- User Since
- Aug 16 2010, 8:06 PM (474 w, 2 d)
Sep 18 2017
One option I liked on the old background images is being able to change the render order to switch having the image always be in front of everything.
Mar 1 2016
I just realized I still had this task open. I'm pretty sure everything works as intended and has been for a while now. Seperate bug reports may still follow if I ever encounter similar weirdness again in the future, but I don't believe this patch serves any public purpose anymore. Closing for now.
Jul 3 2015
Bastien, you are slowly becoming my favourite dev here. ;) Thanks for this quick fix!
@perfection cat (sindra1961) Thank you for helping test this. =) I think you may have gotten lucky with your -resize test though; it seems completely unpredictable and seemingly "random" on my end. So far, around 3 out of 100+ images fails with no apparent link between them, all created the same way.
The PNG has not gone through OCIO. I have a Blender script that calls ImageMagick commands using os.system(). The first path is the same for all images:
Jun 28 2015
Awesome, thank you for this quick and proper fix! =D
Jun 27 2015
@Bastien Montagne (mont29) Anytime! I'm not sure I'd call partial exotic (it's a pretty commonly-used technique isn't it?), but I definitely understand wanting to have a more straight-forward solution for this.
I went ahead and made an attempt at fixing this. It seems to work for my needs so far, but I'm not sure how nice this solution actually is. What I've done is I made bl_obj a @property and added a callback list bl_obj_process_queue.
Jun 26 2015
Figured I'd play with this a bit. I am by no means sure, but here's what I think is happening, based on some tests and print()s. Assuming an FBX with the following layout:
Jun 24 2015
May 23 2015
I understand a decision has been made, and I agree this is better than nothing, but I still believe a scene DnD operator would be very beneficial. So just for future reference I would like to address some of these points one final time:
May 18 2015
+1 on that, for sure. This seems to me like the most useful and complete setup. Holding alt on drop for OSX (or even Windows and Linux) sounds like a great solution as well. No control is lost, and it can be a huge speedup.
May 17 2015
RMB DnD might be probematic for most OS's (Windows and OSX open the context menu directly iirc). That said, I really like:
May 16 2015
After giving this some thought, here's my initial $0.02 on the usability side of things:
Apr 19 2015
@Aaron Carlisle (Blendify) I'm sorry, but I'm afraid tablet input is really beyond the scope of this patch (and my abilities; I'm not very familiar with the Blender source).
Apr 18 2015
And just after posting this I remembered what the original problem was: Two-finger scroll gets picked up as a pan event, not a scroll one. I would love some input on how to best tackle this, or even how to look at it. I feel like adding a toggle just for this switch would now be overkill, but I'm also not sure how else this could be handled nicely. Perhaps some kind of modifier toggle or secondary Value to treat Pan as either 2D Pan, Horizontal Scroll or Vertical Scroll?
Alright, so I'm looking at the current state of Blender and what this patch was meant to accomplish. The goal of this patch was simply to allow "classic" scene navigation using a multi-touch trackpad. For me this only meant having a two-finger scroll to zoom.
Apr 17 2015
Apr 4 2015
Woah, this *is* old! This is actually one of the reasons I still use a modified 2.67 build myself (I have a physical handicap and the current behaviour of pinching and rotating is extremely hard for me to do). At some point the patch stopped working though and I couldn't be bothered to fix it.
Mar 22 2014
Feb 17 2014
Also +1 from me; center-based manipulation feels far more natural i.m.h.o.
Feb 15 2014
Ah, I can see how that could cause some confusion, my apologies! I'll be using just the feature name as headers in the future to prevent this. =)
@Bartosz Moniewski (monio) I may have missed a few, but doing a cmd+F search for "cursor depth" finds me: