[Bf-taskforce25] Updated TODO's

[William Reynish]

Hi All.

After discussion with Joe on IRC we agreed on using the 2D dragging  
approach, where moving the cursor horizontally linearly increases/ 
decreases the value. Moving the cursor vertically increases/decreases  
the sensitivity.

Cheers,

-W


On 14 Aug, 2009, at 5:31 PM, joe wrote:

> I disagree.  As I said before, I think it's something that can be
> solved.  And there's plenty of situations where you need it, where a
> linear step would be even more unusable.
>
> Joe
>
> On Fri, Aug 14, 2009 at 7:35 AM, William  
> Reynish<william at reynish.com> wrote:
>> Well, the exponential nature makes them unpredictable to use. Who  
>> says
>> you want less precision the further away you are from the original
>> number? This assumption makes dialing in numbers with any accuracy
>> really hard, even impossible. This is especially obvious with
>> translations. Drag a transform location field, and you'll see your
>> object move along slowly, until suddenly it shoots along into the
>> distance.
>>
>> There are a few other possibilities though:
>>
>> 1. Use speed of movement to determine accuracy
>>
>> 2. Use vertical height to determine accuracy
>>
>> These are more predictable by users, because the user input is then
>> constant, and doesn't randomly change over time, causing  
>> unpredictable
>> results.
>>
>> First option is also not a 1:1 mapping of movement, bit it's more
>> predictable than having things shoot off exponentially.
>>
>> The second option (suggested by Aligorith in IRC) could work quite
>> well. The idea is that the further up, vertically or the screen, the
>> cursor is, the higher increments the number increases/decreases, and
>> visa versa. But for it to be obvious to the user we'd probably need  
>> to
>> add some sort of visual indication. The downside is that you'd have  
>> to
>> pay more attention to the direction of your mouse gestures.
>>
>>
>>
>> -W
>>
>>
>> On 14 Aug, 2009, at 3:12 PM, joe wrote:
>>
>>> Numbuts originally were linear; this was really annoying, IMHO.  I
>>> think it's be a mistake to revert to that behavior (I find the
>>> exponential version much easier); instead we should make the
>>> exponential more usable; e.g., have it decrease when you move back  
>>> to
>>> the original position, maybe have a fixed exponential range (so  
>>> after
>>> a certain point it's no longer exponential) etc.
>>>
>>> Joe
>>
>> _______________________________________________
>> Bf-taskforce25 mailing list
>> Bf-taskforce25 at blender.org
>> http://lists.blender.org/mailman/listinfo/bf-taskforce25
>>
> _______________________________________________
> Bf-taskforce25 mailing list
> Bf-taskforce25 at blender.org
> http://lists.blender.org/mailman/listinfo/bf-taskforce25