- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev] [PATCH] version 1 of keyboard navigation patch
by Frank de Lange Oct. 16, 2003

Oct. 16, 2003

Glyn Kennington wrote: > I like most of it. But after a very short time, I've found some things > >that I think need improving. So, in order of decreasing importance: > >1) I've found a bug that will cause dillo to crash: > > Go to a page whose first link is in an alt-tagged image. (The one I've > been using for this bugreport is www.guardian.co.uk .) > > Press the key for `link DOWN'. This will make the image link the > current link (status bar reads "http://www.guardian.co.uk/0,6961,,00.html") > > Go to another page. This can be by clicking Home, Back etc, typing in > another URL, or activating a link on the current page. > > Once dillo is about to start rendering the new page, it dies with a > segfault. gdb tells me that it's at the end of Dw_image_destroy(): > g_free (image->alt_text); > >This crash still occurs if I move to the next link (outside the image) >before going to a different page. > > I have not been able to reproduce this yet, will try again and fix anything I find... But, as stated, it does not happen here... Could have something to do with your next remark? [ LIST: has anyone else seen this? ] >2) Patching problems. I know the email said `use cvs 14/10/2003', but >the patch filename includes the string 20031015, so I attempted to patch >against my just-updated checkout. This normally wouldn't be a problem >over the course of only a day, but today there was some change to the >arguments of a_Commands_set_selection which broke the patch. If you >could include on your webpage the exact date and time of the CVS >checkout, near the link to the patch, it would probably prevent this >sort of confusion from happening again. > > Hmmm... tricky. I'll add a GMT date/time to the CVS version to be used. The date-string on the patch is autogenerated by my patch scripts, it just tells the date I made a release. It is not necessarily related to the CVS date the patch is based on. Sometimes I make incremental changes to a patch without updating CVS. >3) The TODO mentions more bindings. Do you intend to make Back > bindable? > > I intend to make everything bindable. That is, everything which can be bound. That is 'the GTK+ way' of doing things... 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." ]

3 2

Re: [Dillo-dev] [PATCH] version 11 of tab/frame patch, added keyboard navigation
by Frank de Lange Oct. 16, 2003

Oct. 16, 2003

Stephan Goetter wrote: >Hi, > >I just got this, after opening a new tab, going to >http://www.foobar123.org/ (yes, it doesn't exist :) >and then trying to close the Tab. > >Program received signal SIGSEGV, Segmentation fault. >Menu_pick_url (widget=0x8158e80, client_data=0x8113f20) at menu.c:112 >112 gtk_object_remove_data(GTK_OBJECT (bw->menu_popup.over_image), "url2"); >(gdb) backtrace full >#0 Menu_pick_url (widget=0x8158e80, client_data=0x8113f20) at menu.c:112 > bw = (struct _BrowserWindow *) 0x8113f20 >#1 0x40142ef6 in gtk_marshal_NONE__NONE () from /usr/lib/libgtk-1.2.so.0 >No symbol table info available. >#2 0x40187e16 in gtk_signal_remove_emission_hook () from /usr/lib/libgtk-1.2.so.0 >No symbol table info available. >#3 0x40186f3b in gtk_signal_handlers_destroy () from /usr/lib/libgtk-1.2.so.0 >No symbol table info available. >#4 0x40184b8e in gtk_signal_emit () from /usr/lib/libgtk-1.2.so.0 >No symbol table info available. >... > >Greets, >Stephan > > Hmmm, this did not happen here, but that was caused by me using a proxy (privoxy) which produces a nice 'not found' page... Will look in to it asap... 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] version 11 of tab/frame patch, added keyboard navigation
by Frank de Lange Oct. 16, 2003

Oct. 16, 2003

Hi'all, The next version (11) of my tab patch is ready (see bottom of this message for download instructions). The big new features are keyboard navigation and configurable key bindings (currently only for keyboard navigation and scrolling, more will be added later). See below for more information on these features. Keyboard navigation is also available in a separate patch, for those who do not want to apply the tab/frame patch. Watch the mailing list and/or see the patch website (http://www.geocities.com/ikbenfrank/) for more information. New features: - keyboard navigation! - configurable key bindings (for keyboard navigation and scrolling) Fixed bugs - solved the 'missing frames' problem for German (and other?) users. Stephan Goetter pointed me in the right direction (the German locale uses the "," as a decimal separator). Some behavioural changes with regard to stock (CVS) Dillo: - Shift-Ctrl-L (instead of Ctrl-L) for "open URL" dialog - Ctrl-L for "focus location bar" - Ctrl-S for "Save page As..." - Ctrl-U for "view source" while in a document, "clear entry" everywhere else (in stock Dillo this key is bound to "focus location bar" while in a document) - Ctrl-B for "Bookmark this page" - Alt-H for "Home" - "/" starts incremental search instead of normal search. For normal (non-incremental) search use Ctrl-F Support for the 0.7.3 branch has been dropped as of this version of this patch. Sorry, folks, but that is the price you pay for progress... Currently, both tab and frame support can be turned off completely, producing a smaller binary. I might (on request of other Dillo developers) remove the conditional compilation, thereby making tab and frame support a permanent feature of Dillo. If you want to keep the conditional compilation feature, speak out on the mailing list! Keyboard navigation can currently NOT be turned off, let me know if this is needed. Frameset (and iframe) support is enabled by default, it can be disabled using the configure option --enable-xhtml-dtd=strict. You can also use --enable-xhtml-dtd=transitional, which will give you iframe but no frameset support. The default option, --enable-xhtml-dtd=frameset, gives both iframe and frameset support. As usual, tab support can be disabled using --disable-tabs. What does not work yet: - the full_screen_off_button is not shown in frame documents. Use double-click instead to toggle the full screen setting... - the frameborder attribute works, but dw_gtk_scrolled_frame always draws a (focus) shadow around itself - even with the border size is set to 0. This can be solved in several ways - adding a 'borderless' attribute to dw_gtk_scrolled_frame, using a special borderless GTK theme for the widget, etc. In GTK2 this is all much easier (just set the widget's style's xthickness/ythickness to 0), but we're not there yet... - borderless frames can not be resized, even though Dillo always shows a thin border (see above). Who cares about resizing borderless frames? Mozilla does it, but I don't think this is an issue. Agree? - iframes always have a border (see above) Keyboard Navigation =================== You can use the keyboard to navigate links and controls in Dillo. In the default configuration, the following keys are used: Alt-Up: link UP Alt-Down: link DOWN Alt-Left: link LEFT Alt-Right: link RIGHT Enter: follow link or activate GTK+ widget Escape: unhighlight link In a non-frames version of Dillo the following keys are also used: Tab: NEXT link Shift-Tab: PREVIOUS link these keys are not bound by default in a frames-capable version of Dillo because they are used to navigate between frames. You are free to bind them (or any other key combination) to these functions though (see Keyboard bindings) Sometimes the result of link UP and link DOWN seems a bit unpredictable. This can happen when the next UP or DOWN link is currently not visible (below the bottom or above the top of the window). The screen coordinates for words are only known when they are in the visible area of the viewport, so off-screen links are accessed sequentially instead. Search and keyboard navigation ------------------------------ Keyboard navigation works with the Search feature: if search hits a link, it will be focused. Press <Enter> in the viewport to follow the link, <Esc> to unhighlight it. When using the status bar for search (an optional feature in the tabbed version of Dillo), press <Esc> to close the search entry followed by <Enter> to follow the link. A second <Esc> will unhighlight the focused link. Pressing <Enter> in the search entry (popup or status bar) repeats the search. History ------- The focus location is saved when navigating to another document in the same window or tab. This makes it possible to go back in history and continue to the next link on the page. Form fields ----------- The same keys used for link navigation work for form field navigation. Do note however that form fields, being embedded GTK+ widgets, can have their own bindings for some functions. As an example, the space bar is bound to the activate function in radio buttons. Keyboard bindings ================= It is possible to change the keyboard bindings used for keyboard navigation, either through editing the dillorc or by adding a section to any of the current GTK+ rc-files. The syntax is the same, the only difference is the addition of the preference keyword "key_bindings" in the dillorc and the fact that the bindings have to be enclosed in double quotes. An example of the configuration syntax can be found in the dillorc and on the mailing list. Tabs, Frames, Keyboard navigation and browser bloat =================================================== version: 0.8.0-pre patch: dillo-20031015-tabs.patch.gz unpatched: 294132 bytes patched, no tabs, no frames: 308404 bytes patched, with tabs, no frames: 315220 bytes patched, no tabs, with frames: 324852 bytes patched, with tabs, with frames: 333108 bytes (gcc 3.2.2, binutils 2.13.90.0.18, CFLAGS "-Os") The increase in size is due to the addition of keyboard navigation. If you want to be able to compile Dillo without this feature (it takes about 10 KB) let me know through the mailing list. Download instructions ===================== The patch is made against 20031014 CVS. Do not forget to run ./autogen.sh after applying the patch. You can find the patch on the usual (ad-happy, use an ad-blocking proxy or sane browser) site: http://www.geocities.com/ikbenfrank/ More specifically: Dillo 0.8.0-pre (CVS) ===================== http://www.geocities.com/ikbenfrank/dillo-20031015-tabs.patch.gz size: 85716 bytes (gzipped, as downloaded) 371416 bytes (uncompressed) md5sum: 358041a00feeef1706b81d24a985613a (gzipped, as downloaded) bf61b1bdb212c69c30ee1fa6fa74f729 (uncompressed) As always, check the site for the most recent version 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." ]

