- Dillo-dev - lists.mailman3.com

[Dillo-dev]Problems with the activation of the URL entry
by Sebastian Geerken June 18, 2004

June 18, 2004

Hi! The current CVS version is supposed to visit the URL in the selection, when the user clicks with the middle mouse button on the "clear URL" button. However, I got it once working (strangely), but currently it does not work. I use Gtk+ version 1.2.10. The function Interface_entry_paste_selected first copies the selection into the entry (as you see from the comment, I did not really understand the code), and then calls gtk_widget_activate, which emits the activation signal of the widget, which is "activate" for GtkEntry, and normally emitted when the user presses the enter key within the entry. This signal is connected to the function a_Interface_entry_open_url. When a_Interface_entry_open_url is called in this case, gtk_editable_get_chars returns an empty string, it seems that the "real" copying of the selection in Interface_entry_paste_selected is somehow delayed. As workarounds, I see two possible solutions: 1. Also queueing our code in an idle function. The attached patch demonstrates this successfully, although it has to be cleaned up (in some cases, the active idle function has to be removed again). 2. Calling a_Interface_entry_open_url directly with the contents of the selection. (But how do I get the latter?) I'm not very familiar neither with selection in general (i.e. the protocols provided by ICCCM), nor with the mechanisms Gtk+ provides. If someone knows a better solution, please describe it (or send a patch). If not, I'd clean up and commit solution 1. Sebastian

2 3

[Dillo-dev]More on the UTF-8 issue
by sam+dillo＠chaosring.org June 16, 2004

June 16, 2004

First of all, I would like to thank Jorge for working on Dillo; it fills the need for there to be a lightweight *nix browser for older systems and times when one doesn't want to devote a lot of system resources just to access the world wide web. It is an excellent browser. When performing cross-browser verification, I discovered that Dillo 0.8.1 does not honor the following line in an HTML header: <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> Despite the presence of this line, Dillo assumed that the web page was encoded in Latin-1. Jorge, in a message with the subject "UTF-8", sent on June 3rd, wondered why someone would encode an English-language webpage with UTF-8. I can not speak for other webmasters, but the reason I personally encode my webpage in UTF-8 is because I do my web design on a Fedora Core 2 box, an OS that generally assumes that plain text files are encoded with Unicode instead of the iso-8859 mess. Also, Unicode gives one the flexibility to have Greek letters for Greek words, Cryllic letters for Russian words, and what not. Not to mention a number of attractive typographical symbols that Latin 1 does not have. Now, obviously, considering Dillo's target audience, it doesn't need full Unicode support. But it would be nice if it could handle basic UTF-8 encoded Latin-1 characters in web pages [1]. As an aside, I think Dillo currently has excellent CSS support (namely, none whatsoever). The thinking behing CSS is that a webpage should be perfectly readable on a browser with no CSS support; the content (if not the presentation) of a CSS webpage is perfectly accessible from a non-CSS browser. I come from the "CSS shouldn't be done at all, or should be done perfectly" camp. Considering the number of hacks web designers have come up to make CSS visible to this browser yet invisible to that browser [2], the last thing a webmaster wants is yet another buggy CSS implementation. - Sam [1] Some hacky code I once wrote, that reads UTF-8 from standard input, and outputs iso 8859-1: if(c<128) { printf("%c",c); } else { if(c < 0xe0) { /* two-byte sequence */ v = c & 0x1f; v <<= 6; c = getc(stdin); v = v + (c & 0x3f); printf("%c",v); } else { /* multi-byte sequence */ while(c & 0xc0 == 0x80 && !feof(stdin)) { c = getc(stdin); } } } (This code, FWIW, is public domain) It is probably better to simply use iconv. [2] http://centricle.com/ref/css/filters/ shows all of the common CSS hacks webmasters do these days.

3 2

[Dillo-dev]building 0.8.1 on slack4.0 (libc5)
by robert w hall June 14, 2004

June 14, 2004

I'm trying to compile dillo 0.8.1 to use on my old 486-based laptop (with slack 4.0, which is libc5-based). I've compiled glib + gtk 1.2.10 from source, but have no other libs like libjpeg, libpng etc. The binary starts to run then collapses in gdk error - error 11 in X-server (or similar message) Any advice? -- robert w hall

2 4

[Dillo-dev]Bug Meter suggestion
by Kyle Amon June 13, 2004

June 13, 2004

Suggestion: On of the bullet items under the "What can I do?" section of the new "About Dillo's Bug Meter" page (http://www.dillo.org/help/bug_meter.html) is ... * Help page authors and webmasters to provide standards-compliant contents (For instance, you can copy&paste the error messages produced by Dillo, and mail them to the author). While a nice idea in theory, almost no one is likely to do such a thing nearly as often as one encounters bugs in websites since it would entail one spend most of their browser time sanitizing the web rather than surfing it. Therefore, to facilitate such activity in a manner the user is likely to engage in, I suggest this be made an automatic dillo option (perhaps even the default behavior). In other words, have dillo automatically email these error messages to webmaster(a)wwwdomain.com so the user doesn't have to do so manually. Since computers are unbothered by the same monotony humans usually don't have much time for anyway, let the computer [via dillo] do the work here. At the very least, along side the W3C and WDG validator links in the Bug Meter, a link/exec to send such an email should be provided. If it isn't convienient (at the very least), people won't do it. -- Kyle P.S. I, for one, do about 75% of my browsing in dillo and would be happy to enable such a feature to nag lousy webmasters about their drift toward fascism in html if it were automatic. Beyond that, I might use a handy click/link/exec once in a while, but I doub't I'd ever send such a nag manually, much less often... and I'm pretty sure I'm not too unique in this regard. -- Kyle Amon email: amonk(a)gnutec.com url: http://www.gnutec.com/~amonk/ KeyID 1024D/4EB96E44 Fingerprint = E9EC 0046 8487 23D7 C91C D757 7B2A 8AE9 4EB9 6E44 "The free man is not anti-American, but anti-imperial. America [now] revisits the time of colonizers drunk on their superiority, convinced of their liberating mission, and counting on reimbursing themselves directly." -- Regis Debray This email Copyright 2004 by Kyle Amon, Inc., a Florida corporation, 12032 Colonial Estates Lane, Riverview, FL 33569-6894: +1 (813) 689-4041 amonk(a)gnutec.com http://www.gnutec.com/~amonk/. All rights reserved.

