W3C home > Mailing lists > Public > www-style@w3.org > September 2010

Re: [css3-images] [css3-values] Inconsistent Angles

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Wed, 8 Sep 2010 00:06:42 -0700
Message-ID: <AANLkTi=1DxXTgJWMHbB2jz9vAeLAjh=4DTNsS+40iejL@mail.gmail.com>
To: fantasai <fantasai.lists@inkedblade.net>
Cc: David Singer <singer@apple.com>, Simon Fraser <smfr@me.com>, "www-style@w3.org" <www-style@w3.org>
On Tue, Sep 7, 2010 at 11:13 PM, fantasai <fantasai.lists@inkedblade.net> wrote:
> On 09/07/2010 08:17 PM, Tab Atkins Jr. wrote:
>>
>> On Tue, Sep 7, 2010 at 8:04 PM, fantasai<fantasai.lists@inkedblade.net>
>>  wrote:
>>>
>>> On 09/07/2010 06:09 PM, Tab Atkins Jr. wrote:
>>>>
>>>> On Tue, Sep 7, 2010 at 5:47 PM, David Singer<singer@apple.com>    wrote:
>>>>>
>>>>> why isn't the starting point for a radial linear gradient
>>>>> "those point(s) on a line perpendicular to the gradient angle,
>>>>> in the most negative position possible, that still just fit(s)
>>>>> within the shape to be filled"?
>>>>
>>>> I can't understand what you're trying to say here.
>>>
>>> He's saying what you're saying, just generalizing it so you don't need
>>> different rules depending on the angle's value.
>>>
>>> Here's the idea: You have an angle. You want a gradient.
>>>
>>> 1. Draw a directed line at that angle.
>>> 2. Place it over the centerpoint of the box. Call this point zero.
>>> 3. Draw a line perpendicular to the line.
>>> 4. Shift the line backwards (i.e. negatively) until the line's
>>>   intersection with the directed line is as negative as possible
>>>   while still intersecting the box.
>>>
>>> The intersection point in step 4 is the start point of the gradient.
>>
>> Oh, got it.  Yeah, that's just the opposite of how I'm currently
>> defining the ending point
>
> Starting point, Tab. He's using this to define the starting point.
> To find the ending point, you'd take step 4 to the most negative
> possible intersection.

...right.  That's what I said.  That definition is the *opposite* of
what I'm using for determining the ending point, where I just project
the perpendicular line forwards rather than backwards.

~TJ
Received on Wednesday, 8 September 2010 07:07:35 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:31 GMT