4 5

RE: [Dillo-dev] FLTK
by Eric Gaudet Oct. 16, 2003

Oct. 16, 2003

> Hi Eric! > Hi Sebastian, > Only some remarks about my experiences. I'm only using Gtk 2.0 and > Pango 1.0, I'll update to see if it makes a difference. > > Startup time is a bit difficult to measure, since the disk cache state > etc. may influence this. What I found out: > ... As I said before, in my tests I always started with a dry run to make sure the libraries and files were loaded or buffered in RAM. The time differences between the first run and subsequent runs was very small in avarage, but could vary a lot. Also, as I said, startup time is slightly longer but still acceptable. If this was the only drawback with gtk2, I would be fine with it. > I've found no difference in drawing time, but this has only played a > bit time in dillo1 (at least for text). I've not configured anything > for Pango, and from what I see on the screen, anti-aliasing (which may > take much time) is turned off. > How do I turn off anti-aliasing? I'd like to try it again if you think this may speed-up redraws. > What *does* take far more time is the initial rendering, while > rewrapping is exactly as fast as before. This (and also that Pango is > only used for drawing, and determing the size of fonts and words) lets > me assume that Dw_page_shape_glyphs is not very efficient. I've tried > to track down this with gprof, but somehow the profile does not give > me much clues. Perhaps it is worth to optimize this code. (The > comparison is a bit unfair because of the different grades of > matureness :-) > I did't try rewrapping (changing the window's size), only scrolling down and up by one full page. I can try to rewrap and scroll line by line if you want. What is the purpose of Dw_page_shape_glyphs? Best, EG

2 1

Development
by Carlos Daniel Ruvalcaba Valenzuela Oct. 15, 2003

Oct. 15, 2003

Hi everyone It sems i lost a lot of the development cycle, now even plugins are implemented! Anyways, here is my wasted effort with dillo-0.7.3: * Fixed dpi.c capi.c, to handle launching multiple dpis via url, example dpi:/bm/ will launch bookmarks plaugs, dpi:/cfg/ will launch the configuration plugin. * Worked on dpilib a bunch of code to initialize comunications from a dpi to dillo, it works pretty well, also handles tag params, and url passed vars. * ddm_srv.c Dillo Download Manager, a download plugin using wget, will be able to download multiple files and show their progress in dpi:/ddm/downloads * cfg_srv.c Configuration, It can tweak a bunch of options for Dillo and DDM. Way back about a month i worked on a plugin daemon too. But with current changes i maded the daemon is not necesary, also dillo can load plugins if necesary. Any comments?, anyone wants the patches? Anyway, cheers. Carlos Daniel Ruvalcaba Valenzuela --------------------------------- Do You Yahoo!? Yahoo! Net: La mejor conexión a internet y 25MB extra a tu correo por $100 al mes.

2 1

Selection leak
by Jorge Arellano Cid Oct. 15, 2003

Oct. 15, 2003

Hi there! Finally the selection riddle is clear: there were some GTK+ warnings very seldomly appearing sometimes, but not in 0.7.3. In 0.7.3 one of the popups was being set as the permanent owner for all the selections, and it was initialized at browser window creation time. As 0.7.3 leaked the popups (not destroying them with the window) the widget remained valid all the time! In CVS the popup leak is fixed, so the selection owner can be destroyed with its window, and thus the warnings. I fixed it like I originally suggested: > Perhaps it'd be better to add a widget parameter to > a_Commands_set_selection: > > void a_Commands_set_selection(GtkWidget *widget, gchar *str) > > So we can set the owner properly. AFAIU the owner gets the > "clear selection" event on the handler it set. For instance, when > the user selects text from the location bar, the "clear > selection" callback is called. This can help to set a custom > handler for when the selected text comes from the rendered area. > This new handler can toggle the selection highlighting off, and > then call a_Commands_clear_selection_callback() to finish the > work. Sebastian: I used "GTK_BIN(bw->docwin)->child" as the selection owner for the rendered area. Is there a better candidate? Commited! Cheers Jorge.-

1 0

Re: [Dillo-dev] Development
by Scorpio Rising Oct. 15, 2003

Oct. 15, 2003

Behold, Carlos Daniel Ruvalcaba Valenzuela scribed in runic symbols on a tube chassis: > * ddm_srv.c Dillo Download Manager, a download plugin using wget, will be able to download multiple files and show their progress in dpi:/ddm/downloads Oy! Could I use this one :-)

1 0

[PATCH] version 1 of keyboard navigation patch
by Frank de Lange Oct. 15, 2003

Oct. 15, 2003

Hi'all, The first version (1) of my keyboard navigation patch is ready. See the end of this message for download instructions. Do note that the tab/frame patch already contains this functionality, so there is no need (and no way) to apply this patch on top of that or the other way around. The patch adds Keyboard navigation with configurable key bindings to (the CVS version of) Dillo. Please test it and let me know what you (don't) like. Even though keyboard navigation does not make much sense with devices without keyboards (pda's, etc), there is no configure option to disable it. Let me know if there is a need for such an option (it is easily implementable). Personally I think there is... For keyboard navigation to work Dillo has to track the position of individual words in a document, which increases memory overhead with 4 bytes per word. However, there seems to be some animosity against the added complexity of conditionally compiled code (enclosed in #ifdef ... #endif). More information on keyboard navigation can be found in the document "Keyboard_Navigation.txt" in the doc/ directory which I'll just reproduce it here: Keyboard Navigation =================== You can use the keyboard to navigate links and controls in Dillo. In the default configuration, the following keys are used: Alt-Up: link UP Alt-Down: link DOWN Alt-Left: link LEFT Alt-Right: link RIGHT Enter: follow link or activate GTK+ widget Escape: unhighlight link In a non-frames version of Dillo the following keys are also used: Tab: NEXT link Shift-Tab: PREVIOUS link these keys are not bound by default in a frames-capable version of Dillo because they are used to navigate between frames. You are free to bind them (or any other key combination) to these functions though (see Keyboard bindings) Sometimes the result of link UP and link DOWN seems a bit unpredictable. This can happen when the next UP or DOWN link is currently not visible (below the bottom or above the top of the window). The screen coordinates for words are only known when they are in the visible area of the viewport, so off-screen links are accessed sequentially instead. Search and keyboard navigation ------------------------------ Keyboard navigation works with the Search feature: if search hits a link, it will be focused. Press <Enter> in the viewport to follow the link, <Esc> to unhighlight it. When using the status bar for search (an optional feature in the tabbed version of Dillo), press <Esc> to close the search entry followed by <Enter> to follow the link. A second <Esc> will unhighlight the focused link. Pressing <Enter> in the search entry (popup or status bar) repeats the search. History ------- The focus location is saved when navigating to another document in the same window or tab. This makes it possible to go back in history and continue to the next link on the page. Form fields ----------- The same keys used for link navigation work for form field navigation. Do note however that form fields, being embedded GTK+ widgets, can have their own bindings for some functions. As an example, the space bar is bound to the activate function in radio buttons. Keyboard bindings ================= It is possible to change the keyboard bindings used for keyboard navigation, either through editing the dillorc or by adding a section to any of the current GTK+ rc-files. The syntax is the same, the only difference is the addition of the preference keyword "key_bindings" in the dillorc and the fact that the bindings have to be enclosed in double quotes. As an example, this configuration (in gtkrc) uses the numerical keypad for link navigation: binding "keypad-keys" { bind "KP_Add" { "focus_move" (0) } bind "KP_Subtract" { "focus_move" (1) } bind "KP_8" { "focus_move" (2) } bind "KP_2" { "focus_move" (3) } bind "KP_4" { "focus_move" (4) } bind "KP_6" { "focus_move" (5) } bind "KP_Enter" { "focus_activate" () } } class "GtkDwScrolledFrame" binding "keypad-keys" You can also use 'widget "*dillo-content" instead of 'class "GtkDwScrolledFrame" if you prefer to use widget names: widget "*dillo-content" binding "keypad-keys" This one (in dillorc, note the enclosing double quotes, the use of single quotes and the preference name 'key_bindings') makes Dillo behave like the vi editor when it comes to link navigation: key_bindings = "binding 'vi-keys' { bind 'k' { 'focus_move' (2) } bind 'j' { 'focus_move' (3) } bind 'h' { 'focus_move' (4) } bind 'l' { 'focus_move' (5) } } class 'GtkDwScrolledFrame' binding 'vi-keys'" The numerical values after "focus_move" are actually GTK enumeration values, but the current implementation does not support this. It would of course look much better to be able to write "focus_move" (GTK_DIR_DOWN) instead of "focus_move" (3)... It is also possible to bind a sequence of actions to a single keypress by adding more actions after the first. Bugz^Hs ======= - Image links without alt text do not work with keyboard navigation. - Image links with alt text work, but Dillo is currently not able to scroll to them. This can lead to 'hidden' focused links, use the scroll keys or scroll bars to bring them in to view. This problem is shared with the Search feature. It is caused by a deficiency in the widget iterator function which is used for scrolling. - The sometimes unpredictable behaviour of link UP/DOWN can be classified as a bug. Or not. It is probably hard to solve this. The root cause is the use of incremental display in dw_page, which is in itself a good feature. To good to mess it up with some hack to get each and every link in the right place before using keyboard navigation. TODO ==== Well, apart from fixing those bugs... Maybe merge focus with search, they have quite a lot in common. This would save some code and memory. It would also mess up the current design, so maybe this is not such a good idea. Add more GTK+ bindings, replacing current key_press event handling. ... Download instructions ===================== There is currently only one version of the patch, made against 20031014 CVS. Sorry, no 0.7.3 support (yet...). As always, have a look at the patch website (http://www.geocities.com/ikbenfrank/) for the latest version of the patch. This patch does NOT apply on top of the tab/frame patch, and the tab/frame patch does not apply on top of this one! The tab/frame patch already contains this functionality, so it is not needed. Also, there is some overlap in the patched files, so I'd have to create a special version of this patch for this purpose. http://www.geocities.com/ikbenfrank/dillo-200310142300-kbnav.patch.gz size: 14336 bytes (gzipped, as downloaded) 62481 bytes (uncompressed) md5sum: 06c21168960cc5de12437ab4abff6d45 (gzipped, as downloaded) a1317f63190ad0df9501bbb257f8c137 (uncompressed) 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." ] -- WWWWW ________________________ ## o o\ / Frank de Lange \ }# \| / \ \ `--| _/ <Hacker for Hire> \ `---' \ +46-734352015 / \ frank(a)unternet.org / `------------------------' [ "Omnis enim res, quae dando non deficit, dum habetur et non datur, nondum habetur, quomodo habenda est." ]

