Age | Commit message (Collapse) | Author |
|
|
|
Avoid margin-bottom being counted twice.
Testcase:
<div style="border: 1px solid black; margin-bottom: 100px">foo bar</div>
<div style="border: 1px solid black">foo bar</div>
|
|
This is not a bug in the event dispatcher but a problem with "hoverTooltip",
which doesn't hold accurate data in some cases. Making it static solves
the problem.
|
|
In this case the ancestor is already in "entered" state, but we need to
emit the signal again, for it to reset its state and resume.
|
|
This patch also disables unused code in Layout::leaveNotify()
|
|
This patch fixes 3f74e4d60ac1 (which is partly wrong).
The idea is to avoid spurious leave/enter events, and thus prevent bugs.
* avoid issuing spurious leave/enter events, and send the necessary ones.
* remove the need for a workaround linkEnter() call in
Textblock::leaveNotifyImpl().
|
|
|
|
Don't draw background for widgets in Textblock::drawLine().
This fixes the border_background_margin.html test in css_compat.
|
|
for the sake of text-indent and tables.
|
|
|
|
thread: http://lists.auriga.wearlab.de/pipermail/dillo-dev/2010-November/007801.html
|
|
|
|
Add lout::misc::roundInt() and use it for double -> int conversion
instead of doing the + 0.5 trick all over the place. It was wrong for
negative values and we might even replace roundInt() with rint() from
libm in the future.
Reported-by: corvid
|
|
|
|
it became unnecessary when anchors were separated out in November
|
|
When a widget changes it's height, this may affect the position of other
words in the same line. Take care by only applying the redraw
optimization if there are no other words in the line.
This special case is quite common e.g with nested <div>'s or table based
layouts and the redraw optimization avoids unnecessary redraws.
Reported by: Dennis Nezic
|
|
Using line height for this and the earlier Word change seems to
make the appearance closer to firefox's. Of course, I'm sure there
will be counterexamples...
|
|
|
|
|
|
|
|
|
|
|
|
|
|
thread:
http://lists.auriga.wearlab.de/pipermail/dillo-dev/2010-March/007420.html
|
|
http://lists.auriga.wearlab.de/pipermail/dillo-dev/2010-February/007368.html
I'm committing a modified version that Johannes made. It moves the
(un)highlight checks into queueDrawRange() and add checks for valid wordIndex
in findLineOfWord().
|
|
|
|
|
|
Group things conceptually, etc. No change in behaviour is intended.
(I would like to rip out all of the DBG stuff someday, which I think
does more harm than good by making the src hard to read.)
|
|
No change in functionality is intended.
|
|
|
|
I noticed that when I was highlighting text in line n, Dillo was also
redrawing line n-1.
Incidentally, it seems that the default arrow cursor points with the tip
of the black arrow itself, and not with the tip of the white border
surrounding it.
|
|
|
|
The reason for this is that they were in the way when adding parbreaks.
http://lists.auriga.wearlab.de/pipermail/dillo-dev/2009-November/007044.html
|
|
|
|
|
|
|
|
|
|
|
|
I had changed it in rev. f231450b81a1 a few days ago.
http://lists.auriga.wearlab.de/pipermail/dillo-dev/2009-October/006943.html
|
|
|
|
http://lists.auriga.wearlab.de/pipermail/dillo-dev/2009-October/006936.html
|
|
|
|
The earlier space is less likely to be styled with the intended appearance.
|
|
|
|
|
|
|
|
|
|
"using namespace" in headers also affects all files that include the
header which is unwanted and unexpected in most cases.
|
|
|
|
|