I had to rewrite our entire scheduling system at work to use Outlook instead of Google Calendar. The guy who wrote the Google Calendar scheduling system made it so unmaintainable that it was faster to just rewrite the entire thing from scratch (1000+ line lambda function with almost 0 abstraction).
At least 90% of what I wrote is just exception handling. There’s ~15 different 4xx/5xx errors that can be returned for each endpoint, but only 1 or 2 200 responses.
This is fair, but it’s at least broken up so they can selectively gut the parts of it they don’t like instead of having to figure out what a 300 line method named “process” does.
I had to rewrite our entire scheduling system at work to use Outlook instead of Google Calendar. The guy who wrote the Google Calendar scheduling system made it so unmaintainable that it was faster to just rewrite the entire thing from scratch (1000+ line lambda function with almost 0 abstraction).
At least 90% of what I wrote is just exception handling. There’s ~15 different 4xx/5xx errors that can be returned for each endpoint, but only 1 or 2 200 responses.
I bet in the future someone who will see your code will also think of the same. Just the nature of things.
This is fair, but it’s at least broken up so they can selectively gut the parts of it they don’t like instead of having to figure out what a 300 line method named “process” does.