make better weighting widget than radio buttons

[chadwhitacre]

@alex:

I have no idea what these radio button things are for.

https://twitter.com/alex_gaynor/status/323499448212799488

[chadwhitacre]

Reticketed from #449.

[dmitshur]

I have to admit I was really confused at first too. Tooltips are good for detailed information, I didn't even think to try them for such basic info. You'd need at least "min" and "max" labels at both ends to establish a direction. Plus I was on a tablet where there's no way to see tooltips.

Perhaps the Humble Bundle can be of some inspiration for how to do splitting/weighing well?

[chadwhitacre]

@coderanger

The scale doesn't say which end is which.

https://twitter.com/whit537/status/323555792378793984

[chadwhitacre]

@rummik mentions the HTML5 range input type.

https://botbot.me/freenode/gittip/msg/2724800/

[rummik]

Needs a fallback, and to be more Angular, but here's something of a mockup: http://jsbin.com/ebidiy/1

[rummik]

Some changes: http://jsbin.com/ebidiy/12 -- is this how everyone expects it to work when divvying up their own gifts to a project, or am I looking at this totally wrong?

[dmitshur]

@rummik I would stick with the original version. It gets too unwieldy to control when they all change at once. Let the user adjust manually one at a time, and potentially normalize afterwards.

There's a problem with the all-sliders-move-automatically approach where by setting one slider to max for a second, you unintentionally reset all the others to be equal (since they all get set to 0 temporarily).

[rummik]

@shurcooL Yeah, I think the HIB sliders adjust based upon their initial values, rather than their current values (which would fix that). It also loses a lot of precision when you set them to different values and wobble one back and forth (since iirc it steps by 1 when not given a step)

In a way I'm a bit partial to the all-sliders-move-automatically approach since it gives a much more visual heads up to their gift distribution. Though the other way does give much more control over the distribution...

[rummik]

It matches the HIB distribution things: http://jsbin.com/ebidiy/17

(Was moreso an exercise in fun than anything else to polish it -- I'm happy with either slidey method)

[rummik]

More realistic? Or just confusing? http://jsbin.com/ebidiy/22

[dmitshur]

Re 22: So that conveys a single extra piece of information. That, if you have all 3 sliders set at 75% or all three at 3%, it doesn't matter - the end result is an even 33% split three ways. I'm not completely sure if it's needed to visualize this or not. I've played with it a for a bit, and I can't tell if it's better with or without that.

[chadwhitacre]

Obsolete.