aboutsummaryrefslogtreecommitdiff
path: root/notes
diff options
context:
space:
mode:
authorrodri <rgl@antares-labs.eu>2021-08-01 11:56:12 +0000
committerrodri <rgl@antares-labs.eu>2021-08-01 11:56:12 +0000
commit7ed124593ebda443fc4d0f1439d1fd8dc602ce72 (patch)
tree9790ce5c3436e4b493ca912d26c6baa1bb8b826b /notes
parentb57dba67153279a84a16bba40ae007c8ae710056 (diff)
downloadmusw-7ed124593ebda443fc4d0f1439d1fd8dc602ce72.tar.gz
musw-7ed124593ebda443fc4d0f1439d1fd8dc602ce72.tar.bz2
musw-7ed124593ebda443fc4d0f1439d1fd8dc602ce72.zip
began building a universe for every party.
get rid of the notes file. i already keep them on paper. adapted the client to use an rframe for drawing. implemented some VModel methods for testing. FPS2MS → HZ2MS. it makes more sense and is shorter.
Diffstat (limited to 'notes')
-rw-r--r--notes17
1 files changed, 0 insertions, 17 deletions
diff --git a/notes b/notes
deleted file mode 100644
index 26e05fb..0000000
--- a/notes
+++ /dev/null
@@ -1,17 +0,0 @@
-• there's, at most, two players waiting in the lobby at a given time.
-it makes no sense to allocate more than two seats since the threadsim
-will consume them whenever they are ready to join the party. i'm
-thinking of using channels to synchronize the two threads, so
-threadsim doesn't loop doing nothing (but sleeping) until at least a
-couple of players join.
-
-• the integrator has to operate with vectors and the different objects
-in the universe, some of which may require their own governing laws.
-
-• think of a way to pack the bullets efficiently. will they be part
-of the global state broadcast? what does the client need to know to
-render and manage them?
-
-• it could be beneficial to do dynamics in the client as well, which
-means sending more data, and probably require a tighter sync, but a
-smoother user experience.