September 2008 - Dillo-dev - lists.mailman3.com

default font_factor

by jcid＠dillo.org

On Sat, Sep 06, 2008 at 09:33:47PM +0000, corvid wrote: > Jorge wrote: > > On Sat, Sep 06, 2008 at 07:17:49PM +0000, corvid wrote: > > > Johannes wrote: > > > > recently the the default font_factor was changed from 1.0 to 1.2. > > > > Is there a reason for this change? > > > > > > I think 1.2 was the uncommented value in the old dillorc2. > > > > > > It would have made more sense for me to leave it as 1.0 and > > > change the elevens and twelves that it gets multiplied by... > > > > Yeah, having it 1.0 and scaling the base size makes sense to me. > > attached. > Committed. -- Cheers Jorge.-

Sept. 8, 2008

an example of page redraw problem when images load

by jcid＠dillo.org

On Sat, Sep 06, 2008 at 10:26:03PM +0200, Johannes Hofmann wrote: > On Sat, Sep 06, 2008 at 04:28:55PM -0400, Jorge Arellano Cid wrote: > > On Sat, Sep 06, 2008 at 06:38:23PM +0000, corvid wrote: > > > Johannes wrote: > > > > On Sat, Sep 06, 2008 at 03:53:34AM +0000, corvid wrote: > > > > > It was like dillo had treated the </div></div><p> as a <br> until that > > > > > click. I see that Textblock::addParbreak() has: > > > > > > > > > > /* Another break before? */ > > > > > if ((word = words->getRef(words->size () - 1)) && > > > > > word->content.type == core::Content::BREAK) { > > > > > word->content.breakSpace = > > > > > misc::max (word->content.breakSpace, space); > > > > > return; > > > > > } > > > > > > > > > > word = addWord (0, 0, 0, style); > > > > > word->content.type = core::Content::BREAK; > > > > > word->content.breakSpace = space; > > > > > wordWrap (words->size () - 1); > > > > > > > > > > For the sake of experiment, I stuck a > > > > > wordWrap (words->size () - 1); > > > > > before the return, and now it seems to work. > > > > > I imagine it must have to do with lastLine->breakSpace > > > > > getting updated... > > > > > > > > Yes, it seems so. Can you please test attached patch? It just updates > > > > lastLine->breakSpace instead of calling Textblock::wordWrap, > > > > because of the comment above Textblock::wordWrap... > > > > > > > > It also fixes some bugs in the redraw optimization. > > > > At least the <ol> issue (google.com) no longer happens here. > > > > > > It appears to be working well. > > > Thanks! > > > > Johannes: Should I commit this patch? > > Yes please. Committed. -- Cheers Jorge.-

Sept. 8, 2008

assertion failed during wildly clicking on links

by akemnade＠tzi.de

Hi, On Sat, 6 Sep 2008 22:03:16 +0200 Johannes Hofmann <Johannes.Hofmann(a)gmx.de> wrote: > Hi, > > On Sat, Sep 06, 2008 at 09:34:29PM +0200, Andreas Kemnade wrote: > > Hi, > > > > I have seen the following, when clikling multiple times very fast on > > a link: > > Nav_open_url: new url='http://www.dwd.de/sid_520SLCXTTJJbtdSgR7JTsLGvkNt6TkS9ssWxDtLpSvVZtPCcyjPG!…' > > Connecting to 141.38.3.62 > > DilloHtml(): content type: text/html; charset=UTF-8 > > *** [unknown] Assertion failed! *** > > anchor 'headline_color_324A92' not defined > > Aborted > > That is typical. I suggest to disable asserts for the release and > some minutes later there is a new report of an assertion failure :-) > > > > > I tried www.dwd.de and then to click Wetter&Warnungen and the corresponding link. > > I can't reproduce it yet. Do you left-click on the link? If so, the > second click is already on the new page, right? > I did not wait for the new page to appear. It was on my freerunner. So I used the touch screen. I saw the assertion failure first when there was some bouncing. Then I tried to do it more systematically. > > > > I have not looked into the source yet, but I think there is a ressource management > > problem. > > Can you post a backtrace? > Well, there is a problem with too much use of -fomit-frame-pointer. So it is not that easy. Hmm, I'll try again with a breakpoint in misc::assert() this evening. Greetings Andreas Kemnade

Sept. 8, 2008

Fixed FltkViewBase::drawRectangle()

by corvid＠lavabit.com

Johannes wrote: > Hi Jorge, > > regarding the last change in fltkviewbase.cc: > > --- a/dw/fltkviewbase.ccSat Sep 06 19:51:25 2008 +0200 > +++ b/dw/fltkviewbase.ccSun Sep 07 10:51:04 2008 +0200 > @@ -384,7 +384,9 @@ > setcolor(((FltkColor*)color)->colors[shading]); > int x1 = translateCanvasXToViewX (x); > int y1 = translateCanvasYToViewY (y); > - ::fltk::Rectangle rect (x1, y1, width, height); > + int x2 = translateCanvasXToViewX (x + width); > + int y2 = translateCanvasYToViewY (y + height); > + ::fltk::Rectangle rect (x1, y1, x2 - x1, y2 - y1); > if (filled) > fillrect (rect); > else > > > To me it looks like this change is only necessary, if > translateCanvasXToViewX() would do some sort of scaling. Was there a > real problem, or did you change it just for consistency? Ah, that's what it was for -- fltkpreview has scaling, which I forgot about. (was fltkpreview intended to be hooked into something somewhere? I have mixed feelings about carrying around code that we don't have any particular plans to do anything with...)

