The main part of this contribution to the special issue of EJM-B/Fluids dedicated to Patrick Huerre outlines the problem of the subcritical transition to turbulence in wall-bounded flows in its historical perspective with emphasis on plane Couette flow, the flow generated between counter-translating parallel planes. Subcritical here means discontinuous and direct, with strong hysteresis. This is due to the existence of nontrivial flow regimes between the global stability threshold Re g , the upper bound for unconditional return to the base flow, and the linear instability threshold Re c characterized by unconditional departure from the base flow.The transitional range around Re g is first discussed from an empirical viewpoint ( §1). The recent determination of Re g for pipe flow by Avila et al. (2011) is recalled. Plane Couette flow is next examined. In laboratory conditions, its transitional range displays an oblique pattern made of alternately laminar and turbulent bands, up to a third threshold Re t beyond which turbulence is uniform.Our current theoretical understanding of the problem is next reviewed ( §2): linear theory and nonnormal amplification of perturbations; nonlinear approaches and dynamical systems, basin boundaries and chaotic transients in minimal flow units; spatiotemporal chaos in extended systems and the use of concepts from statistical physics, spatiotemporal intermittency and directed percolation, large deviations and extreme values. Two appendices present some recent personal results obtained in plane Couette flow about patterning from numerical simulations and modeling attempts.Keywords Transition to Turbulence; Pipe Flow; Plane Couette Flow; Laminar-Turbulent Patterning Cautionary note about the literature cited The number of articles related to the topics examined here is tremendous and, though already referring to more than 150 publications, the bibliography is far from exhaustive by at least one order of magnitude. For a better coverage, the reader is invited to consult the literature cited in the review papers mentioned. I tried not to bias the list according to my personal interests, while choosing what I thought to be the most representative papers in each subtopic, sometimes the most recent publications of given people or groups reviewing related works in their introductions and pointing backwards to earlier relevant papers. For convenience, references are listed in alphabetical order of the first author and next chronologically.The article published by O. Reynolds in 1883 [131] founded the scientific approach to the problem of the transition to turbulence. Already an abstract in itself, its title "An experimental investigation of the circumstances which determine whether the motion of water shall be direct or sinuous and the law of resistance in parallel channels," summarized the main features of the problem and, between the words, identified its control parameter Re nowadays called the Reynolds number. This parameter is a measure of 1 arXiv:1403.6374v1 [physics.flu-dyn]