1 0

RE: [Dillo-dev] FLTK
by Eric Gaudet Oct. 14, 2003

Oct. 14, 2003

> Please back your opinions with links, docs, facts or with > whatever helps better to understand the point. Maybe a single > email can show there's no way to do it, or maybe not! > Two applications I used to run regularly when they were on gtk1.2 were ported to gtk2.0: pan and edonkey2000-gtk. I'm not using them anymore, because their rendering became painfully slow: scrolling the lists down takes several seconds to refresh on my 300MMX laptop. After doing some research, is seems that Pango is the culprit: rendering unicode needs some advanced graphic operation also slows down any text rendering. As I yet have to see a fast application using it, I would recommend NOT going with gtk2.0 for Dillo. Best, Eric

2 1

PLEASE CONTACT ME
by Mr. George Rhoney Oct. 13, 2003

Oct. 13, 2003

Dear Friend, I am an accountant and a member of the Contract Tender Board (CTB) of the Petroleum, Oil & Gas Corporation of South Africa (PTY) Ltd also known as PetroSA. I am presently on a special diplomatic duty in the South Africa foreign office in Amsterdam, the Netherlands. I got your contact through discreet inquiries in my search for a reliable individual or company who can handle a strictly confidential transaction that involves the transfer of a reasonable sum of money to a foreign account. I have no doubt in your ability to handle a financial business of considerable amount, which will form the bedrock of an extensive business partnership in due course. In unfolding this proposal, I want to count on your status that you will handle this transaction with all maturity and confidentiality it deserves. The money we want to transfer to you amount to a total of US$15.500,000.00 (Fifteen Million, Five Hundred Thousand United States Dollars) only. It was extracted from various over-invoiced contracts, masterminded by myself and some of my colleagues for our mutual benefit when the contracts was awarded by the Contracts Tender Board (CTB) of the Petroleum, Oil & Gas Corporation of South Africa (PeTROSA) Ltd. The above over-invoiced sum has been drafted and short-listed for final payment to a foreign bank account and it has been reflected as an amount to be paid to a foreign contractor for contract that has been executed for the Petroleum, Oil & Gas Corporation of South Africa (PTY) Ltd. The money is currently floating in the suspense account of the South Africa Reserve Bank (SARB) awaiting claim by a foreign partner, which status we absolutely want you to assume. The code of conduct bureau in South Africa has made it an offence for civil servants like us to(A) operate a foreign account (B) own a foreign company and (C) take up contract while still in active service. This is why we need your assistance. Now is the optimum period to realize this fund following the recent directive given to all-federal parastatals and corporations to pay off all foreign debts. I therefore seek your assistance to allow the remittance of this fund to be made to your personal or companys bank account. We have agreed that 30% of the total sum will be given to you, 65% will be for me and my colleagues and 5% for any expenses that maybe incured during this transaction, after the successful transference of the fund to your bank account. I want to assure you that all modalities for the successful transfer of this money to your account has been perfectly worked out and as such, there is no fear of failure. However, I would want to believe that you are honest enough and will not raise any misgiving attitude in any aspect of the transaction. Most importantly, you will keep this transaction very confidential so as not to tarnish the confidence reposed on these officials by the nations financial house. If this business proposal interest you, kindly indicate your willingness to me through email or by phone (+ 31-619-310-577) for immediate commencement of the transaction. To enable this fund to be successfully transferred to your account, you will also have to provide the following information: 1.Your company name and address that you wish to use as the beneficiary to claim the fund. 2.Your private telephone, fax Numbers. 3.Your Bank name, address and account number where you would want the money remitted. More information about the remittance of the fund to your account will be sent to you when I receive your response. Meanwhile, your swift response will help to meet the ultimatum given to all foreign contractors in this payment quarter to ensure that funds are claimed. We also wish to discuss investment strategy with you. We have the plan of investing part of our fund in real estate with your assistance in your country. Please, treat this information as confidential. Yours faithfully, Mr.George Rhoney.

1 0

PLEASE CONTACT ME
by Mr. George Rhoney Oct. 13, 2003

Oct. 13, 2003

Dear Friend, I am an accountant and a member of the Contract Tender Board (CTB) of the Petroleum, Oil & Gas Corporation of South Africa (PTY) Ltd also known as PetroSA. I am presently on a special diplomatic duty in the South Africa foreign office in Amsterdam, the Netherlands. I got your contact through discreet inquiries in my search for a reliable individual or company who can handle a strictly confidential transaction that involves the transfer of a reasonable sum of money to a foreign account. I have no doubt in your ability to handle a financial business of considerable amount, which will form the bedrock of an extensive business partnership in due course. In unfolding this proposal, I want to count on your status that you will handle this transaction with all maturity and confidentiality it deserves. The money we want to transfer to you amount to a total of US$15.500,000.00 (Fifteen Million, Five Hundred Thousand United States Dollars) only. It was extracted from various over-invoiced contracts, masterminded by myself and some of my colleagues for our mutual benefit when the contracts was awarded by the Contracts Tender Board (CTB) of the Petroleum, Oil & Gas Corporation of South Africa (PeTROSA) Ltd. The above over-invoiced sum has been drafted and short-listed for final payment to a foreign bank account and it has been reflected as an amount to be paid to a foreign contractor for contract that has been executed for the Petroleum, Oil & Gas Corporation of South Africa (PTY) Ltd. The money is currently floating in the suspense account of the South Africa Reserve Bank (SARB) awaiting claim by a foreign partner, which status we absolutely want you to assume. The code of conduct bureau in South Africa has made it an offence for civil servants like us to(A) operate a foreign account (B) own a foreign company and (C) take up contract while still in active service. This is why we need your assistance. Now is the optimum period to realize this fund following the recent directive given to all-federal parastatals and corporations to pay off all foreign debts. I therefore seek your assistance to allow the remittance of this fund to be made to your personal or companys bank account. We have agreed that 30% of the total sum will be given to you, 65% will be for me and my colleagues and 5% for any expenses that maybe incured during this transaction, after the successful transference of the fund to your bank account. I want to assure you that all modalities for the successful transfer of this money to your account has been perfectly worked out and as such, there is no fear of failure. However, I would want to believe that you are honest enough and will not raise any misgiving attitude in any aspect of the transaction. Most importantly, you will keep this transaction very confidential so as not to tarnish the confidence reposed on these officials by the nations financial house. If this business proposal interest you, kindly indicate your willingness to me through email or by phone (+ 31-619-310-577) for immediate commencement of the transaction. To enable this fund to be successfully transferred to your account, you will also have to provide the following information: 1.Your company name and address that you wish to use as the beneficiary to claim the fund. 2.Your private telephone, fax Numbers. 3.Your Bank name, address and account number where you would want the money remitted. More information about the remittance of the fund to your account will be sent to you when I receive your response. Meanwhile, your swift response will help to meet the ultimatum given to all foreign contractors in this payment quarter to ensure that funds are claimed. We also wish to discuss investment strategy with you. We have the plan of investing part of our fund in real estate with your assistance in your country. Please, treat this information as confidential. Yours faithfully, Mr.George Rhoney.