2 2

Re: [Dillo-dev]Some remarks on the UI
by Jorge Arellano Cid June 10, 2004

June 10, 2004

On Sun, 6 Jun 2004, Sebastian Geerken wrote: > Hi, > > I've just committed several changes, which fix the issues mentioned > here. Please test and report any bugs. If you are not sure, whether > you have detected a bug, it's certainly a misfeature, so report it > anyway :-) I reviewed the patch and commited some changes. Here go my comments: * I liked very much the new menu-title widget! * gcc showed some "missing initializer" warnings. It was the lack of the latest pointer in three GtkTypeInfo structures. Fixed. * The "." shortcut was not working. Fixed. * There was a warning for a signed/unsigned comparison, so I switched _GtkExtButton's pressed_button to guint. Fixed. * The "View Page Bugs" menu item for the page menu doesn't work. It went to a segfault once. NOT fixed. Some inside comments refer to unused code, for instance on Home and reload buttons. Yes, it was meant for future extensions (e.g. a reload images menu item). Now that there's a new way to code this (as with Back and Forward), the old way can be safely deleted. I changed the text for the tooltips of these buttons. Lots of mouses have two buttons and emulate the middle-click when both are pressed, so "right-click" and "middle-click" are a bit better than numbers in this case. It was nice to see the middle-click over "clear-url" button as a shortcut for paste. It'd be perfect if the entry can be focused at the same time. > > On Tue, Jun 01, Jorge Arellano Cid wrote: > > On Mon, 31 May 2004, Sebastian Geerken wrote: > > > 1. Bug meter: The bug meter button reacts on the "pressed" event, to > > > show the window, but it should react on the "clicked" signal. > > > Popping up a window after a "pressed" signal is usually only done > > > for volatile windows (like menus, or overview windows, like in > > > GIMP, which often vanish again, after releasing the mouse button). > > > > > > Reacting on "pressed" is always a bit problematic, also because the > > > pointer grab initiated with this event can cause side-effects. > > Fixed. > > > > Furthermore, it should not have a relief, compare to the "Hide > > > Controls" button. > > Also fixed, I've removed the relief. Yes, the toolbar icons and the "Hide controls" one are OK without a relief, but after watching how the bug-meter button looks without a relief, and how much it hints the user of being a button, I'd much prefer to have the relief back. Note that the "Hide controls" button has a drawed frame around that helps getting the idea of it being a button! Perhaps a more orthogonal way of doing it is to have a relief for both, but it's simpler to keep it as it was. > > > > 2. The menus behind the "Back" and "Forward" buttons: Consider the > > > following scenario: > > > > > > (i) The user presses button 3 on the "Back" button, and the menu > > > is shown. > > > > > > (ii) He keeps the button pressed, moves to one entry in the menu, > > > and releases the button. Nothing happens. > > > > > > Currently, the menu items only react on "button_press", which is > > > anyway wrong (see above). Furthermore, in this situation, the item > > > should already react on the "release" event, either as described > > > above, or in the way (i) click on menu, (ii) click on menu > > > entry. The current behaviour is that much confusing, that a user > > > may think that it does not work at all. (I did!) > > > > Yes. Sometime ago Ricardo Persichetti sent me a patch for this. > > It's just a couple of lines. Now on CVS for testing purposes, > > plus enhanced pixmaps for Back and Forward that hint their right > > click menus. > > I've written two new widgets, GtkExtMenu and GtkExtMenuItem, which > handle this in a cleaner way. Connecting to "button-release-event" has > several disadvantages, what I've seen until now, is: (i) keyboard > navigation was not handled, and (ii) the menu was hidden too late. > > The new widgets resemble the standard way as far as possible, so > problems shouldf be minimized. Good! > > > BTW, it is not very obvious, that there is a menu behind these two > > > buttons. Perhaps there should be, in the icon, a small arrow > > > directed downdwards, or so. > > > > Done! > > You forgot the small icons, this is fixed now. Also, I've added text > to the tooltips. Cute icons :) > > > 3. Attached is the incomplete code of a new widget, which changes the > > > behaviour of GtkButton: > > > > > > (i) It reacts also on button 2 and 3, this would make multiple > > > functions for one button possible. Actually, this feature > > > should be reserved for very few situation (since it is simply > > > confusing), namely the "Clear URL" button, which may also be > > > used for visiting the URL in the selection buffer. > > > > > > (ii) It provides a function for attaching a menu in a better way, > > > than we have currently, for the "Back" and "Forward" menus, > > > and for the bug meter. > > > > > > Simply copy all files into one directory, call "make", and run > > > "testprg". Look, what happens with different mouse buttons. > > > > Interesting. I went through the code a bit. > > > > It's not obvious to me how to modify them to be able to provide > > the same functionality (new window on middle click on history), > > and not to screw the design at the same time. > > Has been enhanced integrated (for forward, backward, and bug > meter). See comments in the code. Good work. It looks cleaner too. Cheers Jorge.-

2 2

[Dillo-dev]fixed font failed in dillo on NetBSD
by tman June 10, 2004

June 10, 2004

I updated packages from source dated June 7 including (amoung others): fontconfig-2.2.95 perl-5.8.4 glib-1.2.10nb6 gtk+-1.2.10nb7 glib2-2.4.2 gtk2+-2.4.2 atk-1.6.1 pango-1.4.0 x11-links-0.23 dillo-0.8.1 When I ran dillo after upgrading, no fonts displayed in the browser. Before updating, fonts displayed. In dillorc I changed vw_fontname from default heletica to lucida, and dillo now displays variable width font. I changed fw_fontname from default courier to fixed, and then lucidatypewriter, but cannot get fixed fonts to display. What busted?

1 0

[Dillo-dev]Dillo: Bug#568
by Jorge Arellano Cid June 9, 2004

June 9, 2004

Hi, There's: 568 Impact: soft error At: run time Type: render bug Comments: When a button tooltip is displayed, and you scroll the page with the keyboard, dillo fails to render the newly exposed content. This is the same behaviour reported for images in bug 289. HTReproduce: Open any page taller than the view area, place the mouse over a dillo control, when the tool tip appears, scroll the page with the keyboard. WorkedBy: nobody Submitter: dmckee*jlab*org in the bug track. I can't reproduce it, so maybe it's a GTK problem. What version are you using? (BTW dillo 0.8.1 or CVS is assumed). I tried several pages, but for instance, going to our home page, making the tooltip for bookmarks appear, and scrolling with the keyboard works OK for me. Can anyone reproduce this bug? Cheers Jorge.-

4 3

[Dillo-dev]patch for bug 211
by Diego Sáenz June 9, 2004

June 9, 2004

Hello. Here i send a patch for the dillo bug 211 (remember save dir on save and open dialogs) based on a pactch for an old version of dillo (0.6.¿?). it was made for 0.8.1 but apply in current cvs too. Each dillo window remember its own save dir. Sorry if this is not the right place to send patchs. DarkSpirit/Diego Sáenz

2 3

[Dillo-dev][전체회신] Re: [Dillo-dev]Downloads File
by vinod＠etri.re.kr June 9, 2004

June 9, 2004

Hi Jorge > Hi all Hi Vinod. I assume you received my email, and after considering, you decided to work on the GUI dpi for downloads. (please confirm me that) # Yes. I did intend to confirm after reply from mailing list regarding work. > I am pretty new to Dillo and would like to work on the downloads file. > [dpi] Please could anyone guide me on the present structure and what > enhancement or modifications to be made ? > > The overview of the current structure : > > 3 Structures - DownloadFile, DownloadStatus and DownloadWindow > > Api's to start and free new download, update status, list and create log, > get download size and time, add download to list and start and close > download. Also, api's for read socket and print list are provided. > > Can anyone tell me how to enhance this structure better ? There're two problems with the GUI dpi as is: 1.- it doesn't work reliably 2.- it's too complex for the things it does So, the work basically is to play with it a bit to know what features are provided (the new one should provide the same), simplify the internal representation and algorithms (the GUI part, the part of the code that comes from the original dpi needs not to be touched), and finally debug and perform stress tests to achieve reliability. # Thanks Jorge, Will peruse through the code this weekend. Cheers Jorge.- Regards Vinod Joseph _______________________________________________ Dillo-dev mailing list Dillo-dev(a)lists.auriga.wearlab.de http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

2 1

[Dillo-dev]Using alternate char for TABBING
by Deon Stoltz June 9, 2004

June 9, 2004

I have recently discovered dillo and am delighted! Very fast and lightweight. We want to use if in a settop box application (thus the lightweight requirement). We need to interface to a IR remote control device so no keyboard are available. Web pages can normally be navigated with the TAB key. In our application we need to associate the TAB key with another key eg: ctrl-A. Do somebody know how, where to do this? I had a look in the code but I am not that strong on c++. ANY help will be appreciated! Thanks Deon

