- Dillo-dev - lists.mailman3.com

why gtk?
by TheStar July 28, 2003

July 28, 2003

What was the basis for using gtk+ in this project? (ie, why gtk+ as opposed to fltk?) Keep in mind that I'm not (yet) a GUI programmer :))

1 0

[PATCH] version 6 of Dillo tab patch now available!
by Frank de Lange July 27, 2003

July 27, 2003

Hej! I have released version 6 of my tabs patch for Dillo. Head for the patch website (geocities... start your junkbusters and privoxies...) and grab it while it is fresh. The patch was made against 20030723 CVS. Notable improvements include: - the mystery of the disappearing full_screen_off_button is solved, it now stays put. Turns out the dw_gtk_scrolled_window does not show (or raise?) the button when needed. (Sebastian, can you confirm this?). Solved it by hiding, then showing the button explicitly. - worked around (what I consider to be) a bug in dw_gtk_scrolled_window which caused a Gdk-CRITICAL warning about a non-existing window (the widget should check if the window exists/has been mapped before it tries to set a cursor for it. Or am I wrong?). The workaround consists of an explicit gtk_widget_map() call on the docwin in commands.c:480 - backspace now goes back ONE step in history, as it should... No known issues this time... Please... test this patch, and tell me what works/what does not/what you want to see changed... The patch does not just give Dillo tab functionality, it also provides a framework for implementing frames (no pun intended) by separating 'document' from 'interface' code. So, the sooner it is in CVS, the better... As always, you can disable the tab functionality with --disable-tabs. You'll still gain the focus improvements and the shift-click functionality, as well as the document/interface separation. Download the patch here: http://www.geocities.com/ikbenfrank/dillo-20030724-tabs.patch.gz size: 41688 bytes (gzipped, as downloaded) 170140 bytes (uncompressed) md5sum: 4bf94a322d2a406969819ebc56b3faf5 (gzipped, as downloaded) eee05fe3c95e86ed7fe6a01a25ade2f6 (uncompressed) As always, check the site for the most recent version: http://www.geocities.com/ikbenfrank Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

5 5

Re: Same bug again... (fwd)
by Jorge Arellano Cid July 25, 2003

July 25, 2003

---------- Forwarded message ---------- Date: Fri, 25 Jul 2003 12:38:08 -0400 (CLT) From: Jorge Arellano Cid <jcid(a)softhome.net> To: Sebastian Geerken <s.geerken(a)ping.de> Subject: Re: Same bug again... (this forwarded message doesn't include the original attachments) ---------------------------------------- Sebastian, > On Thu, Jul 24, Jorge Arellano Cid wrote: > > Sorry for knocking again on the same bug, but please > > acknowledge you've read my report on the mailing list. > > Yes, I received it (and also your other mail today). I'm quite under > stress at work due an x mio EUR (or whatever) project presentation :-( > > I cannot say when I'll have this bug fixed, I won't have much time at > the weekend, but hopefully in the next week. I understand. BTW, I gave a hand at trying to hunt the bug, and reduced the HTML page to reproduce it (attached), but it also requires a customized slow.c (also attached; it sends by bytes instead of chunks and allows changing the speed). A simple: ...slow?/tmp/b2.html will do. The bug has to do with the new way of handling NOWRAP. Adding more lines to the example table doesn't help, a window resize also doesn't help, but a back/forward fixes it. Hope this findings help. > > > I really don't want to upload the 0.8.0-pre tarball before > > 0.7.3 is out, and although a "first impression" is not accurate > > (agreed!), it helps to calm me down! :-) > > I don't know the state of 0.8.0-pre (I had no time to look at it at > all), but it may be an option to skip 0.7.3, depending on how much > time you estimate until 0.8.0 is mature for release. In this case, I > had some time to test and review the new (actually old, but never > used) code for incremental resizing. 0.8.0 may take a while before the final release; it is stable but we don't know what issues may spring. It also needs polishing before end users can play with it. One way to solve this problem is to tag current CVS as 0.7.3-rc1, and then upload the truckload of changes that 0.8.0-pre has. Then you can freely work on fixing the bug on 0.7.3-rc1, and when that's done we release 0.7.3 to the public. That way, I can finally let our developers play with 0.8.0-pre from CVS. You get out of the release pressure, and the only thing to be done after 0.7.3 is ready, is to forward port the fixes to CVS (currently 0.8.0-pre is up to date with 0.7.3 CVS). The other goodie that comes from making a 0.7.3 release is that as a long time has passed since our last release, not making this intermediate one would put a lot of pressure on 0.8.0's timeline. Sounds good for you? Best Jorge.-

1 0

Frames
by Peter Milne July 24, 2003

July 24, 2003

Hi All, Is there any speculation on when Dillo will be frame ready? I would love to use Dillo for Squirrelmail. Thanks, -- ~~~Pete~~~ http://milneweb.com http://nomorevirus.com

4 6

Directions for building on Cygwin (draft)
by Kelson Vibber July 24, 2003

July 24, 2003

Hello! I have a draft of directions for building Dillo on Cygwin: http://www.hyperborea.org/software/dillo/cygwin.html Please let me know if anything is missing, wrong, or unclear. Thanks! Kelson Vibber www.hyperborea.org

2 1

Excellent work... (fwd)
by Jorge Arellano Cid July 24, 2003

July 24, 2003

---------- Forwarded message ---------- Date: Mon, 21 Jul 2003 18:12:36 +0000 From: d4RK <darkiron(a)tiscali.fr> To: Jorge <jcid(a)softhome.net> Subject: Excellent work... Dillo: the ***fast*** browser I was awaiting for ... No boring stuff like other browsers, all is so simple and easy-to-use! I just come to compile the last version (0.7.2) with a customized toolbar (pixmaps.jg.h), this is now PERFECT :) I don't find u really have to add something (even if I need Javascript for online shopping sometimes, & I've to launch another browser ;) POST-IT: just a remark about the PNG support, and the alpha channel: once the browser has loaded the picture, if it's used again -on a different zone- the background (which should be 'transparent') remains with the first background color...a cache pb? LAST WORD: Thank u !!!!