1 0

table-layour revisited
by Ralph Slooten Oct. 10, 2003

Oct. 10, 2003

Hi there all, I posted a while back a bug-report to which I got no reply. This may have been due to me not making myself clear. I'll try again When making a table with a specific width all GUI browsers render this the same, like: <table cellpadding="0" cellspacing="0" width="800"> This gives a table looking like this <--- 800px ---> ------------------------------------------ |table contents | ------------------------------------------ This is no problem with Dillo and other browsers, as they are exactly the same. However when cellpadding and cellspacing are used, like: <table cellpadding="10" cellspacing="10" width="800"> the table should look like this <--- 800px ---> ------------------------------------------ | | | table contents | | | ------------------------------------------ Dillo renders this still incorrectly, by _expending_ the table in width looking like this <--- 820px ---> ---------------------------------------------- | | | table contents | | | ---------------------------------------------- Why is this? Many sites use tables to gain a specific layout in all browsers which I credit them for, but this page layout is altered by this (what I believe to be a) bug. Or am I incorrect as to the specs of html and tables? Greetings Ralph -- http://axljab.homelinux.org/ "...the software said Win95 or better, so I installed Linux"

1 0

[RFC] Configurable keybindings
by Frank de Lange Oct. 10, 2003

Oct. 10, 2003

Hi'all, In my quest for keyboard navigation (which works quite well by now btw) I have come upon Dillo's current inability to define custom keyboard bindings (except for menu bindings). GTK+ offers a mechanism to implement user-configurable keyboard mappings (Bindings) which could be used to add this feature to Dillo. I have made a sample implementation using bindings for GtkDwScrolledFrame, which shows the mechanism to work as it should. The resulting object code is slightly smaller than in the current CVS version, but the source is larger (mainly because of the huge list of 'gtk_binding_entry_add_signal9...0 calls for every key combination). To use the mechanism, widgets need to have signals defined for all user actions. In GtkDwScrolledFrame's case this meant adding "viewport_move" (which replaces the unused user_hchanged and user_vchanged signals), "cursor_move", "cursor_hide" and "cursor_activate" (the latter three for keyboard navigation; it is debatable whether these signals belong in GtkDwScrolledFrame or should move to GtkDwViewport). Keybindings are assigned through GTK+ rc-files or through the dillorc. As an example, this configuration (in gtkrc) uses the numerical keypad for link navigation: binding "keypad-keys" { bind "KP_Add" { "cursor_move" (0) } bind "KP_Subtract" { "cursor_move" (1) } bind "KP_8" { "cursor_move" (2) } bind "KP_2" { "cursor_move" (3) } bind "KP_4" { "cursor_move" (4) } bind "KP_6" { "cursor_move" (5) } bind "KP_Enter" { "cursor_activate" () } } class "GtkDwScrolledFrame" binding "keypad-keys" This one (in dillorc, note the enclosing double quotes and the preference name key_bindings) makes Dillo behave like the vi editor when it comes to link navigation: key_bindings = "binding 'vi-keys' { bind 'k' { 'cursor_move' (2) } bind 'j' { 'cursor_move' (3) } bind 'h' { 'cursor_move' (4) } bind 'l' { 'cursor_move' (5) } } class 'GtkDwScrolledFrame' binding 'vi-keys'" (numerical values after "cursor_move" are actually GTK enumeration values, but the current implementation does not support this. It would of course look much better to be able to write "cursor_move" (GTK_DIR_DOWN) instead of "cursor_move" (3)...) Now I could just add this functionality to the keyboard_navigation patch, or put it in a separate patch. However, since there seems to be some animosity against adding yet even more configuration options, I'd like to know if there is any interest in this feature. If not, I'll just keep things as they are and hardcode some sane defaults (ie. something I like...). In that case, those who want to change keyboard bindings can do it the way they have always done by hacking the source. Any suggestions? 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

dpid
by Jorge Arellano Cid Oct. 9, 2003

Oct. 9, 2003

Hi there! Just made a big commit on dpid (a work we're advancing with Ferdi). It includes the first version of automatical launch; pretty alpha but it should work while we polish. Just be sure to drop the dpid binary in ~/.dillo/ (or whithin PATH), and not to linger with an older version (same applies for the dpis). Cheers Jorge.-

2 2

Re: OpenSSH authentication password problem with dillo
by Jorge Arellano Cid Oct. 9, 2003

Oct. 9, 2003

Hi Mark, > Dear Jorge, > > thank you for dillo, I use it mostly on OpenBSD for fast > browsing from my 266 Mhz computer. However I have an udentified > problem, for which I am not 100% sure that it comes from dillo. > > I have typed in the location bar: http://www. and then I by > mistake pressed ENTER or maybe some other key as well. A > graphical window poped up: > > OpenSSH authentication password for butterfly(a)mock-up.kosmos: > > [X] [X] [X] [ ] [ ] [ ] [ ] [ ] > > > and I was watching this window what it is and where did it come > from because I know dillo does not have any authentication so > far. When I pressed some keys, they were shown like FULL BOX and > in the above example as "X". They were small boxes, filled with > the green color when I pressed keys and they were representing > the password without letting anyone else see it. > > Do you know what it is? > > grep OpenSSH /usr/local/bin/dillo > > did not give results. I did not yet take any source to > investigate and I am not sure if it comes from dillo. > > > Let me know if you know something. Well, if you're using dillo 0.7.3, you have a big problem: that version is SSL unaware. If you're using current CVS, https URLs may be started by dpid, but not with the procedure you describe. It may be that a delayed connection from a previous page could have made it happen. If you can reproduce the window-popup right after the splash, you're probably under attack or something akin. Cheers Jorge.-

1 0

Keybindings
by Ian McDonald Oct. 8, 2003

Oct. 8, 2003

Is there an easy way of breaking the ctrl-<whichever> and right mouse button bindings? I'd rather ask before scratching about in the code. -- ian (building a kiosk)

1 0

[RFC] Keyboard navigation
by Frank de Lange Oct. 7, 2003

Oct. 7, 2003

Hi'all, I'm currently adding keyboard navigation to Dillo. A quick search on the web produces loads of opinions on which keys to use for what purpose, how to navigate using the keyboard, etc. The w3c has produced some guidelines for 'accessibility' (http://www.w3.org/WAI/UA/sect508-UAAG.html#keyboard) which I will take into account for this feature, and the Mozilla project has done quite a bit of brainstorming on the subject. This is a request for some comments from current Dillo users/developers on what you would like Dillo to be like in this respect. My plan is to implement basic keyboard navigation for links and other content which can be activated (input fields, checkboxes, buttons, etc). For this I already have a working implementation. I will probably add accesskey support as well. Currently I use the cursor keys plus a modifier (now CTRL, will probably switch to ALT/META) to navigate to the nearest link (up/down/left/right), while the TAB key navigates sequentially through all 'activatable' content. Shift-TAB should do the same in reverse, but this is currently not working as it should. Enter activates the current link (embedded GTK+ widgets have their own conventions for this, eg. radiobuttons can *also* be selected using the spacebar, etc...). The find feature will be able to set the current focused link (not implemented yet). If you have any suggestions on this subject, send them to the list. Should this feature be optional? On devices without a keyboard (many PDA's) keyboard navigation does not help much after all... On the other hand, the generalised 'focus' mechanism I've added to the viewport can be used for other purposes as well. 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." ]

4 5

Patch to allow pricewatch.com to work
by rcooley Oct. 7, 2003

Oct. 7, 2003

This is a quick and simple patch to address Bug #463. All I did was to _remove_ the 'if' statement in html.c , which ignored the BASE HREF= tag if it was not within HEAD. Does removing that restriction really cause any problems? Just about every other browser handles a BASE tag in the body. Even the w3c.org HTML validator doesn't complain about BASE not being in HEAD. diff -ur /tmp/dillo-0.7.3/src/html.c dillo-0.7.3/src/html.c --- /tmp/dillo-0.7.3/src/html.c Sat Jul 19 16:05:28 2003 +++ dillo-0.7.3/src/html.c Mon Oct 6 20:02:31 2003 @@ -3651,7 +3651,6 @@ const char *attrbuf; DilloUrl *BaseUrl; - if (html->InFlags & IN_HEAD) { if ((attrbuf = Html_get_attr(html, tag, tagsize, "href"))) { BaseUrl = a_Url_new(attrbuf, "", 0, 0); if (URL_SCHEME_(BaseUrl)) { @@ -3665,9 +3664,6 @@ a_Url_free(BaseUrl); } } - } else { - DEBUG_HTML_MSG("the BASE element must appear in the HEAD section\n"); - } } /*

