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
It would be great if the love service exposed an endpoint where you can get an aggregate count of ALL users send/received love counts over a time period. The query would be something conceptually like SELECT user, COUNT(sent), COUNT(received) FROM table WHERE date >= @start AND date <= @end
The text was updated successfully, but these errors were encountered:
The ugly truth of Google Query Language is that typical aggregation functions like COUNT(), SUM(), MAX(), etc. are not even included. To do something as simple as the leaderboard, there has to be a separate table holding weekly love counts. So this operation would require us to iterate over every love sent in the time range and count them up. Which is probably feasible for a small enough time range, but could end up running long and potentially consuming lots of memory, which is a problem because GAE kills requests after 10 seconds, or if they exceed 128 MB of memory.
It's objectively a good idea for a feature, but those are some of the problems we might run into while implementing it.
One thing I'd like for us to do eventually is move off of the old proprietary App Engine APIs. If we migrate to the flexible environment we could use a different datastore that has these capabilities. We could also move to Python 3.
It would be great if the love service exposed an endpoint where you can get an aggregate count of ALL users send/received love counts over a time period. The query would be something conceptually like
SELECT user, COUNT(sent), COUNT(received) FROM table WHERE date >= @start AND date <= @end
The text was updated successfully, but these errors were encountered: