5 Errors in my cata_tests result

[impossible]

Do they affect the gameplay a lot?

make -j2 NATIVE=osx RELEASE=1 LOCALIZE=1 LUA=1 MACPORTS=1 CLANG=1 check

Check distribution of attacks to body parts for same sized opponents. completed in 0.10982s
Check distribution of attacks to body parts for smaller attacker. completed in 0.113276s
Check distribution of attacks to body parts for larger attacker. completed in 0.112718s
Test bounds for mapping x/y/z/ offsets to direction enum completed in 0.000429s
line_to() executed 10000 times in 9654 microseconds.
canonical_line_to() executed 10000 times in 23656 microseconds.
Compare line_to() to canonical line_to() completed in 0.051139s
Radius one 2D square centered at origin. completed in 6.1e-05s
Radius one 2D square centered at -4/-4/0. completed in 4.8e-05s
Radius three 3D square centered at 8/8/1. completed in 0.001872s

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
cata_test is a Catch v1.0 b52 host application.
Run with -? for options

-------------------------------------------------------------------------------
Player body temperatures converge on expected values.
  Nude target temperatures.
-------------------------------------------------------------------------------
player_test.cpp:65
...............................................................................

player_test.cpp:43: FAILED:
  CHECK( high > p->temp_cur[0] )
with expansion:
  9600 > 9633

Completed in 0.689506s
Player body temperatures converge on expected values. completed in 0.689653s
-------------------------------------------------------------------------------
Player body temperatures converge on expected values.
  Lightly clothed target temperatures
-------------------------------------------------------------------------------
player_test.cpp:65
...............................................................................

player_test.cpp:43: FAILED:
  CHECK( high > p->temp_cur[0] )
with expansion:
  8100 > 8172

player_test.cpp:43: FAILED:
  CHECK( high > p->temp_cur[0] )
with expansion:
  9600 > 9627

Completed in 0.760556s
Player body temperatures converge on expected values. completed in 0.760623s
-------------------------------------------------------------------------------
Player body temperatures converge on expected values.
  Heavily clothed target temperatures
-------------------------------------------------------------------------------
player_test.cpp:65
...............................................................................

player_test.cpp:43: FAILED:
  CHECK( high > p->temp_cur[0] )
with expansion:
  9600 > 9605

Completed in 1.64365s
Player body temperatures converge on expected values. completed in 1.64381s
-------------------------------------------------------------------------------
Player body temperatures converge on expected values.
  Arctic gear target temperatures
-------------------------------------------------------------------------------
player_test.cpp:65
...............................................................................

player_test.cpp:43: FAILED:
  CHECK( high > p->temp_cur[0] )
with expansion:
  9600 > 9616

Completed in 2.65271s
Player body temperatures converge on expected values. completed in 2.65291s
Reading a legacy overmap save. completed in 0.211795s
Regression test against old shadowcasting implementation. completed in 0.001124s
===============================================================================
test cases:   11 |   10 passed | 1 failed
assertions: 1952 | 1947 passed | 5 failed

Test world "Alton" left for inspection.
make[1]: *** [check] Error 5
make: *** [check] Error 2

[kevin.granade]

Hmm, I though I had those hammered out, thanks for reporting.

In game terms, the impact of those is negligible, they’re only a little off.
The test is trying to check if player body temperature is converging to the temperature we expect it to be.

In slightly more detail, it creates a player, sets their body temperature to normal, dresses them up in a pre-selected list of clothes, sets the temperature to a test value, then waits until the player’s temperature stabilizes. As you can see the values are close to expected, off by 33, 72, 27, 5 and 16 units. For context the full range is 0 - over 10,000, so being off by less than 100 units isn’t a big deal. It is a little worrying for the test itself though, because it means it can’t reliably tell us when something goes wrong, we’ll have to tweak it to be more reliable.

EDIT: By the way, thanks for checking them out and reporting, the tests aren’t stable enough just yet, but when they are I’m planning on asking people to run them on their computers in case there are weird differences. Early feedback is good though

[impossible]

Glad they are not a big issue now. I’ll keep an eye on your progress.