1 0

thanks (fwd)
by Jorge Arellano Cid July 24, 2003

July 24, 2003

---------- Forwarded message ---------- Date: Fri, 18 Jul 2003 16:09:00 +0200 From: Nicolas <czc(a)free.fr> To: jcid(a)softhome.net Subject: thanks Hello, First of all, sorry for my english :) , I want to thank you and other developers of the Dillo browser, for all that you're already done. I've got a 486 DX4 75Mhz , 24Mo RAM, 3Go HD, on a 640x480 TFT display with Debian Sid. Under Xfree 3.3.6 with the fluxbox window manager, Dillo is what i use to see the web because it's really fast even on my old laptop. I know links and lynx, but i prefer see the web with colors fonts and pictures. Dillo widgets seem to be based on GTK1.2, it's slower than lesstif but reasonnably responsive, and good looking. [...]

1 0

Re: [Dillo-dev] [PATCH] new version of Dillo tab patch, please test...
by Frank de Lange July 23, 2003

July 23, 2003

On Wed, Jul 23, 2003 at 10:36:51PM +0200, Johannes Hofmann wrote: > Hi Frank, > > thanks for the great patch. I noticed a small inconvenience: > > The keyboard shortcut for <Back> (<Backspace>) does not work properly. It > goes back two pages back in the history. The <Back> button works fine though. Yeah, just noticed that as well... Will look into it for the next version of the patch, RSN... Thanks... Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

1 0

[PATCH] new version of Dillo tab patch, please test...
by Frank de Lange July 23, 2003

July 23, 2003

Hi'all! I have released a new version of my tabs patch for Dillo. Head for the patch website (geocities... start your junkbusters and privoxies...) and grab it while it is fresh (see below for download location). The patch was made against 20030723 CVS. Notable improvements include: - focus is on docwin on tab switch, so keyboard navigation works - focus is on location bar on empty new tab/new window - Control-click on link opens new tab, shift-click opens new window - new pref: tab_on_middle_click (default:NO), open a tab (instead of a new window) on link_middle_click - left-to-right tab deletion order (should this be a pref?) known issues: - in fullscreen mode, the 'show controls' button (in the lower right corner of the screen) disappears when switching tabs. Double-click in the window will restore the controls. The button IS reparented, it is just not shown. The widget does not seem to raise it when needed. - when loading a tab in the background, dillo sometimes (quite often actually) gives a Gdk-CRITICAL warning: Gdk-CRITICAL **: file gdkwindow.c: line 1274 (gdk_window_set_cursor): assertion `window != NULL' failed. This seems to be a widget issue (the widget tries to set a cursor on a non-existing window, see earlier messages for more explanation). (Sebastian, have you looked into this yet?) Please... test this patch, and tell me what works/what does not/what you want to see changed... The patch does not just give Dillo tab functionality, it also provides a framework for implementing frames (no pun intended) by separating 'document' from 'interface' code. So, the sooner it is in CVS, the better... As always, you can disable the tab functionality with --disable-tabs. You'll still gain the focus improvements and the shift-click functionality, as well as the document/interface separation. Download the patch here: http://www.geocities.com/ikbenfrank/dillo-20030723-tabs.patch.gz size: 41825 bytes (gzipped, as downloaded) 170684 bytes (uncompressed) md5sum: aa2cd8b800f323c348ff4a381b406bac (gzipped, as downloaded) 03f991b4c3c1cb79609c05879d998efd (uncompressed) As always, check the site for the most recent version: http://www.geocities.com/ikbenfrank Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

1 0

[BUG] segfault on page-back after bookmark delete
by Frank de Lange July 23, 2003

July 23, 2003

While testing my tab-patch I happened upon a segfault in the current CVS version of Dillo. Do this: 1) go to the bookmarks page 2) select the [modify] link 3) select something to delete (can be the 'Unclassified' section in case you have no bookmarks defined) 4) select the delete operation in the drop-down list 5) press [submit] (the bookmark(s)/section(s) will be deleted) 6) click the back button Instead of going back in history, you'll get a page with TWO bookmark sections (a Bookmarks::modify page, followed by a normal Bookmarks:: page). If you press the 'forward' button at this stage, you'll get a page with THREE bookmarks sections. Press back again, and you'll have FOUR bookmarks sections. Et cetera... 5) press back again you should be looking at the 'about:splash' page now 6) to trigger the segfault, either... - press forward or - click any link on the page I submitted a bug for this (bugid = 481) with a short description (that 256 char limit is a bit... limiting...). As this is easy to try for yourself, I have not included trace info... Cheers//Frank -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / +46-734352015 \ \ `--| _/ <Hacker for Hire> \ `---' \ +31-640037120 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

1 0