3 options? > 1) The "out-of-the-box" default is to ask what to do > 2) Save to disk > 3) Specify a helper application. Allow to specify parameters as > required. That's exactly what I was thinking. Good; that probably means neither of us is crazy. (Or else we're *both* crazy, but at least neither of us is crazy alone!) > And for number 3; puhleeeese don't "pull a Firefox" by de-referencing > symlinks. For instance... > > $ ll /usr/bin/abiword > lrwxrwxrwx 1 root root 11 Sep 3 02:53 /usr/bin/abiword -> abiword-2.8 > > When you tell Firefox to use abiword for *.doc files, it de-references > the symlink and uses /usr/bin/abiword-2.8. So a few weeks from now, an > update upgrades me to abiword-2.9 and Firefox whines about not being > able to find /usr/bin/abiword-2.8 when I click on a link to a *.doc > file. See also https://bugzilla.mozilla.org/show_bug.cgi?id=176486 > Warning, some strong language from annoyed users. See especially > comments 46, 47, and 48. What's the point of dereferencing symlinks, anyway? Symlinks in /usr/bin are usually there for a reason, and dereferencing them would actually take *more* code than not. I'm tempted to fire up my Unix machine and just start coding now... and yes, I do keep an OpenBSD install and unpatched mainline sources handy, in case anyone thinks I only code for the one platform :-) ~Benjamin

Sept. 18, 2011

how to open links etc

by waltdnes＠waltdnes.org

On Wed, Sep 14, 2011 at 07:34:33PM -0800, Roger wrote > >Easy solution: ask the user whether to open the file, or just save it to > >disk. > > So, the prompt saves my butt 5% of the time, the other 95% of the time I'm > harassed by a prompt like in Windows. > > IMO, configuring for the fewest clicks possible is best. And like > you said, provide a file with an option to to link in external apps if > the user so wishes. (ie mutt's $HOME/.mutt/mailcap is a good example) > > Keeping a file type pointing to null or commented, gets default > (download) action? Can we have 3 options? 1) The "out-of-the-box" default is to ask what to do 2) Save to disk 3) Specify a helper application. Allow to specify parameters as required. And for number 3; puhleeeese don't "pull a Firefox" by de-referencing symlinks. For instance... $ ll /usr/bin/abiword lrwxrwxrwx 1 root root 11 Sep 3 02:53 /usr/bin/abiword -> abiword-2.8 When you tell Firefox to use abiword for *.doc files, it de-references the symlink and uses /usr/bin/abiword-2.8. So a few weeks from now, an update upgrades me to abiword-2.9 and Firefox whines about not being able to find /usr/bin/abiword-2.8 when I click on a link to a *.doc file. See also https://bugzilla.mozilla.org/show_bug.cgi?id=176486 Warning, some strong language from annoyed users. See especially comments 46, 47, and 48. -- Walter Dnes <waltdnes at waltdnes.org>

Sept. 18, 2011

how to open links etc

by itsme_410＠yahoo.com

--- On Thu, 9/15/11, Benjamin Johnson <obeythepenguin at gmail.com> wrote: > From: Benjamin Johnson <obeythepenguin at gmail.com> > Subject: Re: [Dillo-dev] how to open links etc > To: dillo-dev at dillo.org > Date: Thursday, September 15, 2011, 1:21 PM > On Wed, 14 Sep 2011 23:34:33 -0400, > Roger <rogerx.oss at gmail.com> > wrote: > > >> On Wed, Sep 14, 2011 at 08:16:43PM -0400, Benjamin > Johnson wrote: > > > >> Easy solution: ask the user whether to open the > file, or just save it to > >> disk. > > > > So, the prompt saves my butt 5% of the time, the other > 95% of the time I'm > > harassed by a prompt like in Windows. > > > > IMO, configuring for the fewest clicks possible is > best.? And like you said, > > provide a file with an option to to link in external > apps if the user so > > wishes.? (ie mutt's $HOME/.mutt/mailcap is a good > example) > > > > Keeping a file type pointing to null or commented, > gets default (download) > > action? > > > > I didn't want to say much more without being able to > help code it, but I don't > > want to see another prompt popping-up. (ie. Dillo quit > prompt) :-/ > > Then add a "Remember this choice" option so it won't prompt > you again. > > I disagree that all prompts are automatically bad.? > After a while they do get annoying (and there should be some > way to disable them), but for new users the verbosity can be > helpful.? More importantly, it avoids surprising the > user when the program doesn't behave as expected -- it opens > instead of downloads, or vice-versa -- and of course, > different users will have different expectations for which > behavior is "correct." > > (In my own software, I always err on the side of verbosity, > no matter how annoying it is to me personally -- because as > the developer, I know how to disable it on my own machine!) > > ~Benjamin > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev at dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev Is it possible to consider providing the option on how to open links, etc.....i don't mind having it in some preferences file....i understand it may be some work, but this would certainly be very useful to have....? thanks very much!!