1 0

RPMs for Conectiva 8
by Kelson Vibber Oct. 6, 2003

Oct. 6, 2003

I finally figured out what was preventing Dillo from building on my User-Mode Linux installation of Conectiva 8, so RPMs are now available for both the plain 0.7.3 release and 0.7.3 with tabs+frames patch version 10. Unfortunately, I have not been able to get the virtual network working on this instance, so I cannot route X to my actual desktop to test it. If anyone on this list has a copy of Conectiva 8, I'd appreciate it if you would try both RPMs and verify that the packaged version of Dillo works correctly (or at least as correctly as it does when you build it yourself) and that an icon appears in the KDE/Gnome menus. http://www.hyperborea.org/software/dillo/ On a related note, before I figured out what was missing from my UML installation of Conectiva 8, I decided to try installing an actual copy on spare partition space - and watched it wipe out my primary root partition, my home partition, and my download/media partition. After a long night I was able to recover both data partitions, but I had to reinstall my primary OS. (Not a big deal, since I had my configuration files and my critical data backed up - and I got the rest of my data back.) As a result, I have decided not to install any more distributions directly on that system, so unless/until I can set up a spare system solely for trying things out, anything else I build RPMs for will run under User-Mode Linux (At this point, 5 of the 10 distributions I build RPMs for are running under UML). If there's a pre-built filesystem available, or if it's supported directly by UMLBuilder, it should be easy (assuming I can get the install files). However, for Conectiva 9 or the upcoming Mandrake 9.2 and SuSE 9, I will need to experiment a bit with either building filesystems myself or creating a profile for use with UMLBuilder. (And any SuSE distribution is complicated by the fact that they only provide FTP-based installation unless you buy the boxed set.) The one exception is Fedora Core 1. My current primary OS is Red Hat 9, and I will probably upgrade to Fedora when it's released, assuming they don't rip out anything major. (I was planning to install the latest beta on a spare partition, but I'm a bit reluctant to do that now.) Kelson Vibber www.hyperborea.org

1 0

[PATCH] smart bookmarks
by Tomas Styblo Oct. 6, 2003

Oct. 6, 2003

Hello. I created a patch that adds Mozilla-like "Smart Bookmarks". Smart bookmarks make searching on the web much more convenient. There are two patches attached to this message. The first one is for vanilla dillo 0.7.3, the second one is for dillo 0.7.3 patched with the tabs/frames patch and the SSL patch. The following sample configuration lines should be added to your dillorc file: #------------------------------------------------------------------------- # SMART BOOKMARKS SECTION #------------------------------------------------------------------------- # Use "g dillo" as the URL to search google for "dillo". # The keyword may consist only of alpha-numeric characters and must be # separated from the URL with a semicolon. The URL must contain # the "%s" sequence, which will be replaced with the parameter following # the keyword. smart_bookmark="g;http://www.google.com/search?&q=%s" smart_bookmark="fm;http://freshmeat.net/search/?q=%s&section=projects" ~~~~~~~~~~~~~~~~~~~~ SNIP ~~~~~~~~~~~~~~~~~~~~~ Hopefully someone will find the patch useful. Thank you all for your work. -- Tomas Styblo <tripie(a)cpan.org> PGP: C97EA4B6 / 817E B8A8 1AFD 3256 3181 A730 85CF 7BEB C97E A4B6

1 0

take control of your mortgage!
by cynthia Oct. 5, 2003

Oct. 5, 2003

1 0

Dillo Configuration
by Ringwraith Oct. 3, 2003

Oct. 3, 2003

I'd first like to say thanks for Dillo, its sooo great loading the complete list of CPAN modules in less than 2 seconds, as compared with the nice 5 to 10 seconds with mozilla/firebird :) I do not know if this has been discussed but: One thing I noticed, as Dillo matures, it will obviously need a configuration/preferences menu. I am really starting to get pissed off (pardon lang) at how other browsers use the standard toolkit based widgets to do their configuration. This, IMHO, with the features of today's browsers, clutters up the user interface and causes bloat with the browser. I think having a PURE html interface to doing web browser configuration would do two important things: 1. Gives the flexibility to have a feature rich browser while maintain simplicity. It is a web browser, 90% of the user interface is a web page, so why not make the GUI something the user is familiar with. 2. Reduces bloat. IMHO not so much importantly, 3. Allows for nice customization (themes) of the user interface by someone who knows HTML and vi w/o having to recompile dillo. ;) I am considering to attempt to make a module for the configuration of Dillo that will do this: 1. Configure all aspect of the browser including the (upcoming?) plugin support. 2. Allow pluggable configuration. If plugins have any configurable options (sometimes they do), they will also be configurable using the interface. This module would be accessed through something like: config:// or about:config (i like config:// much better) It would obviously have some security built in so that no one can pull a IE type trick on us and be able to do things they shouldn't. Anyhow this is just a brainstorm, what are your thoughts on doing something like this?

2 1

tab/frame patch ver. 10 makes huge binary?
by Mike Eriksen Oct. 3, 2003

Oct. 3, 2003

Hi I've been running Dillo 0.7.3 from a rpm on a Redhat 9.0 for a while, but wanted to try out the tab/frame patch. So I downloaded the Dillo tarball, applied the patch, ran configure, make and make install. No problem. Since the tarball suggest /usr/local/bin and the rpm-version /usr/bin I'm able to have both versions simultaneously. However: *** ls -lh /usr/bin/dillo -rwxr-xr-x 1 root root 297K <blah blah> ls -lh /usr/local/bin/dillo -rwxr-xr-x 1 root root 3.4M <blah blah> *** A twelve fold increase in size!!!!? Obviously I did something stupid, but what? Gcc is 3.2.2. Mike -- Thinstation FAQ maintainer http://thinstation.sourceforge.net - a light, full featured linux based thin client OS

5 5

patches
by landy Sept. 30, 2003

Sept. 30, 2003

well i admit that either i am blind or can't look correctly but in cvs i do not see any patches. i was told there several patches out there for dillo

4 4

Re: [Dillo-dev] patches
by Kelson Vibber Sept. 30, 2003

Sept. 30, 2003

On Monday 29 September 2003 5:25 pm, landy wrote: > thanks a million, > > i ma having problems with the tab patch, how do i apply it? See http://article.gmane.org/gmane.comp.web.dillo.devel/1108 Just use the right paths for the Dillo source, and update the version number on the patch. (Also, be sure you're using the right version of the patch - if you've downloaded Dillo 0.7.3, use the patch for 0.7.3. If you're using a CVS snapshot, use the patch for CVS.) -- Kelson Vibber www.hyperborea.org

4 3

Yes or No: optional compilation of 'extra' features
by Frank de Lange Sept. 28, 2003

Sept. 28, 2003

Hi'all, Those of you who have tried my tab/frames patch will know that I usually make all 'extra' features optional. This allows those who {do not {need|like}|can not afford} those features to leave them out of the resulting binary. However, it also makes the code somewhat more complex, as the optional features are enabled/disabled through conditional compilation (#ifdef/#ifndef....#endif). The space saved by leaving out tabs and frames is currently about 25 kilobytes (Kib, kB, KB, choose your acronym...). I have gotten some comments on this, suggesting the removal of these conditional statements. This way, Dillo would ALWAYS have the extra features (tabs and frames). Before I say 'yes' or 'no' to these suggestions, I'd like to know what your thoughts are on this. So, speak up. Have a look at the code (look for #ifdef XHTML_DTD_... and #ifndef DISABLE_FRAMES) and tell me whether the savings in space are worth the added complexity... BTW, I do not mind either way... As always, look for the patch at http://www.geocities.com/ikbenfrank A new patch, with some new functionality (sorry Sebastian, functionality was added before you asked me not to add anything new...) will be ready soon. It improves upon the way frame margins are handled, adds an alternative for the find and load popup menu, and includes 'find as you type' (or 'incremental search'). Small parts of the patch (image menu, hand cursor, fixes to widgets) are slowly been added to CVS now. If you want to see the rest added as well, help me out and provide some feedback... 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." ]

