They do put the disliked profiles in S3 storage for extra data analysis. It's possible they store it in a cache or a data structure to prevent showing the same profile twice.
Won't that require too much storage ? Storing all the pairs of likes and dislikes ? That is order of O(n^2). That's too much memory required ? How is tinder solving for that ?
Hi, is tinder not storing the disliked (left swiped) information ? Otherwise, how will tinder know to not show the profile again to the user ?
They do put the disliked profiles in S3 storage for extra data analysis. It's possible they store it in a cache or a data structure to prevent showing the same profile twice.
Won't that require too much storage ? Storing all the pairs of likes and dislikes ? That is order of O(n^2). That's too much memory required ? How is tinder solving for that ?
I'd assume they store the 'profile ID' in a hash data structure on the cache server.
Or it could also be a probabilistic data structure like bloom filter.
Yet I don't think the space complexity is O(n^2), probably each person dislikes a limited number of people compared to the total users of Tinder.
stop asking awkward questions, the author of the article has no idea how it actually works