Sept. 18, 2011

Close dillo with multiple tabs open

by itsme_410＠yahoo.com

--- On Sat, 9/17/11, Jorge Arellano Cid <jcid at dillo.org> wrote: > From: Jorge Arellano Cid <jcid at dillo.org> > Subject: Re: [Dillo-dev] Close dillo with multiple tabs open > To: dillo-dev at dillo.org > Date: Saturday, September 17, 2011, 3:56 PM > On Sat, Sep 17, 2011 at 09:19:42AM > -0700, Globe Trotter wrote: > > > > > > --- On Sat, 9/17/11, Jorge Arellano Cid <jcid at dillo.org> > wrote: > > > > > From: Jorge Arellano Cid <jcid at dillo.org> > > > Subject: Re: [Dillo-dev] Close dillo with > multiple tabs open > > > To: dillo-dev at dillo.org > > > Date: Saturday, September 17, 2011, 12:09 PM > > > On Sat, Sep 17, 2011 at 11:20:29AM > > > +0200, Johannes Hofmann wrote: > > > > On Sat, Sep 17, 2011 at 05:04:49AM +0000, > corvid > > > wrote: > > > > > Roger wrote: > > > > > > >> > you mean, "confirm > to exit > > > prompt".? Ditto.? Rarely, if ever accidentally > > > quit. > > > > > > >> > > > > > > >> I wasn't sure what my > opinion was > > > on this at first, but now I think an > > > > > > >> option sounds > reasonable. > > > > > > > > > > > > > >Why not, if there is demand > for it and > > > the default is safe > > > > > > >(i.e.? off). > > > > > > > > > > > > Ditto. I've never accidentally > quit. > > > > > > > > > > > > Oh, I've used File > Quit once > in a > > > while, but with that many key sequences or > > > > > > clicks, it's tough to > accidentally > > > quit.? Also, few options under the File > > > > > > menu.? Most times, I just halt > -p. > > > > > > > > > > Just occurred to me that everyone's > been saying > > > that they don't need this > > > > > dialog box. I checked firefox and it > doesn't warn > > > when quitting. Maybe we > > > > > don't need it at all. > > > > > > > > Safari has it and given the fact that we > only recently > > > switched from > > > > Ctrl-Q to Crtl-W for closing tabs, people > might > > > accidentally hit > > > > Ctrl-Q when they want to close just the tab. > It > > > happend to me at > > > > least. So for now I would keep it. > > > > > > ? Same for me. > > > > > > ? Now, if somebody wants to add a preference for > that, > > > could be, > > > but the default should be to ask. > > > > I see no harm in this: put in a default but make it > configurable. > >? While I don't have the urge to use ctrl-q much, > I am still stumped why alt-q > >? does? not work now to close dillo, as I had > conditioned myself to. i usually > > immediately rebound and go with the new action.... > > ? ctrl-q is considered more used across apps. (that > made it the default) > > ? You can set alt-q in keysrc if you want. Thanks, i was not complaining, just saying that my mind had become conditioned to using alt-q when exiting dillo (only)....

Sept. 18, 2011

Close dillo with multiple tabs open

by obeythepenguin＠gmail.com

On Sat, 17 Sep 2011 17:08:52 -0400, Roger <rogerx.oss at gmail.com> wrote: > However, how would you like a prompt on your vehicle every time you > pressed the > brakes? Do you really want to stop? Or, how about every time you start > your > vehicle and you're within an emergency situation and get, "Do you really > want > to start your car?" With real world applications, security of preventing > accidental quitting is application response to users intentions. I don't know much about cars, but I do know it's pretty hard to accidentally start one. And computers are, in terms of physical interactions, much more difficult to use than most real-world objects (and less predictable, as well). That's why usability is so tricky to get right, and why most programs' interfaces tend to follow established standards and conventions. By the way, it's worth noting that vim prompts you to prevent accidental quitting, BSD vi prompts you to prevent accidental quitting, even freaking *ed* prompts you to prevent accidental quitting. I find it pretty annoying sometimes, too, but there have been many times when that prompt has saved me from far greater frustration. ~Benjamin