1 0

[Dillo-dev]Content outside table border
by Jorge Arellano Cid June 9, 2004

June 9, 2004

Hi Sebastian, Just found this page: http://www.757.org/~joat/blog/ look at the small calendar! Best Jorge.-

1 0

[Dillo-dev]Some remarks on the UI
by Sebastian Geerken June 6, 2004

June 6, 2004

Hi, several remarks on the UI, which should be fixed IMO. 1. Bug meter: The bug meter button reacts on the "pressed" event, to show the window, but it should react on the "clicked" signal. Popping up a window after a "pressed" signal is usually only done for volatile windows (like menus, or overview windows, like in GIMP, which often vanish again, after releasing the mouse button). Reacting on "pressed" is always a bit problematic, also because the pointer grab initiated with this event can cause side-effects. Furthermore, it should not have a relief, compare to the "Hide Controls" button. 2. The menus behind the "Back" and "Forward" buttons: Consider the following scenario: (i) The user presses button 3 on the "Back" button, and the menu is shown. (ii) He keeps the button pressed, moves to one entry in the menu, and releases the button. Nothing happens. Currently, the menu items only react on "button_press", which is anyway wrong (see above). Furthermore, in this situation, the item should already react on the "release" event, either as described above, or in the way (i) click on menu, (ii) click on menu entry. The current behaviour is that much confusing, that a user may think that it does not work at all. (I did!) (The implementation would become a bit tricky, since button 2 will pop up the page in a new window. Perhaps, I'd even suggest to skip this feature, if a clean solution is not possible.) BTW, it is not very obvious, that there is a menu behind these two buttons. Perhaps there should be, in the icon, a small arrow directed downdwards, or so. 3. Attached is the incomplete code of a new widget, which changes the behaviour of GtkButton: (i) It reacts also on button 2 and 3, this would make multiple functions for one button possible. Actually, this feature should be reserved for very few situation (since it is simply confusing), namely the "Clear URL" button, which may also be used for visiting the URL in the selection buffer. (ii) It provides a function for attaching a menu in a better way, than we have currently, for the "Back" and "Forward" menus, and for the bug meter. Simply copy all files into one directory, call "make", and run "testprg". Look, what happens with different mouse buttons. Comments welcome. Sebastian

2 6

[Dillo-dev]UTF-8
by Jorge Arellano Cid June 4, 2004

June 4, 2004

Hi guys! I just added a small hack to view UTF-8-encoded quotation marks. Yes, it's a mistery to me why someone would want to serve pages in english, encoded in UTF-8 instead of Latin1. ...but you may find some as this one: http://www.seattleweekly.com/features/0422/040602_news_microsoft.php Please let me know if you know of more pages like this one, or if you know the rationale for preferring UTF-8 over latin1 when the latter is enough. Cheers Jorge.-

2 1

[Dillo-dev]very simple question, and many kudos!
by Brian Hechinger June 3, 2004

June 3, 2004

first off, i would like to say that i am completely impressed with dillo's efficiency. i finally found myself to be in the position where i needed to get X11 onto my Toshiba Tecra notebook so that i could use a web browser. it's a 166Mhz with 32M ram, so i asked around what would be the best browser for such a low resource machine, and was told firefox by a few. so i got a copy of firefox. it was so slow, not only was it unusable, it made the rest of my computer unusable. then someone mentioned dillo. so i got it. i have to say, i can't even put my surprise of how fast it was into words. i was floored. this is EXACTLY what i needed. so far, it's perfect for me. the only thing that i have a question about is PgUp and PgDn only moving half a page at a time. is there a way to get them to move a full page? thanks, and keep up the excellent work!! -brian -- "Performing a Google search on the fairly unique name you've chosen weeks ago for a piece of software you're writing, only to discover that the sole occurance of that name is as a pen-name for someone who spends his spare time writing X-rated PowerPuff Girls fan fiction is pretty-much a buzzkill." -- Jonathan Patschke --

5 10

[Dillo-dev][bug] ppc64 crash on right-click / bug tracking system
by Thomas Cort June 3, 2004

June 3, 2004

Hi, I ran into a problem where dillo crashes when I right click on any part of a webpage. A full bug report is below this paragraph. When I went to use the bug tracking system I noticed the "Max 256 chars" notice. Is it possible to get this increased? I couldn't get my bug report into <= 256 characters. Dillo 0.8.1 on GNU/Linux (yellowdog 3.0) ppc64 (Apple G5) configure options: --prefix=/home1/students/tcort To reproduce bug: ssh with X11 forwarding and compression turned on, start dillo, and right click on any part of the webpage. Notes: I don't have local access to this machine, so I don't know if it's an X11 forwarding bug or a ppc64 specific bug. Browsing webpages works fine as long as you don't right click. Using other programs with X11 forwarding from this machine, such as gimp, work fine. Output: tcort@pepsi tcort $ ssh -X -C turing.ubishops.ca tcort(a)turing.ubishops.ca's password: [tcort@turing tcort]$ bin/dillo dillo_dns_init: Here we go! Disabling cookies. Nav_open_url: Url=>about:splash< Gdk-ERROR **: BadWindow (invalid Window parameter) serial 2306 error_code 3 request_code 38 minor_code 0 Gdk-ERROR **: BadAccess (attempt to access private resource denied) serial 2307 error_code 10 request_code 102 minor_code 0 [tcort@turing tcort]$

2 1

[Dillo-dev]https (eg for posting to ps2 bulletin boards)
by robert w hall June 2, 2004

June 2, 2004

Every time I read the playstation2-linux lists I realise how very useful dillo will be when it gets just a little further. How's the short/medium term future of https looking?(*) If an official plugin is in the offing then that's probably what we should all wait for. Otherwise, is the patch from 0.7 resurrectable?? Bob (*You need ssl to post to the ps2 bulletin board) -- robert w hall

3 4

[Dillo-dev]dillo as a debian package ..?
by Jorge Arellano Cid June 1, 2004

June 1, 2004

Hi! It's been a long time since the debian package for Dillo is not updated. Since 0.7.3 to be accurate. :-( I've contacted Phil and, surprisingly, the only reason for them not being made is lack of time. Is here a Debian developer that can take the burden, or help Phil, or at least provide us with an x86 .deb to link to from the download page? It seems important because Debian lingering on 0.7.3 implicitly passes the (wrong) message of the latter versions of dillo not being worth an upgrade. TIA Jorge.- ---------- Forwarded message ---------- From: Jesper Olsen <__lazer __at phreaker__ net__> To: jcid(a)dillo.org Subject: dillo as a debian package ..? hello there .. Is there somewhere around where i can get dillo-0.8.1 as a .deb file?? i can't find it, i've almost looked everywhere :s MaNN, thank you for a wonderful browser .. it runs top stable and fast on my 166mhz .. well, it is a shame that i can't get the new versions of the browser ! DILLO ROCKS!!!!! - lazer

3 5

Re: [Dillo-dev]Some remarks on the UI
by Nicola Girardi June 1, 2004

June 1, 2004

This is what I replied privately to Sebastian: > BTW, it is not very obvious, that there is a menu behind these two > buttons. Perhaps there should be, in the icon, a small arrow > directed downdwards, or so. In fact I didn't ever discovered there was such (useful) menu. :-( > (ii) It provides a function for attaching a menu in a better way, > than we have currently, for the "Back" and "Forward" menus, > and for the bug meter. Another menu I didn't catch, the bug meter menu. :-) > Comments welcome. Very nice. :-)

