- User Since
- May 21 2014, 3:36 PM (320 w, 5 d)
Jan 26 2020
- Make resolve_path backwards compatible with 3.5
I finally had time to work on this. See D6676
Jan 7 2020
Ok, makes sense. I'll try to get a patch for that in the next few days.
Jan 3 2020
- Don't reimport Path
- Skip the symlink path test on Windows
Jan 2 2020
- Alphabetize imports in packing test file
- Test that all paths are rewritten in symlink test
Dec 31 2019
I updated the description and added a test. That's a good question about Windows. Unfortunately, I'm unable to test it at the moment. It already uses the resolved path for all the other assets, so if it would be an issue, it would probably already have problems without this patch.
- Add unit test for resolving symlinks
- Remove redundant path resolve/absolute calls
Dec 30 2019
Jun 19 2018
Jun 5 2018
@Sybren A. Stüvel (sybren) I updated the description to include a .zip file with some backed caches.
May 8 2018
Mar 27 2018
Jan 21 2018
Ok, so it seems that my attempt at recreating the original error was sidetracked by an identical error. I think I managed to figure out what caused the original issue.
Jan 17 2018
Weird, I'm on Windows and when I run:
Jan 16 2018
Jan 14 2018
Dec 15 2017
Jun 1 2017
I've tested it with the svn backend (using a localhost svn server) and it seems to work exactly the same. I've also written a mercurial backend that works as expected. I'd like to use BAM in my own workflow but that requires mercurial support which was the motivation behind this patch.
May 31 2017
I plan on implementing mercurial as a backend. It has fairly good support for binary files and I'm already using it for personal projects (I works out nicely because I can host them on Bit Bucket). I can't give a definite answer for how long I can maintain it, but I plan on staying around for the near future.
May 29 2017
May 25 2017
I updated it to raise BamException instead of calling fatal(). It would be very easy to add more specific sub-classes of BamException and handle those differently if needed. For now I just kept everything as BamException.
May 24 2017
So you are saying fatal () would just raise a BamException (or maybe BamFatalError) but the printing of those errors is still handled in bam_cli.py?
I updated it with the requested changes.
Update with requested changes
May 22 2017
Dec 29 2016
Aug 17 2016
Jul 16 2016
The main reason why vertical nodes would be useful is that they take up less space than horizontal nodes which could be important in some situations (e.g. stringing together lots of math nodes to create a simple math formula). It's also worth mentioning that node-based software (Houdini, Nuke, etc.) tend to use vertical nodes.
Jul 4 2016
Jun 11 2016
I rewrote BM_face_calc_plane to call BM_face_calc_tangent_edge for ngons and BM_face_calc_tangent_edge_pair for quads. It still calculates the tangent for triangles the same way (uses the shortest edge), not sure if it makes more sense to use BM_face_calc_tangent_edge_pair for that too.
Jun 10 2016
I made the requested changes (plus a little cleanup)
Jun 9 2016
Jun 8 2016
@Campbell Barton (campbellbarton) I've implemented those four functions, while doing some testing I thought it might be a useful to add a fifth function calc_tangent_edge_center_diagonal which would use the vector going from the farthest vert to the center of the edge (useful for tris and odd sided ngons) Thoughts?
Jun 4 2016
Should calc_tangent_edge_pair use the longest edge and the second longest edge that isn't connected to the longest? Or should it use the two edges that have the longest length when added together? E.g. in the image below, the red edges when added together are longer than the blue edges, but the longest edge in the entire shape is blue.
Also, what should it do in the case of a triangle where there are no two disconnected edges?
May 25 2016
Would it make more sense to have the H hotkey be cyclic rather than having the Ctrl+Alt+V hotkey? E.g. Press H it hides the node, press it again it makes the node vertical, press it again it resets it back to a normal node.
May 19 2016
May 17 2016
@Brendon Murphy (meta-androcto) I'm probably not going to implement it the C code, but I did write a simple python script that should work just as well:
May 11 2016
@Brendon Murphy (meta-androcto) You're saying it would snap the objects and the cursor to the center? I can see how that could be useful, but it seems like a bit of a corner case. Either way I would like to get some more input form people before implementing it, maybe you could open a feature requrest on rightclickselect.com?
May 5 2016
The reason I implemented Snap Selection to Active is because I found myself frequently doing Snap Cursor to Selection then selecting a bunch of other stuff and doing Snap Selection to Cursor. I don't think snapping stuff to the center is really that common (most of the time Alt-G will suffice). That said, if it would be helpful, then I'd be more than happy to implement it :)
May 3 2016
Apr 8 2016
I just realized that when using the solidify tool in edit mode on the exact same mesh with the same thickness, it works just fine without any intersections. I would expect these tools to work the same...
Apr 4 2016
Mar 18 2016
Aug 28 2015
That should work perfect(maybe even better then my original idea :) Thanks for the clarification!
Forgot to attach the file:
Dec 18 2014
Everything works fine now! Thank you to everyone who looked into this issue :)
Dec 10 2014
The 64 bit version still crashes, but the 32 bit version works fine.
Sep 27 2014
It seams to be some kind of platform issue. I am dual booting Windows 8 and Ubuntu Linux 14.04. I tried the test file in Ubuntu and everything worked fine.
Oops typo, meant to say 2.71. However I did just test it with 2.72 RC and it still crashed. Something interesting to note is that with 2.71 it crashed right away when I pressed render. But with 2.72 RC it finished building the BVH, then it crashed.