Do John Denker's views show a disconnection between (1) academic, textbook
thermodynamics and (2) real-world engineering practices?
1) Traditionally, textbooks interpret dE=dQ+dW as a partitioning of energy
INPUTS to a system on the basis of the mechanism of entry/exit into/from
the system: eg., was it a Heat transfer across a temperature difference,
or was it Work?
I have argued that (so long as the total dE and the final state are
correct) the partitioning (1) is of little practical import in
irreversible processes, and is uniquely specified for reversible processes
by dQ=TdS. (dS is calculated either from its state function, or by
integrating dS=dQ/T over a reversible path between the given end states,
which path is freely choosable so as to make the identification of dQ
unquestionable.)
2) JD interprets dE=dQ+dW as a partitioning of energy CHANGES in a system
on the basis of the properties of the final system state: eg., was the
energy increment thermalized into "Heat" ("thermal energy") so that it
changed the system entropy, or was it Work (with no regard to the
mechanism of its entry/exit into/from the system)?
JD's approach never asks the "entry mechanism question" at all, but goes
directly to an examination of the final system state. He necessarily
marries the second law to (his interpretation of) the first law and
defines dQ=TdS as the increment in the system "thermal energy" or heat
(not a state variable!). Note that this is the dQ for a hypothetical,
reversible process between the end states; it is not the dQ which the
textbook model (1) would assign to the real process under consideration;
this latter quantity does not seem to be of any use or interest in JD's
approach.