7 11

Problems with cookies
by johannes leimbach Sept. 28, 2003

Sept. 28, 2003

Hi ! I am using dillo 0.7.3 on my gentoo box, and it just works great! It's fast, does good rendering, short: I like it One thing bugs me though, it does not save cookies from any message boards (PHPBB). I have cookies enabled (DEFAULT ACCEPT in cookiesrc, file cookies does exist), but still: No writing. I have just noticed google does place a cookie, so it isn't entirely borked. Hey, how can i change this? It would rally make that browser perfect :) Thanks for answer & cu

1 0

Re: Gtk-WARNING with link/image menu
by Jorge Arellano Cid Sept. 28, 2003

Sept. 28, 2003

On Thu, 25 Sep 2003, Frank de Lange wrote: > Jorge, > > Your patch for the link/image menu makes Dillo give Gtk-CRITICAL warnings: > > Gtk-WARNING **: invalid cast from (NULL) pointer to `GtkObject' > > Gtk-CRITICAL **: file gtkobject.c: line 1100 (gtk_object_remove_data): > assertion `object != NULL' failed. > > These are caused by attempts to remove data from non-existing objects in > menu.c: Really weird, as the "non-existing" object is created with the bw and destroyed with it. Anyway to reproduce it? Jorge.-

2 1

[PATCH] Version 10 of tab/frame patch is ready...
by Frank de Lange Sept. 27, 2003

Sept. 27, 2003

Hi'all, The next version (10) of my tab patch is ready (see bottom of this message for download instructions). Control over margins in (i)frames has been improved (no more extraneous margin space in iframe). Dillo (when configured for frameset support) now always renders frameset - even when body content is placed outside the frameset. Frame resize support has been improved a bit. Worked around a bug in GtkContainer which messed up the 'tab' order (focus got stuck in GtkToolbar widgets). Improved/corrected focus handling in dw_gtk_scrolled_frame. New features: - interactive search ('find as you type') on "/" key - new preference: popup_dialogs=([YES]/NO) when set to NO the status bar and location bar are used for search and open URL. Open file and save file still use popup dialogs, though. Changed preferences from previous versions of patch: - frame_blank_in_tab and tab_on_middle_click are now merged into a new preference: tab_instead_of_window=(YES/[NO]) - document_title_use_url is gone, this functionality is now default - (for those who downloaded the interim-patch with find_as_you_type) find_as_you_type is gone, this functionality is bound by default to the "/" key. Use Ctrl-F for normal (non-interactive) search Some behavioural changes with regard to stock (CVS) Dillo: - Shift-Ctrl-L (instead of Ctrl-L) for "open URL" dialog - Ctrl-L for "focus location bar" - Ctrl-S for "Save page As..." - Ctrl-U for "view source" while in a document, "clear entry" everywhere else (in stock Dillo this key is bound to "focus location bar" while in a document) - Ctrl-B for "Bookmark this page" - Alt-H for "Home" - "/" starts incremental search instead of normal search. For normal (non-incremental) search use Ctrl-F The popup menu over images and hand cursor over input image features wihch used to be part of this patch have meanwhile been committed to CVS. The same goes for the bug fixes to dw_image and dw_widget. Support for the 0.7.3 branch will be dropped for the next version of this patch. Some of the new features (interactive search, popup dialog=NO, tab order bug fix) are not availabe in the 0.7.3 version of the patch. if you want the new features you'll have to switch to 0.8.0-pre. Sorry, folks, but that is the price you pay for progress... Currently, both tab and frame support can be turned off completely, producing a smaller binary. I might (on request of other Dillo developers) remove the conditional compilation, thereby making tab and frame support a permanent feature of Dillo. If you want to keep the conditional compilation feature, speak out on the mailing list! Frameset (and iframe) support is enabled by default, it can be disabled using the configure option --enable-xhtml-dtd=strict. You can also use --enable-xhtml-dtd=transitional, which will give you iframe but no frameset support. The default option, --enable-xhtml-dtd=frameset, gives both iframe and frameset support. As usual, tab support can be disabled using --disable-tabs. What does not work yet: - the full_screen_off_button is not shown in frame documents. Use double-click instead to toggle the full screen setting... - the frameborder attribute works, but dw_gtk_scrolled_frame always draws a (focus) shadow around itself - even with the border size is set to 0. This can be solved in several ways - adding a 'borderless' attribute to dw_gtk_scrolled_frame, using a special borderless GTK theme for the widget, etc. In GTK2 this is all much easier (just set the widget's style's xthickness/ythickness to 0), but we're not there yet... - borderless frames can not be resized, even though Dillo always shows a thin border (see above). Who cares about resizing borderless frames? Mozilla does it, but I don't think this is an issue. Agree? - iframes always have a border (see above) Tabs, Frames and browser bloat ============================== version: 0.8.0-pre patch: dillo-20030926-tabs.patch.gz unpatched: 292372 bytes patched, no tabs, no frames: 298624 bytes patched, with tabs, no frames: 306176 bytes patched, no tabs, with frames: 315200 bytes patched, with tabs, with frames: 322784 bytes version: 0.7.3 patch: dillo-0.7.3-20030926-tabs.patch.gz unpatched: 285248 bytes patched, no tabs, no frames: 291852 bytes patched, with tabs, no frames: 298700 bytes patched, no tabs, with frames: 308268 bytes patched, with tabs, with frames: 316556 bytes (gcc 3.2.2, binutils 2.13.90.0.18, CFLAGS "-Os") Download instructions ===================== There are two versions of the patch, one for Dillo 0.7.3 (release) and one for Dillo 0.8.0-pre (CVS). Make sure to download the correct version for the Dillo source you want to patch! If you patch the CVS source, do not forget to run ./autogen.sh after applying the patch. The CVS patch is made against 20030926 CVS, while the 0.7.3 patch is made against Dillo-0.7.3.tar.bz2 (the source tarball on the main Dillo website). Both versions of the patch can be downloaded from the usual (ad-happy, but...) Geocities site: http://www.geocities.com/ikbenfrank/ More specifically for both versions: Dillo 0.8.0-pre (CVS) ===================== http://www.geocities.com/ikbenfrank/dillo-20030926-tabs.patch.gz size: 73796 bytes (gzipped, as downloaded) 315210 bytes (uncompressed) md5sum: e8e1bf9d20ff26991c9d3b7cc1202ee2 (gzipped, as downloaded) 127d224ff29ce263d685718c7ca511f1 (uncompressed) Dillo 0.7.3 (release) ===================== http://www.geocities.com/ikbenfrank/dillo-0.7.3-20030926-tabs.patch.gz size: 88264 bytes (gzipped, as downloaded) 369994 bytes (uncompressed) md5sum: 51df11b26723283124ada4368a9958ea (gzipped, as downloaded) 64769d5f996cec2ac248578a234b32a7 (uncompressed) As always, check the site for the most recent version 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." ]

3 4

Re: [Dillo-dev] Mac OS X Aqua port and new project HEML
by Kelson Vibber Sept. 27, 2003

Sept. 27, 2003

Sorry to extend this thread again... I just wanted to apologize for anything inflammatory I said earlier. Andreas has the right idea: let the past stay in the past and move on. -- Kelson Vibber www.hyperborea.org

1 0

Dillo authorization question...
by Ronald W. Henderson Sept. 26, 2003

Sept. 26, 2003

Group: Does dillo authorization via a popup login authorization dialog? I want to use dillo to manage my linksys router. When I use html management with this router I get the following output from dillo: 401 Unauthorized Authorization Required. I have dillo version 0.7.3 Is there a patch for this? Thanks.... Ronald W. Henderson

2 1

recent changes to nav.c
by Frank de Lange Sept. 25, 2003

Sept. 25, 2003

Jorge, Recently you added a new function (a_Nav_push_nw()) to nav.c. Though that function works fine, it should not be there... Why not? IMO, nav.c should NOT be concerned with windows, tabs, frames or whatever 'visual/physical' location a document is displayed in. It should just concern itself with the state of the document itself. This way, interface and document code are better separated, making it much easier and cleaner to add interface functionality (like tabs, frames, whatever) to Dillo. I have removed this function from my tab/frame patch, as the functionality it provides is - and has for a long time - already present in commands.c. That seems to be a better location for this type of functionality (even though commands.c does end up being a sort of grab-bag for miscellaneous stuff, maybe it should be reorganized a bit...) 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

errno and glibc
by Frank de Lange Sept. 25, 2003

Sept. 25, 2003

Hi'all, Some of you might have run into the same problem as I did when I tried to compile the latest CVS with -Werror: dpi_service.o(.text+0x1d4): In function `dpi_dir': ../../../dillo_tabs/dpid/dpi_service.c:127: undefined reference to `errno' dpi_service.o(.text+0x3fc): In function `rd_dpi_dir': ../../../dillo_tabs/dpid/dpi_service.c:187: undefined reference to `errno' dpi_service.o(.text+0x57b):../../../dillo_tabs/dpid/dpi_service.c:211: undefined reference to `errno' dpi_service.o(.text+0x8df): In function `send_servers_list': ../../../dillo_tabs/dpid/dpi_service.c:290: undefined reference to `errno' This is caused by the failure to include <errno.h>, which is where errno is declared. Just add #include <errno.h> to dpid/dpi_service.c and it compiles cleanly (at least here...) <PATCH> --- dillo/dpid/dpi_service.c 2003-09-15 02:56:41.000000000 +0200 +++ dillo_tabs/dpid/dpi_service.c 2003-09-25 17:13:31.263098504 +0200 @@ -25,6 +25,7 @@ #include <string.h> #include <dirent.h> #include <error.h> +#include <errno.h> #include <glib.h> #ifdef TEST #include "testdat.h" </PATCH> Jorge, can you apply this patch? 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

