[en] The International System of Units (SI) comprises at present seven base quantities and their units. (1) length meter (m); (2) mass - kilogram (kg); (3) time - second (s); (4) electric current - ampere (A); (5) thermodynamic temperature kelvin (K); (6) amount of substance - mole (Mol); (7) luminous intensity candela (cd). The choice of base quantities and units is a matter of convenience, convention and compromise. Temperature could be measured in units of energy. The mole is, ia fact, a number. The candela is a concession to photometry and is defined in terms of the first four units to which I confine myself. They are all defined by natural standards except for the kilogram which is still based on an artificial prototype and has an anomalous name comprising a multiplicative prefix. The discoveries of the Josephson and Quantum Hall Effects opened up new avenues to the definition of base units in terms of fundamental constants. When the Josephson and von Klitzing constants achieve a precision surpassing that of comparing weights the SI base units will have to be redefined and the traditional kilogram abandoned. I propose to choose as base units the quadruple: volt (V), ampere (A), meter (m), second (s). My reasons are as follows. Geometry involves only powers of length, hence its base unit is the meter. Kinematics introduces time and thus adds the second. Dynamics comprises mass, force, energy, power etc. The appropriate units can be conveniently defined in terms of the volt and ampere, the watt providing the bridge: W :=V·A. In all of mechanics V and A always occur only in the combination (VA). The great advantage of this scheme comes to light in electromagnetics. The electromagnetic duality (E↔H,B↔D, etc.) is simply expressed by the interchange V↔A The primary unit should remain the second as defined by the cesium (or any improved) atomic clock. The ac Josephson effect allows to define the volt via the Josephson constant whose agreed value at present is K_J90=4, 835979*10¹⁴ Hz/V. The integral quantum Hall effect defines directly the ohm in terms of the von Klitzing constant K_K-90 25812, 807Ω. However, it is preferable to define as base unit the ampere by A:=3D V/Ω = ({K_J})·{K_K})/(K_J·K_Ks). There can be no objection to this oblique definition. Even the meter is defined indirectly through the second and the speed of light in vacuum c. A variant of this scheme might utilize the weber (Wb≡Vs) and the coulomb (C≡As) rather than the volt and the ampere. The weber would then be given directly by tile inverse Josephson constant while the coulomb would be the inverse product of both K_J and K_K . This might be more appealing from a fundamental point of view since the base quantities would be the magnetic and electric charges (or fluxes) making the action their simple product. Yet I maintain that the VAms scheme is tile simplest and most advantageous from a practical point of view