Sept. 17, 2011

Close dillo with multiple tabs open

by rogerx.oss＠gmail.com

>Safari has it and given the fact that we only recently switched from >Ctrl-Q to Crtl-W for closing tabs, people might accidentally hit >Ctrl-Q when they want to close just the tab. It happend to me at >least. So for now I would keep it. Elinks uses CTRL-C to close tabs. It's kind of tough to think CTRL-Q closes a tab to begin with. Mistake for "close qab" just doesn't sound right to begin with. However, if CTRL-C was mapped to quit, aka "close application", well that would be confusing. Or if "CTRL-Q" were on a keyboard layout next to the "CTRL-C" But still, ELinks does ask to quit before quiting. But this is because I keep typing ":q" (VIM quit), and 'q' is mapped to quit in ELinks. (But this is user error, as I start thinking I'm within VIM at times due to all the console text.) So, some good examples when there's no conflict, and some examples where there is a conflict. Really, when asking, you're just trying to prevent user error. However, how would you like a prompt on your vehicle every time you pressed the brakes? Do you really want to stop? Or, how about every time you start your vehicle and you're within an emergency situation and get, "Do you really want to start your car?" With real world applications, security of preventing accidental quitting is application response to users intentions. <shrugs> I'm sure this has been debated quite a bit, and with Windows, I get so many prompts that I become non-productive & frustrated. -- Roger http://rogerx.freeshell.org/

Sept. 17, 2011

Close dillo with multiple tabs open

by jcid＠dillo.org

On Sat, Sep 17, 2011 at 09:19:42AM -0700, Globe Trotter wrote: > > > --- On Sat, 9/17/11, Jorge Arellano Cid <jcid at dillo.org> wrote: > > > From: Jorge Arellano Cid <jcid at dillo.org> > > Subject: Re: [Dillo-dev] Close dillo with multiple tabs open > > To: dillo-dev at dillo.org > > Date: Saturday, September 17, 2011, 12:09 PM > > On Sat, Sep 17, 2011 at 11:20:29AM > > +0200, Johannes Hofmann wrote: > > > On Sat, Sep 17, 2011 at 05:04:49AM +0000, corvid > > wrote: > > > > Roger wrote: > > > > > >> > you mean, "confirm to exit > > prompt".? Ditto.? Rarely, if ever accidentally > > quit. > > > > > >> > > > > > >> I wasn't sure what my opinion was > > on this at first, but now I think an > > > > > >> option sounds reasonable. > > > > > > > > > > > >Why not, if there is demand for it and > > the default is safe > > > > > >(i.e.? off). > > > > > > > > > > Ditto. I've never accidentally quit. > > > > > > > > > > Oh, I've used File > Quit once in a > > while, but with that many key sequences or > > > > > clicks, it's tough to accidentally > > quit.? Also, few options under the File > > > > > menu.? Most times, I just halt -p. > > > > > > > > Just occurred to me that everyone's been saying > > that they don't need this > > > > dialog box. I checked firefox and it doesn't warn > > when quitting. Maybe we > > > > don't need it at all. > > > > > > Safari has it and given the fact that we only recently > > switched from > > > Ctrl-Q to Crtl-W for closing tabs, people might > > accidentally hit > > > Ctrl-Q when they want to close just the tab. It > > happend to me at > > > least. So for now I would keep it. > > > > ? Same for me. > > > > ? Now, if somebody wants to add a preference for that, > > could be, > > but the default should be to ask. > > I see no harm in this: put in a default but make it configurable. > While I don't have the urge to use ctrl-q much, I am still stumped why alt-q > does not work now to close dillo, as I had conditioned myself to. i usually > immediately rebound and go with the new action.... ctrl-q is considered more used across apps. (that made it the default) You can set alt-q in keysrc if you want. -- Cheers Jorge.-

Sept. 17, 2011

Close dillo with multiple tabs open

by obeythepenguin＠gmail.com

On Sat, 17 Sep 2011 12:45:29 -0400, Rogut?s Sparnuotos <rogutes at googlemail.com> wrote: > Benjamin Johnson (2011-09-17 10:14): >> Speaking of restoring sessions, one of my own users had the very >> clever idea to implement that feature using bookmarks -- save all >> the open tabs in a certain bookmarks section, then load all the >> bookmarks in that section on startup. I haven't coded it yet, but I >> wanted to share since it was such a neat solution to that problem. > > I guess this would be much better than nothing, and separate sessions > could use separate bookmark sections (controlled by a command line > parameter or dialogs)... Though one gets spoiled by the ones like > Firefox, > that take the pain of restoring POST data, this approach is not > bulletproof and perhaps for dillo it would suffice to restore a simple > URL > list into tabs. I should note that this suggestion was for Dillo-Win32, which has the bookmarks code built into the browser (and already supports loading an entire bookmarks section in tabs). So it's quite practical there, since it's just reusing existing code. I doubt the bookmarks approach would work as well in mainline, since it would have to run everything through the DPI, and that code's already complicated enough as it is. But at least the DPI can easily be modified to store bookmarks in a database, for the rare user who has a running MySQL server but lacks the CPU power for Firefox. ;-) ~Benjamin

