When something breaks, the app usually left a note explaining what went wrong and where. That note is in the log.
An error log is the app's running diary: a timestamped list of what it did and what failed. When it breaks, you read the log first.
Wikipedia: "logging is the act of keeping a log of events that occur in a computer system, such as problems, errors or broad information on current operations." The log is that record, written as things happen.
A log line usually carries a timestamp, a level (info, warning, error) and a message. An error line often names the file and line where it failed, and you don't need to code to read it: "cannot read property 'email' of undefined" means something expected an email and got nothing.
Every host has a logs panel (Cloudflare, Vercel and Supabase all do). We rely on ours: when this site returned server errors, they showed up in the host's logs as resource-limit errors, which is how we knew it was the CPU ceiling, not a code bug.
The log is also how you talk to your AI tool about a break. Paste the exact error line in and ask what it means and what causes it. That turns a vague "it's broken" into a specific, answerable question.
Try it with the rules off. Picture two bug reports for the same problem: "it doesn't work," versus the exact error line with a file name and a timestamp. Only one of them can be fixed in a minute, and it's the one that came from the log.
Check yourself
0/4 got itSaved on this device only. No account, no streaks.
Next in when it breaks: What is latency?.




