I would like to know if you guys really use each http status code when you develop API's, cause I noticed that there are many API's that just use 404 when something goes wrong (ALL/ALMOST ALL cases). If I remember well there is a video here talking about that. Let me you know your opinion about that.
Making me able to simply look up a code quickly and see what when wrong (or just recall it from memory) is alot better than being expected to see some homemade error string that might be hard to decode.
An example is 429. I instantly know it its too many requests (instead of guessing the meaning of "you are too eager" or some other random text)
Oh yes agreed. I mean when using some external api. If they return a 404 code when the error is something else. By just looking at an error code, I often know the error instantly