The glass/water interaction is reviewed from early experimental work in bulk samples to the current molecular interpretation of bond rupture in that interaction. The significance of that work to static fatigue in silica optical fibres is discussed. Assumptions concerning the basic equations used to predict the rate of crack propagation, and hence, the lifetime of silica optical fibres, are questioned. Issues such as flaw distribution and the true nature of the crack tip are highlighted along with deficiencies in test methods and data analysis techniques used to obtain fibre "lifetime" parameters. These factors have resulted in no single fibre lifetime model becoming universally accepted with some 15 published lifetime models currently available. It is suggested that lifetime theory based on large, well-defined cracks in bulk material is no longer entirely sufficient to explain the static fatigue behaviour of the nanometre-sized flaws found in current optical fibres. Instead, the literature indicates a two-stage model consisting of a precursor stage followed by the currently accepted bond-rupture mechanism, to be more appropriate.