September 2, 2017, 15:39 (GMT) |
Fixed inverse tube bug. Minor normal issue remains. |
August 29, 2017, 14:51 (GMT) |
Fixed a really hard to find crasher bug and a lot of mesh issues! |
August 29, 2017, 13:35 (GMT) |
Added clamping to input settings. |
August 29, 2017, 13:21 (GMT) |
Mini fixes. Catch some cornercases that shouldn't happen. |
August 28, 2017, 15:34 (GMT) |
Fixed a crasher, mutex was missing initialisation. |
August 25, 2017, 10:18 (GMT) |
Possibly fixed a big portion of the ring missalignments and uncomplete rings. |
August 25, 2017, 08:28 (GMT) |
Fixed some normal orientation issues. B-Ring normals are now aligned A-ring still needs normal alignment. |
August 24, 2017, 15:10 (GMT) |
Mayor rework of the edgering system. Unified code, bugfixes. A lot of bugs still remain. |
August 22, 2017, 16:03 (GMT) |
Fix mini crasher, last commit before mayor rework of the edgering detection/sort. |
August 22, 2017, 09:13 (GMT) |
Fix holes bug in fillets. |
August 20, 2017, 14:49 (GMT) |
Compiler pointer arithmetic fix. |
August 20, 2017, 10:22 (GMT) |
Fix compiler error on win. (Arithmetic on void pointer) |
August 19, 2017, 09:07 (GMT) |
Added subtract mode. "Clipping brush" |
August 18, 2017, 22:35 (GMT) |
Fillets now get generated. Almost feature complete. |
August 18, 2017, 14:31 (GMT) |
Improved triangulation, working on ring assignments. |
August 17, 2017, 14:39 (GMT) |
Added first triangulation to generate the fillets. Still needs ring order and assignment though. |
August 16, 2017, 12:07 (GMT) |
Most data is now exposed to be used for final transitions/fillets. Exact point order can be improved since some are discarded. (todo) |
August 15, 2017, 10:52 (GMT) |
Fixing flooding bug, partial fix. Found BB related flood cause. Realloc fix. |
August 15, 2017, 07:58 (GMT) |
Inner vertices now get calculated as well as edgerings. Problem with holes in the edgerings. |
August 14, 2017, 11:01 (GMT) |
First version calculating the inside and outside vertices. Needs multiple ring support. Holes from earlier misscalculations break the "floodfill" algorithm. |
|