Sept. 7, 2008

Missing include directive for OpenBSD?

by Johannes.Hofmann＠gmx.de

On Sun, Sep 07, 2008 at 02:16:38PM +0200, Christian Kellermann wrote: > * Johannes Hofmann <Johannes.Hofmann(a)gmx.de> [080907 14:09]: > > On Sun, Sep 07, 2008 at 01:51:19PM +0200, Christian Kellermann wrote: > > It's needed for the definition of intptr_t, right? > > Does #include <stdint.h> (without sys/) also help? > > Alternatively we could include <inttypes.h>. > > Aye, stdint.h also does it. Ok, thanks for the report. Johannes

Sept. 7, 2008

[patch] fix compilation on OpenBSD

by Johannes.Hofmann＠gmx.de

Hi, this fixes compilation on OpenBSD. Please verify on Linux. Cheers, Johannes

Sept. 7, 2008

Missing include directive for OpenBSD?

by Christian.Kellermann＠nefkom.net

* Johannes Hofmann <Johannes.Hofmann(a)gmx.de> [080907 14:09]: > On Sun, Sep 07, 2008 at 01:51:19PM +0200, Christian Kellermann wrote: > It's needed for the definition of intptr_t, right? > Does #include <stdint.h> (without sys/) also help? > Alternatively we could include <inttypes.h>. Aye, stdint.h also does it. Cheers, Christian -- You may use my gpg key for replies: pub 1024D/47F79788 2005/02/02 Christian Kellermann (C-Keen)

Sept. 7, 2008

Missing include directive for OpenBSD?

by Johannes.Hofmann＠gmx.de

On Sun, Sep 07, 2008 at 01:51:19PM +0200, Christian Kellermann wrote: > Dear List, > > I have not updated dillo2 for a long time. Now that I have done so > I see that I have missed a lot! I needed the following tiny patch > to allow compilation on my OpenBSD 4.3 System: > > Index: dw/style.cc > =================================================================== > RCS file: /sfhome/cvs/dillo/dw2/dw/style.cc,v > retrieving revision 1.10 > diff -b -u -p -r1.10 style.cc > --- dw/style.cc 5 Sep 2008 12:58:52 -0000 1.10 > +++ dw/style.cc 7 Sep 2008 11:50:08 -0000 > @@ -24,6 +24,7 @@ > #include <string.h> > #include <unistd.h> > #include <ctype.h> > +#include <sys/stdint.h> > > #include "core.hh" > > > I am not sure whether this should have been caught by an autoconfig > check. Can someone point me into the right direction so this can > be changed to a proper patch? It's needed for the definition of intptr_t, right? Does #include <stdint.h> (without sys/) also help? Alternatively we could include <inttypes.h>. Cheers, Johannes

Sept. 7, 2008

Missing include directive for OpenBSD?

by Christian.Kellermann＠nefkom.net

Dear List, I have not updated dillo2 for a long time. Now that I have done so I see that I have missed a lot! I needed the following tiny patch to allow compilation on my OpenBSD 4.3 System: Index: dw/style.cc =================================================================== RCS file: /sfhome/cvs/dillo/dw2/dw/style.cc,v retrieving revision 1.10 diff -b -u -p -r1.10 style.cc --- dw/style.cc 5 Sep 2008 12:58:52 -0000 1.10 +++ dw/style.cc 7 Sep 2008 11:50:08 -0000 @@ -24,6 +24,7 @@ #include <string.h> #include <unistd.h> #include <ctype.h> +#include <sys/stdint.h> #include "core.hh" I am not sure whether this should have been caught by an autoconfig check. Can someone point me into the right direction so this can be changed to a proper patch? Thanks for your efforts and please keep up the good work! Christian -- You may use my gpg key for replies: pub 1024D/47F79788 2005/02/02 Christian Kellermann (C-Keen)

Sept. 7, 2008

load_images and multiple windows

by Johannes.Hofmann＠gmx.de

On Sat, Sep 06, 2008 at 08:28:28PM +0000, corvid wrote: > Johannes wrote: > > the load_images button toggles the global prefs.load_images > > variable but the buttons of other browser windows are not updated > > accordingly. > > Should load_images be a per window variable, or should we update the > > buttons in all other windows? > > Doesn't matter to me; I don't use it. > > > Also, when switching to load_images=true, the "Load images" entry in > > the popup menu get's deactivated, even though the images of the > > current page are not yet loaded. > > I'd been ignoring it since I very rarely use that either, but all right, > here's a patch (will apply at an offset since my tree has forms code in > it that isn't worth putting into CVS until after release). > Very nice. Now it also deactivates the "Load images" menu entry properly when all images have been loaded manually. Thanks, Johannes

Sept. 7, 2008