A review is given of observations and theories relevant to the solar flare of 21 May, 1980, 20 ∶ 50 UT, the best studied flare on record. For more than 30 hr before the flare there was filament activation and plasma heating to above 10 MK. A flare precursor was present ≥6 min before the flare onset. The flare started with filament activation (20 ∶ 50 UT), followed by thick-target heating of two footpoints and subsequent ablation and convective evaporation involving energies of 1 to 2 × 1031 erg. Coronal explosions occurred at 20 ∶ 57 UT (possibly associated with a type-II burst) and at 21 ∶ 04 UT (associated with an Hα spray?). Post-flare loops were first seen at 20 ∶ 57 UT, and their upward motion is interpreted as a manifestation of successive field-line reconnections. A type-IV radio burst which later changed into a type-I noise storm was related to a giant coronal arch located just below the radio noise storm region. Some implications and difficulties these observations present to current flare theories are mentioned.