widget: slider: Add stepping functionality#1875
Conversation
raymanfx
force-pushed
the
master
branch
2 times, most recently
from
5679c05 to
a6bbf2b
Compare
|
What should a |
|
No step at all is reasonable. It makes logical sense to me, you have infinite steps. Maybe its more rust-like use an option and disallow |
|
This is good to go from my point of view now. |
maan2003
left a comment
maan2003
left a comment
There was a problem hiding this comment.
- What will/should happen if initial value is not aligned with the step?
- Should we show the available points(as little dots) on the slider?
|
Huh, the stepping is 10. I can see there being debate on of that should be in the form of With this rounding formula a stepping if 11 would create the same result as with 10. |
because self.max was used in calculation. and self.max / step is fractional |
|
different implementation treats max as a separate step if it is not "well aligned"(min + n * step) |
raymanfx
force-pushed
the
master
branch
3 times, most recently
from
e86bf22 to
5ddbd6c
Compare
A new builder method `with_step(f64)` is added to set the stepping value. Signed-off-by: Christopher N. Hesse <raymanfx@gmail.com>
In case of max % step != 0, the last step is smaller than the others. Signed-off-by: Christopher N. Hesse <raymanfx@gmail.com>
Handle the edge case where we need an extra step separately. This occurs whenever max % step != 0. Signed-off-by: Christopher N. Hesse <raymanfx@gmail.com>
We now handle that case and allow for self.max to be reached. Signed-off-by: Christopher N. Hesse <raymanfx@gmail.com>
Signed-off-by: Christopher N. Hesse <raymanfx@gmail.com>
Signed-off-by: Christopher N. Hesse <raymanfx@gmail.com>
|
@cmyr could you please hit the big scary button? :) |
cmyr
left a comment
cmyr
left a comment
There was a problem hiding this comment.
Works for me! As always this stuff is trickier than expected, but I think it's also fair for us to expect the user to provide sane values.
| if step < 0.0 { | ||
| warn!("bad stepping (must be positive): {}", step); | ||
| return self; | ||
| } | ||
| self.step = if step > 0.0 { | ||
| Some(step) | ||
| } else { | ||
| // A stepping value of 0.0 would yield an infinite amount of steps. | ||
| // Enforce no stepping instead. | ||
| None | ||
| }; |
There was a problem hiding this comment.
No need to change it, but just for posterity I'd have just written:
| if step < 0.0 { | |
| warn!("bad stepping (must be positive): {}", step); | |
| return self; | |
| } | |
| self.step = if step > 0.0 { | |
| Some(step) | |
| } else { | |
| // A stepping value of 0.0 would yield an infinite amount of steps. | |
| // Enforce no stepping instead. | |
| None | |
| }; | |
| if step < 0.0 { | |
| warn!("bad stepping (must be positive): {}", step); | |
| } else { | |
| self.step = Some(step); | |
| } |
There was a problem hiding this comment.
That has a drawback IMO: you'd actually need to check for step <= 0 since otherwise we would divide by zero elsewhere iirc.
We decided to allow 0 as a valid step value though, indicating smooth stepping (no intervals) - but we have to explicitly model this as None.
There was a problem hiding this comment.
lets change that step > 0.0 to step != 0.0 to make it more explicit
4 participants
A new builder method
with_step(f64)is added to set the stepping value.Signed-off-by: Christopher N. Hesse raymanfx@gmail.com