[whatwg] date/time and precision changes
hyatt at apple.com
Wed Jun 16 12:52:08 PDT 2004
Ok, missed the "any" value in the spec. That solves my problem. I do
think you should state that the value property for sliders must be
locked to a step multiple.
On Jun 16, 2004, at 11:39 AM, David Hyatt wrote:
> Also a float slider is simply continuous... it effectively has no step
> value... I don't see how to specify a continuous float slider without
> having to say something hacky like step="0.001". If you must have a
> step attribute, I think you should have a special value, continuous,
> to denote that the entire range of values is allowed.
> This is another reason why I think the precision value was necessary.
> Both a float slider and int slider should by default be considered
> continuous, i.e.to support all possible values. While a step
> attribute is definitely useful, I don't think it's intuitive to use it
> to set the data type of the slider control. Logically the default
> value of step should be "continuous" IMO and precision="integer|float"
> should be supported to determine what "continuous" means.
> Adding numeric values for step also introduces an ambiguity of how the
> slider should behave when dragging. Does it jump in increments or
> does it smoothly slide and do rounding to the nearest step increment?
> This is perhaps something for a CSS property to control.
> On Jun 16, 2004, at 11:05 AM, Dave Hyatt wrote:
>> I strongly disagree with dropping the precision attribute for <input
>> type=range>. I at least like the ability to distinguish between
>> float and integer.
>> How does a user tell the form control to submit as either a float or
>> an integer now? I already implemented precision="float" to
>> accomplish this. The precision as an integer ensured that even if
>> the slider was handing out float values that they would be
>> auto-converted when accessing/submitting/setting the value attribute.
>> With your changes this ability seems to have been lost.
>> On Jun 16, 2004, at 7:40 AM, Ian Hickson wrote:
>>> Based on all your feedback (thanks everyone!) I've reworked the
>>> types and attributes:
>>> types: datetime, date, time, expdate, week
>>> attributes: precision, min, max
>>> Let me know if you are not happy with the new text, and if you have
>>> more suggestions.
>>> Apologies to anyone who'd already implemented this stuff...
>>> Ian Hickson U+1047E )\._.,--....,'``.
>>> http://ln.hixie.ch/ U+263A /, _.. \ _\
>>> ;`._ ,.
>>> Things that are impossible just take longer.
More information about the whatwg