define side effects

[Niklaus]

This is one of the posts that i got.
------------------------------
A "side effect" of an operation is something that
*happens*, not something that *is produced*. Examples:
In the expression 2+2, the value 4 *is produced*. Nothing
*happens*.
Thus, 4 is the value of the expression, and it has no side effects.
In the expression g=2.0, the value 2.0 is produced. What *happens*
is that 2.0 is assigned to g. Thus, 2.0 is the value of the
expression,
and its side effect is to assign 2.0 to g.
In the expression (int)g, the value 2 is produced. Nothing happens.
Thus, 2 is the value of the expression (int)g, and it has no side
effects.
In the expression (a=1,++a), the value 2 is produced. What happens
is that first 1 is assigned to a, and then a is incremented; those are
the side effects of the expression.
-----------------------------

This seems to be a good definition in most of the cases.

I argue that in k=(int)2.0 , 2 is produced and the truncation of 2.0
-> 2 is that happens. I am really confused. Can some one give me a
more clear definition of side effects ?

If i apply the same definition to i=j++ where j=3 then i=3 and j=4
are the side effects and value 3,4 are produced. Am i correct ? Can
someone throw more light on how to define side effects without
ambiguity ?

-
Nik

 

[Richard Bos]

This is one of the posts that i got.
------------------------------
A "side effect" of an operation is something that
*happens*, not something that *is produced*. Examples:
In the expression 2+2, the value 4 *is produced*. Nothing
*happens*.
Thus, 4 is the value of the expression, and it has no side effects.
In the expression g=2.0, the value 2.0 is produced. What *happens*
is that 2.0 is assigned to g. Thus, 2.0 is the value of the
expression,
and its side effect is to assign 2.0 to g.
In the expression (int)g, the value 2 is produced. Nothing happens.
Thus, 2 is the value of the expression (int)g, and it has no side
effects.
In the expression (a=1,++a), the value 2 is produced. What happens
is that first 1 is assigned to a, and then a is incremented; those are
the side effects of the expression.
-----------------------------

This seems to be a good definition in most of the cases.

I argue that in k=(int)2.0 , 2 is produced and the truncation of 2.0
-> 2 is that happens.

No; the 2.0 is not changed. Nothing happens "behind the scenes"; the
value 2.0 does get truncated, but only for the direct reason of
calculating the value of the expression. Assigning this truncated value
to k _is_ a side effect.

I am really confused. Can some one give me a more clear definition
of side effects ?

Well, according to the Standard,

# [#2] Accessing a volatile object, modifying an object,
# modifying a file, or calling a function that does any of
# those operations are all side effects, which are changes
# in the state of the execution environment.

Whether that is more clear, well... it's unambiguous, anyway.

If i apply the same definition to i=j++ where j=3 then i=3 and j=4
are the side effects and value 3,4 are produced. Am i correct ?

No; 4 is never produced. j is increased to 4, but that value is never
passed on to any other sub-expression; its previous value, 3, is.

Richard

 

[Gautam]

This is one of the posts that i got.
------------------------------
A "side effect" of an operation is something that
*happens*, not something that *is produced*. Examples:
In the expression 2+2, the value 4 *is produced*. Nothing
*happens*.
Thus, 4 is the value of the expression, and it has no side effects.
In the expression g=2.0, the value 2.0 is produced. What *happens*
is that 2.0 is assigned to g. Thus, 2.0 is the value of the
expression,
and its side effect is to assign 2.0 to g.
In the expression (int)g, the value 2 is produced. Nothing happens.
Thus, 2 is the value of the expression (int)g, and it has no side
effects.
In the expression (a=1,++a), the value 2 is produced. What happens
is that first 1 is assigned to a, and then a is incremented; those are
the side effects of the expression.
-----------------------------

This seems to be a good definition in most of the cases.

I argue that in k=(int)2.0 , 2 is produced and the truncation of 2.0
-> 2 is that happens.

No; the 2.0 is not changed. Nothing happens "behind the scenes"; the
value 2.0 does get truncated, but only for the direct reason of
calculating the value of the expression. Assigning this truncated value
to k _is_ a side effect.

I am really confused. Can some one give me a more clear definition
of side effects ?

Well, according to the Standard,

# [#2] Accessing a volatile object, modifying an object,
# modifying a file, or calling a function that does any of
# those operations are all side effects, which are changes
# in the state of the execution environment.

Whether that is more clear, well... it's unambiguous, anyway.

If i apply the same definition to i=j++ where j=3 then i=3 and j=4
are the side effects and value 3,4 are produced. Am i correct ?

No; 4 is never produced. j is increased to 4, but that value is never
passed on to any other sub-expression; its previous value, 3, is.

Richard

regarding side - effects
what exactly does this sequence 'point mean'

 

[Joona I Palaste]

This is one of the posts that i got.
------------------------------
A "side effect" of an operation is something that
*happens*, not something that *is produced*. Examples:
In the expression 2+2, the value 4 *is produced*. Nothing
*happens*.
Thus, 4 is the value of the expression, and it has no side effects.
In the expression g=2.0, the value 2.0 is produced. What *happens*
is that 2.0 is assigned to g. Thus, 2.0 is the value of the
expression,
and its side effect is to assign 2.0 to g.
In the expression (int)g, the value 2 is produced. Nothing happens.
Thus, 2 is the value of the expression (int)g, and it has no side
effects.
In the expression (a=1,++a), the value 2 is produced. What happens
is that first 1 is assigned to a, and then a is incremented; those are
the side effects of the expression.
-----------------------------

This seems to be a good definition in most of the cases.

I argue that in k=(int)2.0 , 2 is produced and the truncation of 2.0
-> 2 is that happens.

No; the 2.0 is not changed. Nothing happens "behind the scenes"; the
value 2.0 does get truncated, but only for the direct reason of
calculating the value of the expression. Assigning this truncated value
to k _is_ a side effect.

I am really confused. Can some one give me a more clear definition
of side effects ?

Well, according to the Standard,

# [#2] Accessing a volatile object, modifying an object,
# modifying a file, or calling a function that does any of
# those operations are all side effects, which are changes
# in the state of the execution environment.

Whether that is more clear, well... it's unambiguous, anyway.

If i apply the same definition to i=j++ where j=3 then i=3 and j=4
are the side effects and value 3,4 are produced. Am i correct ?

No; 4 is never produced. j is increased to 4, but that value is never
passed on to any other sub-expression; its previous value, 3, is.

regarding side - effects
what exactly does this sequence 'point mean'

It's a point during the evaluation of an expression, when all side
effects are guaranteed to have taken place. Sequence points include:
- The terminating ; in a statement
- The && and || operators
- The ?: operator
- The , operator
Also, AFAIK when a function is called, its entry point forms a
sequence point for the expressions in its arguments.

 

[Neil Kurzman]

Simply

The same as in medicine. A drug cures one problem, but causes another.

In code you fix one bug, but cause another.

 

[Arthur J. O'Dwyer]

Simply
The same as in medicine. A drug cures one problem, but causes another.
In code you fix one bug, but cause another.

An interesting quotation, but why do you say so? If this is
supposed to be a new topic of discussion (incidentally, one more
suited to comp.programming than comp.lang.c, which is dedicated
specifically to fixing bugs in *C* code then you might have
considered starting a new thread rather than piggybacking on an
existing one.
We've seen people here before who apparently thought they should
"conserve threads" by posting irrelevant replies to old threads;
that's really not necessary.
And if you *didn't* mean your reply to be irrelevant to the old
thread, then you ought to have quoted some context so people could
tell to what you were responding. Google "usenet faq" for more
information.

-Arthur