1 0

Re: [Dillo-dev]page increment (was: very simple question, and many kudos!)
by Sebastian Geerken May 31, 2004

May 31, 2004

On Mon, May 31, Brian Hechinger wrote: > On Mon, May 31, 2004 at 03:09:03PM +0200, Sebastian Geerken wrote: > > > > Scrolling half a page is actually standard behaviour of Gtk, look > > e.g. at gtk_layout_size_allocate() (in gtklayout.c), which is more or > > less responsible for this in our case, since our code uses > > GdkAdjustment::page_increment, which is set there. > > hmmmmmm. can't say i care for that behavior. maybe i should whine to the GTK > folks and see if i can either get that changed, or at least get it setup so > there is some way to override it at run/compile time? They won't probably change it, especially since dillo depends on Gtk 1.2, which development has been closed, AFAIK. > > The attached patch "solves" it, but I'm not sure, wether the standard > > behaviour should be overridden. > > how hard would it be to make it a runtime option? something in dillorc? that > way you can have standard behavior and also the option to change its behavior > to satisfy people like me? Actually, it envolves reading Havoc Pennington's essay "Free software and good user interfaces" (http://ometer.com/free-software-ui.html) especially the section "The Question of Preferences", and considering, whether or not this configuration option should be added. Actually, I'd rather like it either having applied or not (not configurable), the others should apply the patch. (Adding the code is BTW rather simple, just look at how other options are handled in "prefs.c".) > how about this. let me know when the CVS tree stablizes and i'll write that > for you. It will probably still take some time, it is a bit more difficult than I thought. Sebastian

1 0

[Dillo-dev]helper applications and their uses
by Randy May 25, 2004

May 25, 2004

Hello, First of all,let me say that i don't know a command-line from a chorus line-my feeble knowlege of Linux consists mainly of downloading Apt and Synaptic for my RH7.3 and FC1 distros,and handling it all from the GUI. Just letting you know that your'e dealing with a novice here. : ) I like the concept behind Dillo,simple and fast.I understand that you might be working on a version for frames later on,which would be even better of course.My question though-without reliance on the usual Mozilla plug-ins(which Galeon,Epiphany and Firefox also use) is there anything I can download as a helper application? Would the full Java,for example,interact at all with Dillo when browsing the Web? Would Realplayer,or Gxine with Real plug-ins(i'm on Gnome,of course). Thanks in Advance, Randy

2 1

[Dillo-dev]Dillo RPM available for Yellow Dog Linux 3.0
by Kelson Vibber May 24, 2004

May 24, 2004

In case anyone's interested, I posted an RPM for Yellow Dog Linux 3.0 over the weekend. The usual place: http://www.hyperborea.org/software/dillo/ -- Kelson Vibber www.hyperborea.org

1 0

[Dillo-dev]0.8.1 text centering error? (fwd)
by Jorge Arellano Cid May 23, 2004

May 23, 2004

---------- Forwarded message ---------- Date: 18 May 2004 04:15:11 PDT From: Gregg <dude(a)scorpiorising.ca> To: Dillo <dillo-dev(a)lists.auriga.wearlab.de> Subject: [Dillo-dev]0.8.1 text centering error? Hello, I have compiled and run Dillo 0.8.1 flawlessly on Debian woody. However, I notice the browser "centering text" that wasn't on 0.8.0 and earlier. Example: http://www.cknw.com - my daily news, now shows the headlines as centred. Is this supposed to be? Thanks, -- Gregg _______________________________________________ Dillo-dev mailing list Dillo-dev(a)lists.auriga.wearlab.de http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

1 0

[Dillo-dev]Re: White spaces handling (was: Weird glitch with rendering)
by Jorge Arellano Cid May 23, 2004

May 23, 2004

Hi Sebastian, > Just FYI: I remember from the old discussion, that DwPage originally > collapsed spaces (i.e. calls to a_Dw_page_add_space), and that I > suggested some changes, so that this is not done anymore. Your patch > would reverse this change again. > > The discussion goes attached. I read the whole thread you sent me. BTW, the whole whitespace issue re-started when I found one of the patches you mentioned. I then read the past emails, and started to try to follow. It is not easy to make a clean picture of it though! --This mainly because of the way it used to be handled inside Dillo and because the SPECS seem not definitive in this matter. Here I'll try to cite past emails and comment my current thoughts: > [...] > This is because a_Dw_page_add_space is called twice, and this function > does actually not add a space, but change the current one. I've worked > on two solutions: > > 1. The first patch (underlined-spaces-clean.diff) changes the > behavior of a_Dw_page_add_space, but needs some changes in the > HTML parser, to control better when this function is called, > i.e. to ignore spaces after <A> and before </A>. This one seems the best candidate. With regard to fixing the parser for ignoring these spaces, see my comments below. > 2. The second one (underlined-spaces-kludgy.diff) tries to adjust > spaces, depending on rather hairy conditions. It works already > (if I'm not wrong, it has quite the same results as the code > before, except the bug). This option is no longer valid. As you clearly described: <q> > > > 2. The second one (underlined-spaces-kludgy.diff) tries to adjust > > > spaces, depending on rather hairy conditions. It works already > > > (if I'm not wrong, it has quite the same results as the code > > > before, except the bug). > > As I've now noticed, this will not work for CSS, e.g. the following > code: > > One non-underlined word > > will be displayed: > > One non-underlined word > ---- ----- > > just because this patch lets DwPage assume something about the > document structure (change from non-underline to underline == > beginning of a tag) which it can (and should) not. > </q> Now, > > Despite of the file names, I'm not sure if the changes in the HTML > parser can be done cleanly. So, if you think that this is difficult to > realize, apply the second patch. Especially, there is probably also an > other DwPage function necessary to remove again the last space, when > </A> is read after a space. I believe the parser is not very hard to modify for ignoring the spaces as patch 1.- requires. The problem is that the SPEC is not clear about exactly how these spaces should be collapsed. For instance: > A different case is "Some text". Your patch will make > "Some text" of it, but it should be really be > "Some text." Yes, I agree, "collapsing" here should be: 'Some text' => 'Some text' as you note. but what do we do with this: 'Some text' If we ignore white space after the start tag and before the end tag, it becomes 'Sometext' (with no space at all!) If we "collapse" as the SPEC says should be done, we have two possibilities: 'Some text' (as it was: underline the whitespace) and 'Some text' (move the space out of the tag) AFAICT, the SPEC leaves the choice open, and advices HTML authors against whitespace inside the tags. IMO, always collapsing white space after the start tag and before the end tag is the simplest to implement. Even more, as the SPEC doesn't define what to do in this case, it's an option left to the User Agent: <q source='HTML4.01 SPEC, 9.1'> In order to avoid problems with SGML line break rules and inconsistencies among extant implementations, authors should not rely on user agents to render white space immediately after a start tag or immediately before an end tag. Thus, authors, and in particular authoring tools, should write: We offer free <A>technical support</A> for subscribers. and not: We offer free<A> technical support </A>for subscribers. </q> Now, this solution would also account for the special SGML line break rules: <q source='HTML-4.01 SPEC B.3.1'> SGML (see [ISO8879], section 7.6.1) specifies that a line break immediately following a start tag must be ignored, as must a line break immediately before an end tag. This applies to all HTML elements without exception. The following two HTML examples must be rendered identically: Thomas is watching TV. Thomas is watching TV. So must the following two examples: <A>My favorite Website</A> <A> My favorite Website </A> </q> Note that Firebird doesn't follow the SGML line break rules! This is not to say we should follow, but that we may find some buggy pages out there, but at least at this point the SPEC is quite clear. ;) > I'm not completely sure, we should carefully evaluate the specs, I'd > suggest to delay this for the 0.8.2 release. Neither I am sure! :-) At this point I see that applying patch 1, plus making the parser ignore whitespaces after a start tag and before an end tag is a correct solution that can endure a reality test. I'll start coding it. Please let me know your thoughts or any point I'm still missing. Best Jorge.- PS: This work is for 0.8.2.

