Hyper Text Markup Language (HTML) Question:

HTML validators check HTML documents against a formal definition of HTML syntax and then output a list of errors. Validation is important to give the best chance of correctness on unknown browsers (both existing browsers that you haven't seen and future browsers that haven't been written yet).
HTML checkers (linters) are also useful. These programs check documents for specific problems, including some caused by invalid markup and others caused by common browser bugs. Checkers may pass some invalid documents, and they may fail some valid ones.
All validators are functionally equivalent; while their reporting styles may vary, they will find the same errors given identical input. Different checkers are programmed to look for different problems, so their reports will vary significantly from each other. Also, some programs that are called validators (e.g. the "CSE HTML Validator") are really linters/checkers. They are still useful, but they should not be confused with real HTML validators.
When checking a site for errors for the first time, it is often useful to identify common problems that occur repeatedly in your markup. Fix these problems everywhere they occur (with an automated process if possible), and then go back to identify and fix the remaining problems.
Link checkers follow all the links on a site and report which ones are no longer functioning. CSS checkers report problems with CSS style sheets.