A new thermodynamic phase, the vortex-glass superconductor, is proposed in bulk, disordered, type-II superconductors. This phase lacks conventional off-diagonal long-ranged order but is a true superconductor with zero dc resistance. Metastable currents decay as $ (\ln t)^{-1/\mu} $, with $ \mu \leq 1 $ a universal exponent. The phase is relevant to experiments on high- $ T_c $ oxides. Disorder in type-II superconductors pins vortex lines, leading to low voltage (creep rate) and destroying translational long-ranged order of the Abrikosov flux lattice. In high- $ T_c $ oxides, significant hysteresis and irreversibility are observed, indicating disorder-induced disorder in the flux lattice. The original Anderson-Kim theory predicts a crossover from fast to slow dynamics but no sharp phase boundary. The author argues that in bulk disordered systems, a sharp equilibrium phase boundary exists, separating the normal phase from the flux-creep phase. Below this boundary, a new phase, the vortex-glass superconductor, exists. Glass phases in dirty superconductors have been suggested in numerical simulations and mean-field treatments. The author considers a model with microscopic pinning and finds strong evidence for a true vortex-glass phase in 3D but not 2D. This phase is characterized by a nonzero Edwards-Anderson order parameter but lacks conventional off-diagonal long-ranged order. It is a true superconductor with zero dc resistance. The mixed state of an extreme type-II superconductor is described by a Ginzburg-Landau Hamiltonian. The dual representation of the lattice version of this Hamiltonian in terms of 2D bosons allows study of disorder in the mixed state. The appropriate second quantized boson Hamiltonian is analyzed, with the partition function involving a trace over both the boson field and the gauge field. Three important energy scales are identified: pinning energy, interaction energy, and entropic contribution. High- $ T_c $ oxides with large $ T_c $ and modest $ H_{c1} $ are likely affected by entropic effects well below $ T_c $. For $ H > H_{c1} $, there is a competition between thermal fluctuations and pinning disorder. At high temperatures, thermal fluctuations dominate, leading to an unpinned vortex-liquid phase. Below this, pinning disorder and interactions dominate, leading to a vortex-glass phase. The vortex-glass phase is characterized by a nonzero Edwards-Anderson order parameter and is a true superconductor with zero dc resistance. The dynamics of the vortex-glass phase are studied, showing that metastable currents decay as $ (\ln t)^{-1/\mu} $. Experiments on bulk Y-Ba-Cu-O crystals indicate frequency-dependent "irreversibility" lines, suggesting aA new thermodynamic phase, the vortex-glass superconductor, is proposed in bulk, disordered, type-II superconductors. This phase lacks conventional off-diagonal long-ranged order but is a true superconductor with zero dc resistance. Metastable currents decay as $ (\ln t)^{-1/\mu} $, with $ \mu \leq 1 $ a universal exponent. The phase is relevant to experiments on high- $ T_c $ oxides. Disorder in type-II superconductors pins vortex lines, leading to low voltage (creep rate) and destroying translational long-ranged order of the Abrikosov flux lattice. In high- $ T_c $ oxides, significant hysteresis and irreversibility are observed, indicating disorder-induced disorder in the flux lattice. The original Anderson-Kim theory predicts a crossover from fast to slow dynamics but no sharp phase boundary. The author argues that in bulk disordered systems, a sharp equilibrium phase boundary exists, separating the normal phase from the flux-creep phase. Below this boundary, a new phase, the vortex-glass superconductor, exists. Glass phases in dirty superconductors have been suggested in numerical simulations and mean-field treatments. The author considers a model with microscopic pinning and finds strong evidence for a true vortex-glass phase in 3D but not 2D. This phase is characterized by a nonzero Edwards-Anderson order parameter but lacks conventional off-diagonal long-ranged order. It is a true superconductor with zero dc resistance. The mixed state of an extreme type-II superconductor is described by a Ginzburg-Landau Hamiltonian. The dual representation of the lattice version of this Hamiltonian in terms of 2D bosons allows study of disorder in the mixed state. The appropriate second quantized boson Hamiltonian is analyzed, with the partition function involving a trace over both the boson field and the gauge field. Three important energy scales are identified: pinning energy, interaction energy, and entropic contribution. High- $ T_c $ oxides with large $ T_c $ and modest $ H_{c1} $ are likely affected by entropic effects well below $ T_c $. For $ H > H_{c1} $, there is a competition between thermal fluctuations and pinning disorder. At high temperatures, thermal fluctuations dominate, leading to an unpinned vortex-liquid phase. Below this, pinning disorder and interactions dominate, leading to a vortex-glass phase. The vortex-glass phase is characterized by a nonzero Edwards-Anderson order parameter and is a true superconductor with zero dc resistance. The dynamics of the vortex-glass phase are studied, showing that metastable currents decay as $ (\ln t)^{-1/\mu} $. Experiments on bulk Y-Ba-Cu-O crystals indicate frequency-dependent "irreversibility" lines, suggesting a