2 9

[Dillo-dev]BugTrack
by Jorge Arellano Cid May 22, 2004

May 22, 2004

Hi there! Dillo's bugtrack was just updated so email addresses are mangled with images to protect from SPAM robots. The "submitter" camp was also added, so we can contact the poster for details. Cheers Jorge.-

1 0

[Dillo-dev]javascript links
by Jorge Arellano Cid May 21, 2004

May 21, 2004

Hi there, The CVS is just updated for 0.8.2-pre. The first patch is a tiny handler for javascript links. It's very simple, it will try to make a normal link from a javascript-based href. Please test it and tell me how it works. Cheers Jorge.-

1 0

[Dillo-dev]White space handling
by Sebastian Geerken May 21, 2004

May 21, 2004

Hi! This mail is just to collect what I have found in the specs, and how I understand it. I'll comment Jorge's mails seperately. Sebastian. ---------------------------------------------------------------------- HTML ---- Unfortunately, I do not have the SGML spec at hand, so I only refer to the parts of the HTML spec, which refer to general SGML topics. There is something in chapter 6, although I did not find anything about how #PCDATA (which is used in the DTD for the texts, which are relevant in this context), only about #CDATA: > 6.2 SGML basic types > > The document type definition specifies the syntax of HTML element > content and attribute values using SGML tokens (e.g., PCDATA, CDATA, > NAME, ID, etc.). See [ISO8879] for their full definitions. The > following is a summary of key information: > > * CDATA is a sequence of characters from the document character set > and may include character entities. User agents should interpret > attribute values as follows: > > o Replace character entities with characters, > o Ignore line feeds, > o Replace each carriage return or tab with a single space. > > User agents may ignore leading and trailing white space in CDATA > attribute values (e.g., " myval " may be interpreted as > > "myval"). Authors should not declare attribute values with leading > or trailing white space. > > For some HTML 4 attributes with CDATA attribute values, the > specification imposes further constraints on the set of legal > values for the attribute that may not be expressed by the DTD. About HTML: > 9.1 White space > > [...] > For all HTML elements except PRE, sequences of white space separate > "words" (we use the term "word" here to mean "sequences of non-white > space characters"). When formatting text, user agents should > identify these words and lay them out according to the conventions > of the particular written language (script) and target medium. > [...] This is, what is currently the difference between calls to a_Dw_page_add_text() (which adds a word), and a_Dw_page_add_space(), which adds a "word separation". > B.3.1 Line breaks > SGML (see [ISO8879], section 7.6.1) specifies that a line break > immediately following a start tag must be ignored, as must a line > break immediately before an end tag. This applies to all HTML > elements without exception. > > The following two HTML examples must be rendered identically: > > Thomas is watching TV. > > > Thomas is watching TV. > > > So must the following two examples: > > <A>My favorite Website</A> > > <A> > My favorite Website > </A> Notice this: "This applies to all HTML elements without exception." That is, also for <pre>. But it does not say anything about white spaces in general. XHTML ----- First of all from the XML 1.1 spec: > 2.10 White Space Handling > > In editing XML documents, it is often convenient to use "white > space" (spaces, tabs, and blank lines) to set apart the markup for > greater readability. Such white space is typically not intended for > inclusion in the delivered version of the document. On the other > hand, "significant" white space that should be preserved in the > delivered version is common, for example in poetry and source code. > > An XML processor MUST always pass all characters in a document that > are not markup through to the application. A validating XML > processor MUST also inform the application which of these characters > constitute white space appearing in element content. First of all, we do not have a distinction between the (basic) XML processor, and the application. (In the CSS prototype, this is clearer, and because it is always a good decision to make the implementation structure resemble the structure of the specifications, this should be kept in mind.) How I understand this basicly: It is (in most cases, see below) the task of the application (in this case XHTML) to determine how to deal with whitespaces. Anyway, I did not find anything in the XHTML 1.0 specification, so I believe that the same rules apply, as for HTML 4.01. > A special attribute named xml:space MAY be attached to an element to > signal an intention that in that element, white space should be > preserved by applications. In valid documents, this attribute, like > any other, MUST be declared if it is used. When declared, it MUST be > given as an enumerated type whose values are one or both of > "default" and "preserve". For example: > > <!ATTLIST poem xml:space (default|preserve) 'preserve'> > <!ATTLIST pre xml:space (preserve) #FIXED 'preserve'> > > The value "default" signals that applications' default white-space > processing modes are acceptable for this element; the value > "preserve" indicates the intent that applications preserve all the > white space. This declared intent is considered to apply to all > elements within the content of the element where it is specified, > unless overridden with another instance of the xml:space > attribute. This specification does not give meaning to any value of > xml:space other than "default" and "preserve". It is an error for > other values to be specified; the XML processor MAY report the error > or MAY recover by ignoring the attribute specification or by > reporting the (erroneous) value to the application. Applications may > ignore or reject erroneous values. If you look at the DTDs of XHTML 1.1, you'll find three elements setting this: <style>, <script>, and <pre>. As to the patch: <style> and <script> are currently rather unimportant, while for <pre>, the HTML parser switches into the DILLO_HTML_PARSE_MODE_PRE mode, where white spaces are handled differently. > The root element of any document is considered to have signaled no > intentions as regards application space handling, unless it provides > a value for this attribute or the attribute is declared with a > default value.