Mac OS X Aqua port and new project HEML
by Robin Rowe Sept. 25, 2003

Sept. 25, 2003

FYI, a screenshot of dillo running native on Mac OS X Aqua. http://heml.sourceforge.net/pix/dillo.aqua.811x734.png That and my Windows port are still under development. Looking to have those ready for users by the end of the year. Devout Linux advocates on the dillo team will be relieved to know that these ports will not be known as dillo, that I've founded a new project on SourceForge called HEML. We will support Linux, Mac, and Windows without prejudice. http://heml.sourceforge.net/ I'm letting you know about it to avoid confusion. Not seeking your help or interest. Ciao, Robin --------------------------------------------------------------------------- Robin.Rowe(a)MovieEditor.com Hollywood, California www.CinePaint.org Free motion picture and still image editing software

4 4

Image submenu for images that are links
by Jorge Arellano Cid Sept. 23, 2003

Sept. 23, 2003

Hi there, I just commited this patch (see subject). I got some GTK+ warnings while developing it but have no longer been able to reproduce them. Maybe they were wiped out maybe not. Please test it. Cheers Jorge.-

1 0

dillo browser repaint
by Vikram Sept. 21, 2003

Sept. 21, 2003

Hi, I force dillo to browse to a new page based on some external event; the code is as follows [with text containing the url to goto] <snip> gchar* new_text = a_Url_string_strip_delimiters((gchar*)text); DilloUrl* url = a_Url_new(new_text, NULL, 0, 0); if (url) { a_Nav_push(transfer_bw,url); a_Url_free(url); } </snip> however i notice that the though the browser DOES goto the new url the screen [browser window] doesn't get painted properly. i seem to have to scroll up and down a couple of times to get it to repaint completely. Any idea why this would be ? [I don't know if this is related but the url's dillo is forced to goto are served typically by a tomcat servlet engine] Thanks, Vikram

2 1

patches
by landy Sept. 15, 2003

Sept. 15, 2003

where can i find patches for dillo

2 1

Image popup
by Jorge Arellano Cid Sept. 15, 2003

Sept. 15, 2003

Hi, A right-mouse-button popup for images has been added, please test and enjoy! :-) When the image is also a link, the link popup will appear instead. The plan is to add an "Image menu..." submenu entry at the end of it, and handle it in way similar to the "Jump to..." submenu in the page popup. Cheers Jorge.-

1 0

Re: [Dillo-dev] Re: [Patch] Search feature and more
by Jorge Arellano Cid Sept. 15, 2003

Sept. 15, 2003

Hi Johan! It took a long time to analyze the three patches, but finally here we are: [I'm not quoting the previous email bacause it got too big and can be easily found at the web archives] The patches were: 1.- Made main document window grab focus on startup and after open url. Commited! In the meanwhile, Owen taylor answered my post to GTK-list and it become clear that the "Nirvana"-widget issue is solved in GTK2. Thanks a lot for your comments on this. You may find interesting to read the answer: http://mail.gnome.org/archives/gtk-list/2003-June/msg00307.html 2.- Accelerator keys. Huh! These are to be picked very carefully! It took me a lot of work to figure simple mnemonic shortcuts that integrate well with the UI model. The chosen ones are: Shortcut Mnemonic ------------------------------------------------------------ Ctrl-U "Ctrl-URL" (to edit the URL) Ctrl-H "Ctrl-Hide" (Hide/show controls) Ctrl-S "Ctrl-Search" (search the web) Ctrl-R "Ctrl-Reload" (reload current page) ------------------------------------------------------------ (integrating with the existent) Ctrl-L "Ctrl-Location" (enters a new URL) Ctrl-F "Ctrl-Find" (find text) Ctrl-Q "ctrl-Quit" (quit window) (Note: Ctrl-S is not implemented yet) These should provide good defaults as Havoc suggests. Please give them a test drive to get the feel. Ah, while thinking of it, I thought it'd be good to add the ESC key as a way to switch from the location box into the main window, it can be easily done with the key handler you wrote. 3.- Search feature Uncommited yet, the functionality is perfect. Just some minor details in the UI: A small icon to the right of the location box (hideable as the othe icons), with Ctrl-S as shortcut, that pops up a dialog (just as find text and open url). The icon's tooltip can be "Search the Internet", "Search the Web", or something better :-) The final touches: * While working on the location bar shortcuts, I thought it'd be good to add the ESC key as a way to switch from the location box into the main window, it can be easily done with the key handler you wrote. * It'd be also good to add middle-click over the clear-url button, as a way to "paste-push" an URLs. Instead of the the current three step sequence: - Press clear-url - Paste URL - Press enter I carefully considered your previous email and this model should take good care of the several points we talked about. Half of the work is on CVS now, and the above mentioned details are pending. Please send me a new patch with the modifications to fast-track-commit-it-in! Best Jorge.-

3 2

Cosmetic 0.7.3 patch
by Andrew R. Greenwood Sept. 13, 2003

Sept. 13, 2003

Greetings, I originally started changing the following for fun, but after I had finished I found these cosmetic touches give Dillo a more polished and appealing look. I felt it was a nice improvement, so I decided to diff the changes (as instructed on the site) and subscribe to this mailing list. I hope others like these changes as much as I do. Here is a summary of what I've done: * Added a space at the beginning and end of the image and page progressbar text to offer a nice clean padded look in all of the panel_size modes * Cleaned up case inconsistencies and wording in context menus; verbage from Mozilla * Removed trailing menu separators and uppercase title menu entries from context menus to give them a more polished look * Changed the "View Source" window title to "Dillo: View Source" for window management consistency Any feedback will be greatly appreciated =) Andrew R. Greenwood <shadearg(a)sdf.lonestar.org>

1 0

ipaq dash keys
by Vikram Sept. 12, 2003

Sept. 12, 2003

Hi I am using familiar 0.7 on a h3670. I have a modified version of dillo which adds some custom functionality to it. I need to be able to trigger certain events on pressing the ipaq dash buttons [these are the ones which launch calendar, filemanager etc and are wrapped all round the arrow keys of the ipaq]. I see that dillo via GdkEventKey [in interface.c] senses the arrow keys but this function does not seem to get events from the ipaq dash key presses. Is there anyway I can get such keypresses ? Is it necessary I have change some system wide seting [like use xmodmap?] ? Thanks, Vikram

1 1

New dpid patch
by Ferdi Franceschini Sept. 11, 2003

Sept. 11, 2003

There is a new dpid patch for dillo-0.8-pre available from http://www.geocities.com/ferdif2003/ Plugins are now stored in a directory structure under libdir/dillo/dpi. You need to put dpi_dir=/usr/local/lib/dillo/dpi in ~/.dillo/dpidrc so that dpid can find the plugins, there should be no other entries in this file. Dpid now uses a .filter suffix in the plugin name to identify plugins which read and write to stdio (eg hello.filter.dpi). The following changes affect the CVS repository: Renamed dpiddir.[ch] to dpi_socket_dir.[ch] New files dpi_service.[ch] Cheers, Ferdi

2 1

[RFC] Find as you type, statusbar instead of popups, etc
by Frank de Lange Sept. 11, 2003

Sept. 11, 2003

Hi'all, My current tab/frame patch (which will probably be released AFTER the 20th of september, as I am going abroad for a while) contains a find_as_you_type implementation (like Mozilla, search results are updated interactively). I also included an option to use the status bar and location bar instead of popup windows for search and open_url (popups are quite often not worth the hassle...). User controls for find (and find_as_you_type) are the same as the normal find dialog ("/" or CTRL-F to start find, ESC to close, enter to search/search next), but find now retains the last search string between searches. Code size impact is minimal, in case of a popup_less version code is actually quite a bit less complex (and a bit smaller) than the current CVS implementation. Is this something you'd like to see in Dillo, or should I keep these to myself? On the frameset front, control over margins in (i)frames has been improved (no more extraneous margin space in iframe). Dillo (when configured for frameset support) now always renders frameset - even when body content is placed outside the frameset. Frame resize support has been improved a bit. These changes introduce two new preferences (popup_dialogs and find_as_you_type). Both have defaults which reflect unpatched Dillo's behaviour. Please let me know if this sounds like an improvement (or the opposite) to you. 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." ]

