[Bf-committers] new features without release log

[Martin Poirier]

Hi,

--- Ben Stabler <ben at half-dome.net> wrote:

> "From the angle at the tip of the pyramid, you can
> derive the zoom factor."
> 
> I still don't understand how this is possible; if
> you imagine a large cube
> beyond a smaller cube, and use alt-b to draw the
> pyramid around the cubes
> from the camera, they can both 'only just' fill the
> cross-section of the
> pyramid, even though one is far larger than the
> other.

How is that a problem?

> If the sides of the
> pyramid are at normals to the camera, how is it
> possible to calculate the
> correct offset?

They are normal to the camera, that's the whole point
(and that's exactly how the pyramid is original
defined, by taking the normals at the four corners of
the box).

> The angle at the tip gives the zoom
> factor of the camera,
> but I don't understand how that tells you how far
> along the pyramid the user
> intends to 'zoom' to (keeping in mind that the
> center of rotation is pretty
> much irrelevant at this stage).

It doesn't (and I don't think I ever suggested it
did).

The pyramid define a viewport orientation and zoom,
not where the focus is along the depth of the viewport
.

> "can't you just dolly by a certain amount
> proportional to the wheel movement
> and the zoom of the viewport (prior to the current
> zoom operation)?"
> 
> Yes, but this would break if the user is working on
> a large scene, and wants
> to zoom into a small detail- the zoom factor is
> large because they are
> rotating around a large scene, and dollying towards
> a detail won't change
> the zoom factor (as it is a translation not a zoom),
> so the camera would be
> likely to fly straight past the detail.

Like I said: "proportional to the wheel movement and
the zoom of the viewport"

Big Zoom -> Small Dolly

> Hmm, this also feeds into the last point- if the
> user zooms into a detail,
> and the center of rotation is as far past the detail
> as the camera was
> previously, then you can't really rotate about it
> once you're zoomed in on
> it.

Which is why people have to be mindful about viewport
center placement.

Moreover, that is not a problem if they use Rotate
Around Active (in object mode, it still doesn't work
correctly in edit mode)

> "Is that really the desired behavior? I mean, when I
> set the viewport center
> manually, I'd be pretty pissed if zooming on
> something messed it up."
> 
> Well, this has to happen in some way, as by
> definition both 'zoom to cursor'
> and 'border zoom' are going to translate the
> viewport center to some extent.

Yes, it's going to change the position, but not the
depth relative to the camera

> Of course, I'm neither an interface expert or
> particularly good at coding
> either, so these are just thoughts!

Until someone codes it, yes, these are just thoughts
and ideas.

Martin

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com