1 0

[Dillo-dev]Tabs/Frames/Kbnav
by Jorge Arellano Cid May 21, 2004

May 21, 2004

Hi, Now that dillo 0.8.1 is done, I want to clarify what happened with the Tabs/Frames/KeybNav patch. Sometime has passed since, but at the time we decided to not stir an already ungrateful situation. Here I'm forwarding a couple of emails that should be enough to do it (dillo-dev archives holds the rest). Sincerely Jorge.- -------------------------------------------------------------------------- From akemnade(a)informatik.uni-bremen.de Mon May 17 08:44:51 2004 Date: Sat, 27 Mar 2004 14:25:31 +0100 From: Andreas Kemnade <akemnade(a)informatik.uni-bremen.de> To: Sebastian Geerken <s.geerken(a)ping.de> Cc: jcid(a)dillo.org Subject: Re: [Dillo-dev]Frame support in CVS Hi Sebastian, Hi Jorge, On Sun, 7 Mar 2004 22:14:49 +0100 Sebastian Geerken <s.geerken(a)ping.de> wrote: > Hi Andreas. > > (I'm writing this in English to make forwarding etc. simpler.) > > Jorge and I would indeed like to incorporate Frank's work into the > main trunk, but unfortunately Frank refused to clean up the code, and > finally, he quit the project, and plans to make an own version of > dillo. I've a mail of 8 Jan, which I can forward you, if you are > interested. > Please send me that mail. Ok, perhaps I was not deep enough in the archives. I just did not find any statement about what has to be done with the patch. I think other readers of dillo-dev do not know about that, too. Perhaps it would be better if that was more clarified but I also understand that you do not want to repeat yourself too often. > If you want to commit some work, you may clean up the patch, > especially split it up into several parts. Most important, there are > three parts: DilloDoc (which I'd like to be renamed), frames, and > tabs. They are of course related, but it would be handy to have single > patches for them (because of the over-linear correlation of patch size > and review time). > Most the time I was just happy to be able to apply the frames patch. I was not forced to look at the code. But with the massive conflicts the patch produces when I applied it to 0.8 I really had to start to look at the patch. Sometimes being not a maintainer has really big advances... If I checked the patch size before fixing the conflicts I guess I would have given up. I have never noticed before how big the patch was. If splitting does not work, then there is something really wrong with the code. So let's divide and conquer. I think an important part is to remove nice (but not very important) features from the patch which can be added separetely later on. I think I can find time in the next month to split and clean up the frame patch. Greetings Andreas Kemnade -------------------------------------------------------------------------- From s.geerken(a)ping.de Mon May 17 08:44:37 2004 Date: Sat, 10 Jan 2004 13:04:11 +0100 From: Sebastian Geerken <s.geerken(a)ping.de> To: Jorge Arellano Cid <jcid(a)dillo.org> Subject: Fwd: Re: Notes on keyboard navigation FYI. [ Part 2: "Included Message" ] Date: Thu, 08 Jan 2004 22:31:10 +0100 From: Frank de Lange <frank(a)unternet.org> To: Sebastian Geerken <s.geerken(a)ping.de> Subject: Re: Notes on keyboard navigation Sebastian Geerken wrote: >Hi Frank. > >First of all, a happy new year! > >Several notes on keyboard navigation: > >1. Jorge and I decided, that we do not immediately support keyboard > configuration, also because it is rather unclear what widget > toolkit will be used in the future. So the version of keyboard > navigation, which will be integrated the first time, should have > hard-coded keys. > >2. Some days ago, I committed some changes to iterators, you may be > interested in. There is some description within doc/Dw.txt, section > "Anchors and Scrolling", subsection "Scrolling". Especially, > scrolling has now two dimensions, there is a new scrolling > position, DW_[HV]POS_INTO_VIEW, and DwIterator::scroll_to has been > replaced by the more general DwIterator::get_allocation. > >3. I noticed some performance problems with stress tests, I'm not sure > they play a role in real life. Anyway, you wrote on the list that > you are already working on in. > >Sebastian > > Happy new year as well... and thanks for the notice. For keyboard navigation to work in a sensible way x coördinate support is useful... so useful that I had already implemented it. The performance problems have been solved in my tree as well, as I have stated in my last message to the list. They are caused by the use of g_list_search_custom to find the focusable in focus.c. Using a hash table solved this. But... I have decided to discontinue support to the Dillo project as of the 'infringer' message to the list, so I will no longer post or update patches to Dillo. Instead, I am working on a new browser which should fill the same niche as Dillo (small, fast, enough features for most sites so it can be used as main browser). Starting point for this is my patched version of Dillo, with additional changes: - dpi and dpid and related things are gone (DONE) - library plugins for everything dpi did, and more (mime decoders, schemes, etc) (DONE) - ccc is on the way out, to be replaced by something less convoluted and more general (WORK IN PROGRESS) - DOM support with new xml parser (WORK IN PROGRESS) The project currently uses GTK+ as toolkit, but might switch to something else (FLTK is a candidate). A side goal for this project is to be developer-friendly. Patches are welcome, they will get their own space on the site (which will be on savannah, which is still not completely recovered from the break-in yet... hence the delay). I may post a short notice to the list if/when the project is up and ready for testing. Good luck with Dillo. 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

[Dillo-dev]Re: White spaces handling (was: Weird glitch with rendering)
by Jorge Arellano Cid May 19, 2004

May 19, 2004

Hi Sebastian, One thing that puzzles me about the patch that collapses whitespace at the parser level, is that some tables no longer are centered! For instance: http://www.dillo.org/stats/ http://www.dillo.org/stats/usage_200405.html Cheers Jorge.-

1 0

[Dillo-dev]0.8.1 text centering error?
by Gregg May 18, 2004

May 18, 2004

Hello, I have compiled and run Dillo 0.8.1 flawlessly on Debian woody. However, I notice the browser "centering text" that wasn't on 0.8.0 and earlier. Example: http://www.cknw.com - my daily news, now shows the headlines as centred. Is this supposed to be? Thanks, -- Gregg

1 0

[Dillo-dev]dillo-0.8.1 was officialy released.
by Jorge Arellano Cid May 15, 2004

May 15, 2004

Hi folks, The new dillo-0.8.1 is just released now, waiting to appear on the freshmeat front page. It's been a long road 'til here, so please let me thank all of you that have participated in dillo-dev. Now, let's party! ;-) Jorge.-

1 0

Re: [Dillo-dev]dillo-0.8.1.tar.bz2 final
by Jorge Arellano Cid May 15, 2004

May 15, 2004

Hi there, Has anyone tested on *BSD yet? and Solaris? Cheers Jorge.-

3 2

Re: [Dillo-dev]dillo-0.8.1.tar.bz2 final
by Jorge Arellano Cid May 14, 2004

May 14, 2004

On Fri, 14 May 2004, Maarten Sander wrote: > Hi, > > > It's a simple test, if you have a working dillo binary after > > 'make', it's OK. That's all. > > > > Please don't be shy and report success on your platform. If > > nobody reports problems before Kelson is done (or Saturday), it > > will be released. ;-) > > Well, it works overhere :) > > I've just compiled and used Dillo on my Fedore Core 1 (Athlon-)machine. > It's updated with all the latest and greatest (*cough*) packages from > Dag, ATrpms, etc. GCC is 3.3.2, glibc 2.3.2, GTK 1.2.10. During > ./configure it gave a big warning that it could not find libpng-config, > but it detected png.h after that and compiled without problems. libpng is not easy to detect, maybe the Fedora team decided to go back to the simple header/library pair to avoid problems. Don't worry. You have a quite current GNU/Linux. Can you tell whether Dillo was linked against NPTL? Cheers Jorge.-

2 2

[Dillo-dev]dillo-0.8.1.tar.bz2 final
by Jorge Arellano Cid May 14, 2004

May 14, 2004

Hi there, It's being a long period of testing. As I don't want to delay the next release anymore, the final tarball is there, ready for download! It is not yet announced, probably on Staurday, after Kelson is ready with the RPMs. The final tarball is not exactly the same as "pre2", it comes with a few fixes: - A fix on iterartors (on a very rare critical warning). - A fix to gray up the stop button on downloads (1 line fix). - A new link in the splash, magazines section. - A fix on pthreads detection code. The last one _NEEDS_ testing, on every machine! I may have said only on non Linux-based ones, but the bug was detected on an old libc5 based GNU/Linux! It should work because it more or less leaves the settings as they were before, but with the enhancement of not linking pthread to every binary. It's a simple test, if you have a working dillo binary after 'make', it's OK. That's all. Please don't be shy and report success on your platform. If nobody reports problems before Kelson is done (or Saturday), it will be released. ;-) Good Luck! Jorge.- PS: Please note that this release shifts the state from alpha to "very stable beta". PS2: http://www.dillo.org/download/dillo-0.8.1.tar.bz2

2 3

Re: [Dillo-dev]dillo-0.8.1.tar.bz2 final on ps2
by Jorge Arellano Cid May 14, 2004