Sept. 17, 2011

Close dillo with multiple tabs open

by rogutes＠googlemail.com

Benjamin Johnson (2011-09-17 10:14): > On Sat, 17 Sep 2011 01:04:49 -0400, corvid <corvid at lavabit.com> wrote: > > >Just occurred to me that everyone's been saying that they don't need this > >dialog box. I checked firefox and it doesn't warn when quitting. Maybe we > >don't need it at all. > > Of course, Firefox has the ability to restore the last open session > when it starts up; accidentally quitting isn't so bad when you can > just restart and resume where you left off. (Out of curiosity, does > it do that by default?) I think Firefox asks how it should behave regarding sessions when quiting for the first time. > Speaking of restoring sessions, one of my own users had the very > clever idea to implement that feature using bookmarks -- save all > the open tabs in a certain bookmarks section, then load all the > bookmarks in that section on startup. I haven't coded it yet, but I > wanted to share since it was such a neat solution to that problem. I guess this would be much better than nothing, and separate sessions could use separate bookmark sections (controlled by a command line parameter or dialogs)... Though one gets spoiled by the ones like Firefox, that take the pain of restoring POST data, this approach is not bulletproof and perhaps for dillo it would suffice to restore a simple URL list into tabs. -- -- Rogut?s Sparnuotos

Sept. 17, 2011

Close dillo with multiple tabs open

by itsme_410＠yahoo.com

--- On Sat, 9/17/11, Jorge Arellano Cid <jcid at dillo.org> wrote: > From: Jorge Arellano Cid <jcid at dillo.org> > Subject: Re: [Dillo-dev] Close dillo with multiple tabs open > To: dillo-dev at dillo.org > Date: Saturday, September 17, 2011, 12:09 PM > On Sat, Sep 17, 2011 at 11:20:29AM > +0200, Johannes Hofmann wrote: > > On Sat, Sep 17, 2011 at 05:04:49AM +0000, corvid > wrote: > > > Roger wrote: > > > > >> > you mean, "confirm to exit > prompt".? Ditto.? Rarely, if ever accidentally > quit. > > > > >> > > > > >> I wasn't sure what my opinion was > on this at first, but now I think an > > > > >> option sounds reasonable. > > > > > > > > > >Why not, if there is demand for it and > the default is safe > > > > >(i.e.? off). > > > > > > > > Ditto. I've never accidentally quit. > > > > > > > > Oh, I've used File > Quit once in a > while, but with that many key sequences or > > > > clicks, it's tough to accidentally > quit.? Also, few options under the File > > > > menu.? Most times, I just halt -p. > > > > > > Just occurred to me that everyone's been saying > that they don't need this > > > dialog box. I checked firefox and it doesn't warn > when quitting. Maybe we > > > don't need it at all. > > > > Safari has it and given the fact that we only recently > switched from > > Ctrl-Q to Crtl-W for closing tabs, people might > accidentally hit > > Ctrl-Q when they want to close just the tab. It > happend to me at > > least. So for now I would keep it. > > ? Same for me. > > ? Now, if somebody wants to add a preference for that, > could be, > but the default should be to ask. I see no harm in this: put in a default but make it configurable. While I don't have the urge to use ctrl-q much, I am still stumped why alt-q does not work now to close dillo, as I had conditioned myself to. i usually immediately rebound and go with the new action....

Sept. 17, 2011

Close dillo with multiple tabs open

by jcid＠dillo.org

On Sat, Sep 17, 2011 at 11:20:29AM +0200, Johannes Hofmann wrote: > On Sat, Sep 17, 2011 at 05:04:49AM +0000, corvid wrote: > > Roger wrote: > > > >> > you mean, "confirm to exit prompt". Ditto. Rarely, if ever accidentally quit. > > > >> > > > >> I wasn't sure what my opinion was on this at first, but now I think an > > > >> option sounds reasonable. > > > > > > > >Why not, if there is demand for it and the default is safe > > > >(i.e. off). > > > > > > Ditto. I've never accidentally quit. > > > > > > Oh, I've used File > Quit once in a while, but with that many key sequences or > > > clicks, it's tough to accidentally quit. Also, few options under the File > > > menu. Most times, I just halt -p. > > > > Just occurred to me that everyone's been saying that they don't need this > > dialog box. I checked firefox and it doesn't warn when quitting. Maybe we > > don't need it at all. > > Safari has it and given the fact that we only recently switched from > Ctrl-Q to Crtl-W for closing tabs, people might accidentally hit > Ctrl-Q when they want to close just the tab. It happend to me at > least. So for now I would keep it. Same for me. Now, if somebody wants to add a preference for that, could be, but the default should be to ask. -- Cheers Jorge.-

Sept. 17, 2011