daniel wrote:We'll just have to find something else than 'ALT' (it's already used to select multiple entities).
ALT is the modifier that is probably most commonly used for this, at least for DCC applications. It's what Maya uses  and nearly everybody else has either adopted that convention or at least offers an alternative navigation scheme that matches Maya. In the realm of point cloud software Sequoia is the only one off the top of my head that offers "Maya like" navigation.
That being said, I'm not a big fan of ALT since your hand ends up in a weird position if you want to keep your other fingers on the home row, say to easily access other shortcuts.
daniel wrote:Moreover, in the 'interactive segmentation tool' and the 'interactive transformation tool', you can already 'pause' the tool (with the space bar) to move the camera and then restart the tool (hit the space bar again).
The space bar would be my personal preference over ALT. It's what Houdini uses  and it keeps your hand in a much more comfortable position. Conceptually it also makes sense, when typing space is probably the most commonly used key and when working in 3D, navigation is probably the most commonly used function. In the point cloud realm PolyWorks uses the space bar for navigation, but like the interactive tools in CC it is a toggle between navigating and interacting. A toggle works OK, but I think using a continuous press offers some benefits. For example, if a lower resolution version of the cloud is being rendered during navigation releasing the modifier is a clear signal that navigation is done and I want to see the full resolution model, even if I'm not ready to use the interactive tool yet. Using press and hold as a modifier also keeps the possibility open of using a short press on the space bar to do something else important down the road. Maybe pop up a tool palette under the cursor.
daniel wrote:And last but not least, I'm not sure we should let this happen in the 'interactive segmentation tool' as the segmentation polyline is a 2D entity linked to the screen (in pixels) and if you move the object / camera in the 3D world, the polyline won't have a meaning anymore...
I wouldn't immediately discount this behavior. It is exactly how Mari's Paint Buffer works  and can be really useful. Instead of directly painting/selecting in 3D you actually work on an invisible 2D plane in front of the camera. Your 2D work then gets projected, i.e. "Baked', to 3D as a separate step. Although with the default setting this extra step is largely invisible to the user. That two step process offers the benefit that you don't need to get your 2D work perfect the first time. For example, if the initial 2D polyline I draw is off by a few pixels in one area, instead of drawing a new one maybe I just move the camera/object a little bit so that it fits within the outline. Doing the projection as a separate step also offers some performance benefits. You can build up a potentially complex 2D selection in multiple steps and only commit to the expensive 2D-to-3D projection once you're happy with it.
I haven't looked at all of the permutations, but it seems like the hold-modifier-to-nav paradigm could be made to work with the segment tool's current behavior. However, with some tweaks I think that tool could also be made much more efficient. That's a topic for another day though.
 http://help.autodesk.com/view/MAYAUL/20 ... ew_hotkeys