May 14, 2004

On Fri, 14 May 2004, robert w hall wrote: > works OK on playstation2-linux with a 2.2.21 kernel (from xrhino). > I haven't tried it on the default (sony) 2.2.1 kernel (kids seem to have > deleted it!) - but dimly recollect pthreads is a library not a kernel > feature, (so it shouldn't matter???) POSIX threads is one of "those" complex matters. It depends on several things including the kernel and glibc. To add flavour, Linux comes/came with linux threads, which is close to POSIX but not fully. The New POSIX Threads Library (NPTL) is POSIX compatible, and requires special kernel threading (as of 2.5.36), glibc 2.3 and other bindings (RedHat 9 was going to be the first to provide such environment). Furthermore, there are several different threading libraries among the platforms we support, so go figure! You can read much more than you wanted to know here: http://lwn.net/Articles/10465/ http://people.redhat.com/drepper/nptl-design.pdf As you may have begun to suspect after this short exposition, it's hard to know, and certainly I don't know, whether it will also run with Sony's 2.2.1 kernel! :-) Regards Jorge.-

1 0

[Dillo-dev]pre release
by Frank W McCormick May 13, 2004

May 13, 2004

Hi all Compiled and running the pre on Linux 2.4.24. No problems yet, but one thing that's been bothering me since Dillo was given DL capabilities....right clicking on a direct link gives you the DL but were it not for the eth0 indicator in ICEWM I wouldn't have any indication that DL was going ahead. Is this normal ? -- -------------- Frank McCormick Montreal --------------

2 1

[Dillo-dev]Re: null byte in HTML
by Jorge Arellano Cid May 13, 2004

May 13, 2004

---------- Forwarded message ---------- Date: Thu, 13 May 2004 12:33:13 +0300 (EEST) From: Jukka K. Korpela <jkorpela(a)cs.tut.fi> To: Jorge Arellano Cid <jcid(a)dillo.org> Subject: Re: Is the null byte allowed in HTML? On Wed, 12 May 2004, Jorge Arellano Cid wrote: > I can't yet found whether the null byte character is allowed in > HTML. Can you shed some light on this? It is not. You could use http://validator.w3.org to check for disallowed characters (it reports "non SGML character number 0"), but the ultimate reference is a) for HTML 4, the SGML declaration http://www.w3.org/TR/html4/sgml/sgmldecl.html where UNUSED effectively means 'disallowed' b) for XHTML, the XML specification, see http://www.w3.org/TR/REC-xml/#charsets which say, among other things, that all characters below 9 (HT) are disallowed. Thanks for a good question - I'm just finalizing a book on XHTML (in Finnish, sorry) and I realized that I had forgotten to discuss the character issue in sufficient detail. (I just realized that various generators may produce data with control characters.) -- Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/

2 1

[Dillo-dev]null byte in HTML
by Jorge Arellano Cid May 12, 2004

May 12, 2004

Hi there! Does anyone know whether a null-byte ('\0') is allowed inside HTML or SGML or XML? Cheers Jorge.-

1 0

[Dillo-dev]download problem on OpenBSD
by Riccardo Mottola May 11, 2004

May 11, 2004

Hello, using pre2 I went to sourceforge do download a file. On the mirror page I selected one and clicked on the "here" message for the non-stadard refresh which gives a direct link to the file (I could use it for d/l with wget). Clicking on it brought up a file location requester, I selected my home directory and gave ok. Dillo crashed with: Nav_open_url: Url=>http://prdownloads.sourceforge.net/ayttm/ayttm-0.4.6.tar.gz?use_mirror=hean…< Connecting to 66.35.250.217 Nav_open_url: Url=>http://heanet.dl.sourceforge.net/sourceforge/ayttm/ayttm-0.4.6.tar.gz< Dns_server [0]: heanet.dl.sourceforge.net is 0x721cb8 Connecting to 193.1.219.87 HTTP warning: Unhandled MIME type: <application/x-gzip> dpi.c:71: a_Dpi_rd_dpi_socket_dir: fopen : No such file or directory - /home/multix/.dillo/dpi_socket_dir debug_msg - init_sockdir: The dpi_socket_dir file does not exist dpid started Riccardo

2 2

Re: [Dillo-dev]dillo interface and motif, athena etc
by Riccardo Mottola May 11, 2004

May 11, 2004

david feldman wrote: > > I think your questions are very good, I hope you get good answer! > unfortunately there was no follow up. I wonder if anyone else beside you was interested. -R

2 1

[Dillo-dev]BUG#522
by Jorge Arellano Cid May 10, 2004

May 10, 2004

Hi, Can anyone reproduce BUG#522? 522 Impact: soft error At: run time Type: crash Comments: Url=>http://www.rti-zone.org/dillo/Html.testsuite/jorge/c2.html< Gdk-ERROR **: BadValue (integer parameter out of range for operation) serial 4689 error_code 2 request_code 45 minor_code 0 HTReproduce: debian unstable WorkedBy: nobody Cheers Jorge.-

2 1

Re: [Dillo-dev]BUG#522
by Jorge Arellano Cid May 10, 2004

May 10, 2004

On Mon, 10 May 2004, Riccardo Mottola wrote: > Jorge Arellano Cid wrote: > > > > Hi, > > > > Can anyone reproduce BUG#522? > > [...] > > No I cannot reproduce it on solaris. > > I compiled dillo on Solaris, UltraSparc too (alyazs solaris 2.6, gcc > 3.3) You might add this to the compatibility list too. So we have: Solaris 2.6, gcc 3.3, Sparc v8 processor (=32bit) Solaris 2.6, gcc 3.3, UltraSparc (64bit?) What SunOS do they report? Cheers Jorge.-

1 0

[Dillo-dev]dillo-0.8.1-pre2.tar.bz2
by Jorge Arellano Cid May 10, 2004

May 10, 2004

Hi there! It took a long time, but finally, dillo-0.8.1-pre2.tar.bz2 is ready for download. The problem was that the patch for the problem of left-clicking an unhandled MIME type involved changes, inside the cache module, that were not trivial. I pondered the situation carefully trying to decide whether to make the patch and delay the next release, or leave it for 0.8.2. After studying the situation and developing a patch, important changes were made in the cache (for instance all the cached URLs use new code now, and the main scheduler was also changed, and the posibility of deleting an active URL-transfer was made possible). After all that, it only remained to offer a download dialog for the aborted transfer, and as the patch was already major, I went ahead and the result was one of those improvements you can't leave out. What does it do? well, now you can left-click on any URL; if it happens to be of an unhandled MIME type, the transfer is aborted, and a download dialog appears to offer the possibility of saving the resource locally. Simple and clean. Now you can click carelessly on those fake-download links from SF until you reach the real source ;). The tradeoff: the possibility of labelling this release as "beta stable" is not as clear as before. This tarball requires plenty of testing, so please download it and give it a hard time! Any use of the browser serves, specially testing the new download feature and multiple windows at the same time. Get it from: http://www.dillo.org/download/dillo-0.8.1-pre2.tar.bz2 and report back to dillo-dev. Cheers Jorge.- PS: Kudos are also welcome! ;)

