Pros: This is an awesome idea! I feel like Github's discovery is broken, and the fact that this includes data from Github as well as Reddit and HackerNews makes it really cool. I love that languages that get a little less love are included as well.
Cons: It looks like I'm hitting a rate limit for the Github API. The API request to Algolia is working fine, but then when the browser tries to load all the info about the Github repos, it's giving me 403s and saying `API rate limit exceeded for {{My public IP Address}}`. That might be because I'm at work and other people here are looking at this too, so one thing that might help would be having people log in with their Github account to up the rate limit. This could also maybe be improved by using the Github GraphQL API to get all the repo info in one request so that it's not doing an N+1 request.
@dnsbty Ah nice I didn't know about the Github GraphQL API! That's a nice idea and very efficient trick! Otherwise I saw one web app asking for user's github API token to increase the limit!
As somebody who spends all day at a computer, this makes a lot more sense to me than the iOS app (which I do have as well). Mind you, I'm not sure how it will affect my productivity. :-)
A filter for HN and Reddit only would be cool, I'm seeing a lot of foreign characters repo's in the top that look like spam that I'm assuming is coming from Github itself?
@wololodev I was thinking about adding this filter. The first 25 repos are coming from Github trending and then the next from HN and Reddit. There are very often repositories in Chinese in the top 25, it's not spam, it's just that the comments are in Chinese, we can't do much about, that it's based on Github algorithm.
Llamao
Llamao
Stay22
Llamao