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

Fixed FltkViewBase::drawRectangle()

by Johannes.Hofmann＠gmx.de

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? Cheers, Johannes

Sept. 7, 2008

default font_factor

by corvid＠lavabit.com

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.

Sept. 6, 2008

load_images and multiple windows

by jcid＠dillo.org

On Sat, Sep 06, 2008 at 09:00:33PM +0200, Johannes Hofmann wrote: > Hi, > > 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? A per window setting looks OK to me. That way people that use it only on certain pages (like me) are happy, and those that prefer text mode most of the time can have load_images=NO in dillo2rc. > 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. This is a bug. -- Cheers Jorge.-

Sept. 6, 2008

an example of page redraw problem when images load

by Johannes.Hofmann＠gmx.de

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. Cheers, Johannes

Sept. 6, 2008

default font_factor

by jcid＠dillo.org

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. -- Cheers Jorge.-

Sept. 6, 2008

load_images and multiple windows

by corvid＠lavabit.com

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).

Sept. 6, 2008

an example of page redraw problem when images load

by jcid＠dillo.org

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? -- Cheers Jorge.-

Sept. 6, 2008

assertion failed during wildly clicking on links

by Johannes.Hofmann＠gmx.de

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 have not looked into the source yet, but I think there is a ressource management > problem. Can you post a backtrace? Cheers, Johannes

Sept. 6, 2008

assertion failed during wildly clicking on links

by akemnade＠tzi.de

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 I tried www.dwd.de and then to click Wetter&Warnungen and the corresponding link. I have not looked into the source yet, but I think there is a ressource management problem. Greetings Andreas Kemnade

Sept. 6, 2008

default font_factor

by corvid＠lavabit.com

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...

Sept. 6, 2008