3 6

[Dillo-dev]compile problems on AIX + gcc
by Riccardo Mottola May 7, 2004

May 7, 2004

Hello, I'm trying to compile 0.8.1-pre1 on AIX 4.1 : One never knows, maybe som4e progress was made :) I had still to use CFLAGS="-lpthreads" to have configure pthread working at configure. I notice the following warnings gcc -DHAVE_CONFIG_H -I. -I. -I.. -DDILLORC_SYS='"/usr/local/etc/dillorc"' -I/u sr/local/include -I/usr/local/include/gtk-1.2 -I/usr/local/include/glib-1.2 -I/u sr/local/lib/glib/include -lpthreads -D_REENTRANT -D_THREAD_SAFE -Wall -Waggreg ate-return -c `test -f 'dw_bullet.c' || echo './'`dw_bullet.c gcc: -lpthreads: linker input file unused since linking not done should I have used LDFLAGS or something else? isn't there a more elegant solution to work around this pthread regnition problem? like a configure option? Cheers, Riccardo

1 0

[Dillo-dev]Re: about gtk on solaris 2.6
by Jorge Arellano Cid May 6, 2004

May 6, 2004

Riccardo, On Thu, 6 May 2004, Riccardo Mottola wrote: > > Hey Jorge, > > just a quick precisation. On the box where i have the old 0.7.3 dillo > which gives GTK errors, I have GIMP installed and that one works fine, I > have edited quite some images. > > SO I suspect it is something in the way dillo calls GTK that could > trigger the error, before saying GTK is bugged on that platform totally. There may be some good news for you: Try this patch: diff -pru dillo/src/dillo.c dillo-0.8.0/src/dillo.c --- dillo/src/dillo.c 2004-04-21 16:47:34.000000000 -0400 +++ dillo-0.8.0/src/dillo.c 2004-05-06 11:11:31.000000000 -0400 @@ -216,6 +216,8 @@ gint main(int argc, char *argv[]) /* Initialize GUI and parse GTK related args */ gtk_init(&argc, &argv); gdk_rgb_init(); + gtk_widget_set_default_colormap (gdk_rgb_get_cmap()); + gtk_widget_set_default_visual (gdk_rgb_get_visual()); /* Handle command line options */ while ((opt_flg = Dillo_get_opt(Options, argc, argv, &opt_argv, &idx))) { It fixes the colormap and visual detection/settings on some systems. The documentation is a bit vague about this calls, they are deprecated now, formerly suggested as initialization for most cases and removed from theGTK+2.0 API (with compatibility defines). If you're lucky, it may help with AIX too. Please test it and report. Note: The patch is in CVS now, but use a "pre1" patched tarball for the test. That asserts the use of the same configure script. Cheers Jorge.-

1 0

[Dillo-dev]downloads and pre2
by Jorge Arellano Cid May 6, 2004

May 6, 2004

On Tue, 4 May 2004, Yves de Champlain wrote: > Hi > [...] > Everything works fine except for this : > > av_open_url: Url=>http://ftp.gnu.org/gnu/ddd/ddd-3.2.1.tar.gz< > Connecting to 199.232.41.7 > warning: Unhandled MIME type: <application/x-tar> > > ** CRITICAL **: file web.c: line 57 (a_Web_dispatch_by_type): assertion > `dw != NULL' failed. > > Of course, this does not start any download but after this, > I still get messages in the shell but I will > not have those from wget once I do "save link as" > like I have if I do "save link as" before. Effectively, a left click on a non-dillo-viewable MIME type URL triggers this error message. The code inside dillo assigns it to a "null-client" which means it continues to download but it's "ignored". This has some side effects, one of them is the transfer not being aborted (so it continues wasting bandwidth and memory). It also keeps a cache entry, so when a dpi download is requested it hits and gets fed from the cache! (not from a dpi). That's clearly a bug. This days I've been reviewing the code with a view to patch it, but it seems that a full patch (aborting the transfer, cleaning the cache entry, stopping the clients and offering a download dialog) may be too complex to debug in this pre-release testing stage. Aborting the download (which would allow for the later dpi download) looks simpler. I'm currently working on this. That, plus the two patches in CVS seem to round "pre2" and most probably become 0.8.1. Cheers Jorge.-

1 0

[Dillo-dev]dillo-0.8.1-pre1.tar.bz2
by Jorge Arellano Cid May 6, 2004

May 6, 2004

Hi there, It took a long time to prepare the pre1 tarball for the next release. Mainly because complex problems were discovered in the last minute. For instance, a race condition in the dpi chain that produced segfaults (it took three days to find and fix), and several problems with the downloads server (it had troubles when saving unnamed pages as "index.html" when the file already existed, it was also not renaming those pages, etc). The "pre1" tarball can be found at: http://www.dillo.org/download/dillo-0.8.1-pre1.tar.bz2 When testing it, please pay special attention to the new detection routine for socklen_t (at ./configure time), and give the downloads server a hard test. Feedback is very important now, because I want to remove the "alpha software" label from dillo and name it as stable (beta). Thanks in advance Jorge.-

3 11

[Dillo-dev]CVS renamed ?
by Jens Arm May 5, 2004

May 5, 2004

Hi If I try to do a CVS checkout I get: [tux@tux dillo]$ cvs -z9 update -dP /sfhome/cvs/dillo: no such repository What happend ? Jens

2 1

[Dillo-dev]libpng-1.2.4 detection
by Jorge Arellano Cid May 4, 2004

May 4, 2004

Hi, Is anyone here using libpng-1.2.4? Rubén sent me a bug report and effectively we were not detecting libpng-1.2.4 right (this only happens with 1.2.4). A patch for this is in CVS, and it will appear in "pre2". I'll keep expecting more feedback from "pre1" because the png lib detection is the only change so far (it's minor and doesn't affect the work of the browser, just compilation). If anyone is using libpng-1.2.4, please let me know. Cheers Jorge.-

2 1

[Dillo-dev]GTK FLTK and other Render
by shuishuila May 3, 2004

May 3, 2004

Hi, dillo-dev， I want to port dillo to set-top box as menu(OSD). But it don't have X11,GTK. How many work need to do if I have to write widget myself? I think it's a pity that dillo combine tightly with glib/gtk, so it is hard to port to X,FrameBuffer,DirectFB etc. If dillo have it's own widget and event handle, it is a excellent broswer for embedded system. Regards, 　　　　　　　　shuishuila 　　　　　　　　shuishuila(a)hotmail.com 　　　　　　　　　　2004-04-30

2 1

Re: Re: [Dillo-dev]GTK FLTK and other Render
by shuishuila May 3, 2004

May 3, 2004

Hi,Jorge, Thank you for your reply. FLTK need X, so I need to port microwindows at least. Why no let dillo's render engine more scalable? And I think that few desktop system have FLTK. Regards, shuishuila(a)hotmail.com 　　　　　　　　　　2004-05-03 ======= 2004-05-02 10:41:00 You Wrote：======= >On Fri, 30 Apr 2004, shuishuila wrote: > >> Hi, dillo-dev， > > Hi. > >> >> I want to port dillo to set-top box as menu(OSD). But it don't have >> X11,GTK. How many work need to do if I have to write widget myself? > > That's a lot of work, but... > >> >> I think it's a pity that dillo combine tightly with glib/gtk, so it >> is hard to port to X,FrameBuffer,DirectFB etc. >> >> If dillo have it's own widget and event handle, it is a excellent >> broswer for embedded system. > > We're almost done with a plan to port Dillo to FLTK. > > You can read about it in our mailing list archives. Start here: > > http://lists.auriga.wearlab.de/pipermail/dillo-dev/2004-April/001882.html > > > If you have FLTK knowledge and are willing to help us doing the >port it'd be great. > > Cheers > Jorge.- = = = = = = = = = = = = = = = = = = = =

1 0