3 3

[PATCHES] GTK1 focus problem, solution
by Frank de Lange Sept. 10, 2003

Sept. 10, 2003

Jorge, Sebastian, world, In a previous message I described a focus problem with Dillo (focus is 'trapped' in invisible widget). I have traced the problem to the GtkToolbar class. GtkToolbar is a GtkContainer, so it always gets focus even when GTK_CAN_FOCUS is unset. Hoewever, it redefines the focus signal function to NULL, thereby trapping focus inside the widget. To get normal focus behaviour, the GtkToolbar class's focus function needs to be changed to something like { return FALSE; }. I have patched interface.c to this extent. Patch is attached to this message. To get a complete, two-way tab focus chain (ability to tab around all focusable widgets in Dillo, forward and backward) my focus patch for dw_gtk_scrolled_frame is needed, otherwise dw_gtk_scrolled_frame will trap focus inside itself. You can find the patches on my patch site: http://www.geocities.com/ikbenfrank/dillo-200309101437-interface_focus.patc… (size: 660 bytes) http://www.geocities.com/ikbenfrank/dillo-200309101116-focus_chain.patch.gz (size: 910 bytes) 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

[PATCHES] hand_cursor and image_popup, some remarks
by Frank de Lange Sept. 10, 2003

Sept. 10, 2003

[ Jorge, your mailserver is not healthy... I tried to send this 3 times without success, so it goes to the list instead... ] On Mon, 8 Sep 2003 11:13:32 -0400 (CLT) Jorge Arellano Cid <jcid(a)softhome.org> wrote: > You're right. In that case it causes more troubles than help. > Please send me the patch to commit it to CVS. It is part of the tab/frame patch... It uses the document abstraction. In non-patched Dillo, it would be something like: if(bw->nav_stack_ptr != -1) gtk_widget_grab_focus(GTK_BIN(bw->docwin)->child); else gtk_widget_grab_focus(bw->location); Question is, how much sense does this make in a single-document window? When the window is created, the navigation stack is empty, so focus will always be on the location bar... By the way, there are some problems with 'hidden focus' in Dw. These problems only become directly noticeable in the tab/frame patched version, but they are present in CVS Dillo as well. To see the problem: 1. start dillo on a longish directory listing or a long document (so that the vertical scroll bar appears) 2. use up and down arrow keys to scroll the document. This works in unpatched Dillo. In the (current) patched version, it only works AFTER pressing the UP key once... 3. press SHIFT-TAB twice. Focus is now set on... something. Something invisible. There is no visible focus cue 4. press DOWN key. Nothing happens, no matter how often you press it... Same for TAB, ENTER, RIGHT, ... 5. press UP or SHIFT-TAB or LEFT key. Focus is now on the document, and scrolling keys work as they should. The problem does not show (in patched or unpatched Dillo) when in fullscreen mode. I've searched for the cause of this problem, but have not yet found it. > Clashing with other apps. is almost unavoidable (unless we > clone), but we have chosen to rely over an underlying cognitive > model for our UI. It has been very succesful, and praised! > > Now, as a matter of serendipity :), the bouncing keyboard may > have shown us both a better way. If we make Ctrl-L focus the > location bar (as Ctrl-U does now) and a second Ctrl-L to clear > it, that may be a good choice. Well, Ctrl-U is generally used for 'clear entry' in GTK1 (not in GTK2 though), so I think it is best to use it for that purpose in Dillo as well. For GTK2, following the toolkit default bindings is the logical thing to do. Even better is to allow users to decide for themselves which bindings they want. GTK2 does not allow this by default, but it can be re-enabled. No matter how well thought out a model is, the user should *always* be able to override the designer's defaults. Even if that user messes up a perfectly thought out cognitive model. That is also why I make just about everything user-selectable using preferences with sane default values. So, one of my next goals will probably be to implement user-selectable accelerators in Dillo... Based on GTK's default mechanisms for doing such a thing. > Of course, a more orthogonal way to do it, is the same > functionality, but inside the location popup. But please > experiment with the above suggested idea, and tell me your > findings. Generally I find that popup menu's are more a hinder than a help. This is more of a problem with window managers which do not correctly manage them than with the basic concept of popups, but as the state of the art currently stands I try to avoid them as much as possible. I will therefore try to implement the functionality of Dillo's transient popups (find and open url) in the main window, the former probably in the status bar and the latter in the location bar. And, of course, make a preference to allow the user to choose which version to use - popup or main window. As for using CTRL-L for clearing the location bar, I do not think this is a good idea. But, it is not for me to say it is a bad idea either. > PS: If you send me the requested patches, chances of fast commit > are very high Attached two patches, one for hand_cursor, one for right_click on image. Some remarks: The 'image save as' function in the image popup works, but is dependent on the dpi for the actual save function. That dpi (or, rather, wget) does not know how to handle some link types (eg. file: links), so it will fail on those. The hand_cursor patch works fine, but input_images do NOT work as stated in my previous message. Reverse the patch to dw_button.c to get them back to live... Maybe the image popup should be a submenu on other popups? See the frame/frameset popups (in the tab/frame patch) for an example of what I mean. That way, the image can also be accessed in links (and, maybe, in input images, even those currently do not react to RMB-clicks for obvious reason...) Please be aware of the fact that I will have to keep on adding these patches to the tab/frame patch while the document abstraction has not been committed to CVS... So I will have to patch my own patches, not something I want to keep on doing too long... 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." ]

2 2

FW: [Dillo-dev] How do your clear the dillo cache
by Ronald W. Henderson Sept. 9, 2003

Sept. 9, 2003

Folks: Let me be more accurate with my problem: The graphic images (which are *.png files) are the ones which are cached. If I hit the reload botton the html code updates ok but the graphics don't. I need to start and stop Dillo for the graphic images to be updated. Folks: How do you clear the Dillo cache? I am using dillo (0.7.3) to view an updated local html file: "file:xxx.html". I have to stop and restart dillo to see the updates. The "Reload" Button or open file menu option just keeps loading the cached pages... Thanks.. _______________________________________________ Dillo-dev mailing list Dillo-dev(a)lists.auriga.wearlab.de http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

2 1

Two small things that need implementation...
by Jorge Arellano Cid Sept. 9, 2003

Sept. 9, 2003

Hi, These things have waited for a long time unimplemented. Both don't require much code, but some knowledge: 1.- A right-click popup for images (we have all the code and functions of it, what we need is to hook an empty popup to right-click over images). 2.- Input images work, but there's a detail lacking. The arrow- shaped cursor should change to a hand-shaped one (as with links). Anyone? Cheers Jorge.-

3 3

Feature request - Open tab in background
by Colin Michael Yates Sept. 8, 2003

Sept. 8, 2003

Hi, Is it possible to open a tab in the background, instead of it always getting focus? I generally go to a forum (for example), and open the messages that I want to read in a seperate tab, when I have selected all the messages I want to read, I then go through them one by one. Is this feature already implemented, or does anybody with more C coding skill then me want to implement it :) Cheers, Colin

2 1

How do your clear the dillo cache
by Ronald W. Henderson Sept. 7, 2003

Sept. 7, 2003

Folks: How do you clear the Dillo cache? I am using dillo (0.7.3) to view an updated local html file: "file:xxx.html". I have to stop and restart dillo to see the updates. The "Reload" Button or open file menu option just keeps loading the cached pages... Thanks..

1 0

Check for unterminated quoted strings
by Andreas Kemnade Sept. 6, 2003

Sept. 6, 2003

I found this tag on pages at www.reichelt.de: <A HREF="Info.html?P1mqhKwQAQsAAHVmTEY+Suche+0+50+LM%20324%20D%20SMD+lm324" onmouseover="return overlib('<TABLE><TR><TD><font face=\'Verdana, Arial, Helvetica, sans-serif\' size=1 color=\'#FFFFFF\'></font></TD></TR></TABLE>',CAPTION,'LM 324 D SMD');" onMouseout="return nd();"> It looks like correct html to me or am I wrong? But dillo does not render it correctly. It is rendered like this: <A HREF=... onmouseover="...TABLE"><TR><TD>... So the check for unterminated quoted strings confuses the table handling. With a small change to html.c, it will be rendered correctly: /* Skip over quoted string */ buf_index++; buf_index += strcspn(buf + buf_index, - (ch == '"') ? "\">" : "'>"); + (ch == '"') ? "\"" : "'"); if ( buf[buf_index] == '>' ) { /* Unterminated string value? Let's look ahead and test: * (<: unterminated, closing-quote: terminated) */ I don't know why that check is there. Does HTML allow quoted strings to be not terminated? I guess there is a better solution than mine. Greetings Andreas Kemnade

1 0