02:24 Changeset [815] by Maciej Komosinski
Use double for consistency (using float to store double value of 0.2 would cause imprecision when 0.2f becomes 0.2 in Part shape properties)


21:17 Changeset [814] by Maciej Komosinski
Passes all arguments (if any) to python tester
20:35 Changeset [813] by Maciej Komosinski
Improved the script that runs unit tests
18:34 Changeset [812] by oriona
F1 genotype test added to geometry tests. Command line argument allowing to use valgrind added to go.sh.
17:24 Ticket #52 (ModelGeometry: test geometry measure using boundary test cases and valgrind) closed by oriona
fixed: In 811: […]
17:24 Changeset [811] by oriona
Tests for geometry and similarity added [fixes #52]


16:22 Ticket #51 (ModelGeometry: sqrt(negative)) closed by Maciej Komosinski
fixed: In 810: […]
16:21 Changeset [810] by Maciej Komosinski
Don't unnecessarily calculate sqrt(negative_value) [fixes #51]
15:06 Ticket #44 (ModelGeometry: for cuboids, sampling points should be located on edges) closed by Maciej Komosinski
fixed: The result is correct, sizexyz should not be 2.0 in this case as one might expect. For example, the first value ("length") is the longest diagonal in the cube with edge length of 2, so sqrt(12)=3.464101615137754587. The other values ("width" and "depth") are also analogously correct.


20:20 Ticket #54 (ModelGeometry and fF: fix conv_fF and scale<1 and find_hole==-1) closed by Maciej Komosinski
fixed: Most probably fixed in r667, just in case the LOG_ERROR message will be printed if find_hole() returned less than zero. Integration with ModelGeometry? postponed until ModelGeometry? becomes perfect and more flexible than it is now.
Note: See TracTimeline for information about the timeline view.