We consider the question of the concepts of entropy and temperature in arbi
trary nonequilibrium conditions in the framework of the so-called Informati
onal Statistical Thermodynamics. This is the approach to Thermodynamics bas
ed on the statistical-mechanical foundations provided by a Gibbs ensemble-l
ike algorithm in nonequilibrium situations. The resulting nonequilibrium te
mperature-like variable dubbed as quasitemperature - is shown to be a quant
ity measurable with appropriate "thermometric devices". A comparison of qua
sitemperatures that arise in different approximated nonequilibrium statisti
cal-thermodynamic descriptions of the dissipative system is done. The valid
ity of these different approximations is evaluated, and (in the framework o
f the theory) generalized Gibbs, Clausius, and Boltzmann's relations, as we
ll as properties of the corresponding entropy-like function (or information
al entropy in Jaynes-Shannon sense), that the theory introduces, are presen
ted. Conceptual and physical aspects of the question an also discussed, and
a partial comparison of these concepts with those arising in other approac
hes to irreversible thermodynamics is briefly attempted.