/
open_questions.txt
35 lines (32 loc) · 1.62 KB
/
open_questions.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
so, fixing initial map loads of dynamic entities... my options are:
- add (a) new step(s) that trigger after the stale entities check
- will work, easy enough to make work
- may have complications
- won't solve future instances of similar problems
- hacky; "feels wrong"
- modify the location of the existing init so it happens after stale ent
- may have complications
- not straight-forward
- modify the stale ent so it happens before init
- same as above
- modify the stale ent call so it doesn't clear some stuff
- hacky
- modify the init call so its stuff won't get cleared by stale ent
- hacky
- modify the whole system so that stale and init don't clash
- no idea how to do this
- a couple days later: did this!
- got rid of "stale", it was an umbrella to make it so I wouldn't
have to think about when to change maps and how to load entities,
at a time when bidirectional communication between managers was
very hard to do without opening up holes big enough to drive
dump-trucks through.
Now that it's easy to (in a limited space) tap a manager on the
shoulder and ask it nicely to set up the entities, it's easy to
do so within the change_maps function. So stale_entities isn't
actually needed at all, and the calls to get entities aren't in
most places either.
... also, this resolved the problem! It required a bit more
intelligence around if/when a map is loaded, and I think
that method is getting called every frame, so... I might
want to find a way to lock that down. Later. :D