T
TonyMc
As we all now use widescreen televisions, XBoxes and Kinects, I would
like to propose a new and potentially much more efficient method for C
programming. Instead of boring old keyboards and ASCII characters, what
about the use of *gestures* to enter our program code? I know that
similar (but much less adventurous) suggestions have been made recently,
and I acknowledge that I have learned from them, but everyone should
realise that this is really a dramatically new developlment in the
history of the C programming language, and the standards committee will
be forced by the sheer beauty and logical necessity of my proposal to
halt work immediately on the forthcoming standard and devote all their
attention and effort to realising my proposal immediately.
Proposal: in future, all C programs should be entered by standing in
front of the TV and making appropriate (and occasionally inappropriate)
gestures. The Kinect will then interpret these gestures and translate
the code into an intermediate form suitable for input to a C compiler.
For example, if I hold my right middle finger like this, the software
should insert a "not-equals" symbol, which may or may not be the same as
the digraph != . If I wave my left hand like this, I evidently wish to
invoke the strlen() function on the string my right hand is pointing to.
(Note: the immediacy and intuitiveness of the connection between pointers
and the act of "pointing at" should simplify discussion on comp.lang.c
enormously.) If I turn my back on the television and expose my behind,
the software should insert "void main()". If I smile at the screen, my
evident intention is "return EXIT_SUCCESS", while a frown would signify
"return EXIT_FAILURE". If I were to put my left hand... here... and my
right arm twisted like... this... (sorry, I am no spring chicken and
such contortions are not without a little discomfort) the editor would
insert a call to a Batcher's odd-even merge (recursive version) on an
array of "const investment_t*" items. At this point the efficiency
advantage of this proposed method becomes compelling.
In anticipation of some objections, I must acknowledge here that there
may be some disadvantages to the proposed method of coding. I ignore
the fact that there may be some Luddite programmers who refuse to
embrace new technology, their objections are based on prejudice and
paranoia and will not be countenanced. I also ignore all objections
from people who disagree with me - that is sufficient grounds for
considering them incompetent. However, I am keen to allow proper
debate, in the interest of further developing my ideas, so I give proper
consideration to the objection that this input method is both culturally
specific and prone to misinterpretation. The first objection relates to
this gesture, which, as everyone knows, in France means "please serve me
steak tartare with a side of dauphinoise", while in Ecuador it can be
taken to mean "there is a dead coypu in my bed". I am sensitive to this
objection but believe that it can be shown to be equivalent to the
problem of getting everyone to agree on an interpretation of
Shakespeare's Macbeth, and therefore is solvable in principle. The
second objection is more serious. I have myself inadvertently invoked
undefined behaviour while scratching thus or waving like this at the
postman, so I can see that extra care is needed from the programmer.
But is that problem really anything new? Isn't it just as easy to type
"void *p = 0; int q = *p;" accidentally and never notice what we have
done? However, again this is solvable in principle by better diagnosis
of the programmer's twitches and by making allowance for the possibility
that a cat might stroll by the controller.
I am sure you will agree that this is a demonstrably better method of
writing C. If you disagree for other than argumentative reasons, I
welcome a discussion and promise only to insult people who refuse to be
persuaded that my way is best.
Tony
like to propose a new and potentially much more efficient method for C
programming. Instead of boring old keyboards and ASCII characters, what
about the use of *gestures* to enter our program code? I know that
similar (but much less adventurous) suggestions have been made recently,
and I acknowledge that I have learned from them, but everyone should
realise that this is really a dramatically new developlment in the
history of the C programming language, and the standards committee will
be forced by the sheer beauty and logical necessity of my proposal to
halt work immediately on the forthcoming standard and devote all their
attention and effort to realising my proposal immediately.
Proposal: in future, all C programs should be entered by standing in
front of the TV and making appropriate (and occasionally inappropriate)
gestures. The Kinect will then interpret these gestures and translate
the code into an intermediate form suitable for input to a C compiler.
For example, if I hold my right middle finger like this, the software
should insert a "not-equals" symbol, which may or may not be the same as
the digraph != . If I wave my left hand like this, I evidently wish to
invoke the strlen() function on the string my right hand is pointing to.
(Note: the immediacy and intuitiveness of the connection between pointers
and the act of "pointing at" should simplify discussion on comp.lang.c
enormously.) If I turn my back on the television and expose my behind,
the software should insert "void main()". If I smile at the screen, my
evident intention is "return EXIT_SUCCESS", while a frown would signify
"return EXIT_FAILURE". If I were to put my left hand... here... and my
right arm twisted like... this... (sorry, I am no spring chicken and
such contortions are not without a little discomfort) the editor would
insert a call to a Batcher's odd-even merge (recursive version) on an
array of "const investment_t*" items. At this point the efficiency
advantage of this proposed method becomes compelling.
In anticipation of some objections, I must acknowledge here that there
may be some disadvantages to the proposed method of coding. I ignore
the fact that there may be some Luddite programmers who refuse to
embrace new technology, their objections are based on prejudice and
paranoia and will not be countenanced. I also ignore all objections
from people who disagree with me - that is sufficient grounds for
considering them incompetent. However, I am keen to allow proper
debate, in the interest of further developing my ideas, so I give proper
consideration to the objection that this input method is both culturally
specific and prone to misinterpretation. The first objection relates to
this gesture, which, as everyone knows, in France means "please serve me
steak tartare with a side of dauphinoise", while in Ecuador it can be
taken to mean "there is a dead coypu in my bed". I am sensitive to this
objection but believe that it can be shown to be equivalent to the
problem of getting everyone to agree on an interpretation of
Shakespeare's Macbeth, and therefore is solvable in principle. The
second objection is more serious. I have myself inadvertently invoked
undefined behaviour while scratching thus or waving like this at the
postman, so I can see that extra care is needed from the programmer.
But is that problem really anything new? Isn't it just as easy to type
"void *p = 0; int q = *p;" accidentally and never notice what we have
done? However, again this is solvable in principle by better diagnosis
of the programmer's twitches and by making allowance for the possibility
that a cat might stroll by the controller.
I am sure you will agree that this is a demonstrably better method of
writing C. If you disagree for other than argumentative reasons, I
welcome a discussion and promise only to insult people who refuse to be
persuaded that my way is best.
Tony
