It's beginning to dawn on me that the whole
OOP/functional/aspect-oriented/etc. scene is about software engineering, not
just about egos and pet languages.
My software engineering education kind of stopped with Structured System
Design á la Yourdon and Wirth's structured programming mantras in the 70's.
After that, I saw fads come (and and almost as often go: AI, 4th-generation
DBMS's, talking paperclips) but didn't see anything that made me want to
change my basic take on programming.
Then I got into Python about a year ago and it's been a bit of a rocket
ride, educationally. In the beginning, the air was a little too thin at the
requisite levels of abstraction and took some getting used to, and will
probably remain the biggest barrier to entry. It's actually the second
Revenge of the Nerds, and a Darwinian selection process is taking place as
we speak: I think a lot of people didn't make the transition from COBOL to
C, and another group will have trouble going from pointers to references and
making sense of the multiheaded XML hydra.
Add to this the fact that many people now believe that you can teach
object-oriented languages without recapitulating the conceptual models
underlying the whole progression: toggled-in machine language routines
wriggling in the primordial ooze of absolute machine addresses, relocating
assemblers darting about the shallows, "high-level" languages like FORTRAN
and COBOL lumbering out of the primal memory sea and expiring on the land,
clever amphibians like C scurrying on the shore but returning to the memory
sea to lay their pointery little eggs, and finally the Pythons and Javas
that preserve only a faint echo of their marine origins in the IDs of their
objects.
I've taught computer languages professionally, and, as the age gap between
myself and my students widens, I feel more and more guilty that I had the
big advantage of having seen so much of the evolution happen. It helps me
appreciate both what's great about what we have now, and why certain
limitations are still there. The new folks don't have that advantage unless
they do enough assembler programming to understand how things look from the
instruction set level.
'Nuff said about that. Many thanks to everyone at the python and twisted
lists for their help as I reinvent myself again.
OOP/functional/aspect-oriented/etc. scene is about software engineering, not
just about egos and pet languages.
My software engineering education kind of stopped with Structured System
Design á la Yourdon and Wirth's structured programming mantras in the 70's.
After that, I saw fads come (and and almost as often go: AI, 4th-generation
DBMS's, talking paperclips) but didn't see anything that made me want to
change my basic take on programming.
Then I got into Python about a year ago and it's been a bit of a rocket
ride, educationally. In the beginning, the air was a little too thin at the
requisite levels of abstraction and took some getting used to, and will
probably remain the biggest barrier to entry. It's actually the second
Revenge of the Nerds, and a Darwinian selection process is taking place as
we speak: I think a lot of people didn't make the transition from COBOL to
C, and another group will have trouble going from pointers to references and
making sense of the multiheaded XML hydra.
Add to this the fact that many people now believe that you can teach
object-oriented languages without recapitulating the conceptual models
underlying the whole progression: toggled-in machine language routines
wriggling in the primordial ooze of absolute machine addresses, relocating
assemblers darting about the shallows, "high-level" languages like FORTRAN
and COBOL lumbering out of the primal memory sea and expiring on the land,
clever amphibians like C scurrying on the shore but returning to the memory
sea to lay their pointery little eggs, and finally the Pythons and Javas
that preserve only a faint echo of their marine origins in the IDs of their
objects.
I've taught computer languages professionally, and, as the age gap between
myself and my students widens, I feel more and more guilty that I had the
big advantage of having seen so much of the evolution happen. It helps me
appreciate both what's great about what we have now, and why certain
limitations are still there. The new folks don't have that advantage unless
they do enough assembler programming to understand how things look from the
instruction set level.
'Nuff said about that. Many thanks to everyone at the python and twisted
lists for their help as I reinvent myself again.