The test should definitely include width == 0 for now.
I don't know exactly why. I'd guess it has to do with initializing
resizable single line texts' whose width is still 0.
This will be much easier to resolve when lfc/sprite callback model is
cleaned up as part of this change:
'Change 20100410-hqm-e Summary: UPDATED ^2 BY MAX: make autosizing
inputtext recompute it's height properly when user types'
At least it appears to be in the LFC and not runtime-specific.
Can you file a separate bug to address this nit?
On 4/15/10 2:45 PM, P T Withington wrote:
What we are asking is should the test be explicitly for "width == 0"? If so,
could you elucidate us? None of us understand why we need this, er, quirk.
On 2010-04-14, at 19:40, Max Carlson wrote:
the width == 0 test is required for test/lztest/lztest-textheight.lzx to pass.
The changes to the swf9 kernel are enough to fix the issue - but it still shows
up in DHTML...
On 4/14/10 12:54 PM, André Bargull wrote:
<canvas height="500" width="100%">
<view width="100" height="100" bgcolor="blue">
<simplelayout axis="y" spacing="3"/>
<text name="dots" text="..." bgcolor="red"
oninit="this.setAttribute('width', 0);"/>
<button name="width0" text="change width to 0">
<handler name="onclick"><![CDATA[
parent.dots.setAttribute("width", 0);
]]>
</handler>.
</button>
<button name="width20" text="change width to 20">
<handler name="onclick"><![CDATA[
parent.dots.setAttribute("width", 20);
]]>
</handler>.
</button>
</view>
</canvas>
--
Regards,
Max Carlson
OpenLaszlo.org
--
Regards,
Max Carlson
OpenLaszlo.org
