I'll be damned. It works. I swear that I tried this yesterday and it
didn't work. I must be losing it.
Quoting P T Withington <[EMAIL PROTECTED]>:
> You can do the same thing in Javascript, since true -> 1 and false -> 0
> in a numeric context.
>
> On 2006-09-25, at 20:34 EDT, [EMAIL PROTECTED] wrote:
>
>> Hmmm. I have an idea. In my situation, I just need to stop the width
>> from going below zero.
>>
>> In C I'd do something like this...
>>
>> width = (parentwidth-20} * (parentwidth-20 > 0);
>>
>> Is there some mathematical trickery I can use to convert true/false to
>> 1/0 in a constraint?
>>
>> The best I can come up with is the ternary operator...
>>
>> width="${(parent.width-20) * (parent.width-20 > 0)?1:0}"
>>
>>
>>
>>
>>
>> Quoting P T Withington <[EMAIL PROTECTED]>:
>>
>>> If I wanted to be a pedant (ok, I do) I would tell you that
>>> [functions](http://en.wikipedia.org/wiki/
>>> Functional_programming#Pure_functions) are just fine in constraints,
>>> it is 'procedures' that are not.
>>>
>>> On 2006-09-25, at 16:21 EDT, John Sundman wrote:
>>>
>>>> Please consult the Developer's Guide on the hazards of using
>>>> functions within constraints.
>>>>
>>>> http://labs.openlaszlo.org/trunk-nightly/docs/guide/
>>>> constraints.html#d0e13355
>>>>
>>>> It can sometimes be done, but there be dangers there, and you do so
>>>> at your own risk.
>>>>
>>>> jrs
>>>>
>>>> On Sep 23, 2006, at 2:48 PM, Benjamin Shine wrote:
>>>>
>>>>>
>>>>> Constraints can hold complex javascript. To cap constraints to a
>>>>> range (100-500 in this example), I use something like
>>>>> width="{max(100, min(500, sizeslider.value))}"
>>>>>
>>>>> A nest of trinary ? expressions would probably be faster by avoiding
>>>>> two function calls. With whitespace added for clarity:
>>>>> width="${ sizeslider.value < 100
>>>>> ? 100
>>>>> : ( sizeslider.value > 500 ?
>>>>> 500
>>>>> : sizeslider.value)) }"
>>>>>
>>>>> It is sometimes clearer to express this as a handler for
>>>>> sizeslider.onvalue, something like this:
>>>>> <handler reference="sizeslider" event="onvalue"
>>>>> method="setClampedSize" />
>>>>> <method name="setClampedSize" args="x">
>>>>> var xclamp = min(100, max(500, x));
>>>>> mybox.setAttribute("width", xclamp);
>>>>> </method>
>>>>>
>>>>> The compiler probably turns the first and third forms above into the
>>>>> same thing; if the compiler has an optimization for min and max (but
>>>>> I don't think it does) then all three forms would be equivalent.
>>>>>
>>>>> -ben
>>>>>
>>>>> On Sep 22, 2006, at 10:25 PM, [EMAIL PROTECTED] wrote:
>>>>>
>>>>>> In my example below, I create a view with a smaller nested view.
>>>>>> The
>>>>>> inner view gets its size from constraints based on the its parent.
>>>>>> The problem is, when the parent is resized too small, the
>>>>>> constraints
>>>>>> on the inner view become negative. If you run my app, you can
>>>>>> adjust
>>>>>> the size with the slider and see what happens. Basically, I need
>>>>>> some
>>>>>> way to have a constraint that tracks the parent, but cannot become
>>>>>> negative. Is there some way to do this in a contstraint? Or
>>>>>> possibly
>>>>>> a different solution that doesn't have the negative constraint
>>>>>> problem?
>>>>>>
>>>>>>
>>>>>> <canvas>
>>>>>>
>>>>>> <view x="20" y="20" layout="axis:y">
>>>>>> <view layout="axis:x">
>>>>>> <text text="outer view size: "/>
>>>>>> <slider id="sizeslider" width="100" showvalue="true"
>>>>>> value="50"/>
>>>>>> </view>
>>>>>> <view layout="axis:x">
>>>>>> <text text="inner view size: "/>
>>>>>> <text text="${box.subviews[0].height}" bgcolor="silver"/>
>>>>>> </view>
>>>>>> </view>
>>>>>>
>>>>>> <view id="box" x="20" y="80" bgcolor="blue"
>>>>>> width="${sizeslider.value}" height="${sizeslider.value}">
>>>>>> <view bgcolor="green" x="10" y="10"
>>>>>> width="${parent.width - 20}" height="${parent.height -
>>>>>> 20}"/>
>>>>>> </view>
>>>>>>
>>>>>>
>>>>>> </canvas>
>>>>>>
>>>>>>
>>
>>
>>
>> _______________________________________________
>> Laszlo-user mailing list
>> [email protected]
>> http://www.openlaszlo.org/mailman/listinfo/laszlo-user
_______________________________________________
Laszlo-user mailing list
[email protected]
http://www.openlaszlo.org/mailman/listinfo/laszlo-user
