You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
For the anti map repetition code to work we need a way of telling when a game was started from a particular queue. Right now there is only one queue games started from it have the gametype ladder1v1. But for future tmm queues this will not be the case.
Proposed solutions
Add a nullable matchmaker_queue_id column to game_stats table.
The problem is that game_stats is a big table, and the majority of games will have a null value here. This is a lot of wasted space.
Introduce some sort of m:n table for linking queues to games.
Takes less space, but requires an extra join when querying. Can also link multiple queues to the same game, but I'm not sure what the use case for that would be.
The text was updated successfully, but these errors were encountered:
* Add matchmaker_queue_game table
* Write originating mm_queue on game start
Not using it for recent match lookup yet
* Use matchmaker_queue in LadderService.get_game_history
* Temporarily add migration to travis
* Fix flake8 checks
* Add id column to matchmaker_queue_game table
* Store and use matchmaker queue id
* Remove manual migration
* Dont shadow id in function parameter
* Add tmm2v2 ladder history tests
* Update DB_VERSION for travis
For the anti map repetition code to work we need a way of telling when a game was started from a particular queue. Right now there is only one queue games started from it have the gametype
ladder1v1
. But for future tmm queues this will not be the case.Proposed solutions
matchmaker_queue_id
column togame_stats
table.The problem is that
game_stats
is a big table, and the majority of games will have a null value here. This is a lot of wasted space.Takes less space, but requires an extra join when querying. Can also link multiple queues to the same game, but I'm not sure what the use case for that would be.
The text was updated successfully, but these errors were encountered: