When is a function not a function?

[David Mark]

I've got a collection of functions that accept a function or object
(paired with a method name) as a callback.

For the longest time I have relied on this test.

(typeof cb == 'function')

This should work as well, but I am not sure how well it degrades in
older browsers. I think there are issues with functions created in
another context (eg frame) as well.

(cb instanceof Function)

Then I came across this.

(!!fn && typeof fn != "string" && !fn.nodeName && fn.constructor !=
Array && /function/i.test( fn + "" ))

I'm sure it is wrong as it is from jQuery, but it makes me wonder if
there are bugs in older browsers that will mislead my test.

As an aside, I was looking at the jQuery source as some browser
sniffer I was arguing with on a blog cited it as part of an "everybody
is doing it" argument. I had glanced at it previously and dismissed
it based on its resemblance to Prototype. Upon closer inspection, it
makes Prototype look inspired. Suffice to say that anybody who
recommends it hasn't read the code. It's a horror show.

[Peter Michaux]

On Sep 25, 7:09 pm, David Mark <dmark.cins...@ gmail.comwrote:

I've got a collection of functions that accept a function or object
(paired with a method name) as a callback.

For the longest time I have relied on this test.

(typeof cb == 'function')

This should work as well, but I am not sure how well it degrades in
older browsers. I think there are issues with functions created in
another context (eg frame) as well.

(cb instanceof Function)

Then I came across this.

(!!fn && typeof fn != "string" && !fn.nodeName && fn.constructor !=
Array && /function/i.test( fn + "" ))

I'm sure it is wrong as it is from jQuery, but it makes me wonder if
there are bugs in older browsers that will mislead my test.

I don't like to assume incorrect because it is in jQuery. I am
surprised if they are testing in older browsers. The hip libraries
don't seem to do that.

As an aside, I was looking at the jQuery source as some browser
sniffer I was arguing with on a blog cited it as part of an "everybody
is doing it" argument. I had glanced at it previously and dismissed
it based on its resemblance to Prototype. Upon closer inspection, it
makes Prototype look inspired. Suffice to say that anybody who
recommends it hasn't read the code. It's a horror show.

In a talk at Google, John Resig, jQuery founder, said something like
"after fighting with browser bugs long enough an experienced
JavaScript program just uses navigator.userA gent." At least that is
how I understood his talk. He does say many really good things. You
can watch the video.

http://ejohn.org/blog/building-a-javascript-library/

I left a comment on that blog page about sniffing and he never did
really respond. Perhaps that future blog post to which he refers later
in the comments is still in the works. I'm interested to read it.

Tricky one: How do you feature test opacity support?

Peter

[Peter Michaux]

On Sep 26, 1:02 pm, David Mark <dmark.cins...@ gmail.comwrote:

jQuery uses dollar sign variables, which I don't think are allowed
under the specification.

Dollar signs are allowed. See sec 7.6. The spec then says "The dollar
sign is intended for use only in mechanically generated code."

After reading all the uproar about dollar signs in variables or even
one character dollar sign variable, I think it would have been better
if they had left this statement of intention out of the spec. A dollar
sign can or cannot be in an identifier. If it can then it is up to the
programmer to decide how to use it. If I was writing JavaScript to
last 100 years and knew JavaScript would be backwards compatible with
the exception of new keywords, I would probably start all my
identifiers with a dollar sign (eg. $foo, $bar) to avoid clashes with
new keywords.

Peter

[David Mark]

On Sep 26, 5:22 pm, Peter Michaux <petermich...@g mail.comwrote:

On Sep 26, 1:02 pm, David Mark <dmark.cins...@ gmail.comwrote:

jQuery uses dollar sign variables, which I don't think are allowed
under the specification.

Dollar signs are allowed. See sec 7.6. The spec then says "The dollar
sign is intended for use only in mechanically generated code."

I understand that, but I don't think the authors of these libraries
are robots.

>
After reading all the uproar about dollar signs in variables or even
one character dollar sign variable, I think it would have been better
if they had left this statement of intention out of the spec. A dollar
sign can or cannot be in an identifier. If it can then it is up to the
programmer to decide how to use it. If I was writing JavaScript to
last 100 years and knew JavaScript would be backwards compatible with
the exception of new keywords, I would probably start all my
identifiers with a dollar sign (eg. $foo, $bar) to avoid clashes with
new keywords.

The biggest problem I have is naming functions "$" or "$$" or anything
like that. If you minify your production code then the only argument
for these is to reduce the total number of keystrokes required to
write a script. Apparently these people have never heard of macros.

[Thomas 'PointedEars' Lahn]

Peter Michaux wrote:

On Sep 26, 1:02 pm, David Mark <dmark.cins...@ gmail.comwrote:

>jQuery uses dollar sign variables, which I don't think are allowed
under the specification.

Dollar signs are allowed. See sec 7.6. The spec then says "The dollar
sign is intended for use only in mechanically generated code."

Which is exactly why jQuery should not.

After reading all the uproar about dollar signs in variables or even
one character dollar sign variable, I think it would have been better
if they had left this statement of intention out of the spec. A dollar
sign can or cannot be in an identifier.

This subsection addresses only identifiers. You missed the point completely.
PointedEars
--
realism: HTML 4.01 Strict
evangelism: XHTML 1.0 Strict
madness: XHTML 1.1 as application/xhtml+xml
-- Bjoern Hoehrmann

[Thomas 'PointedEars' Lahn]

Peter Michaux wrote:

[...] "Richard Cornford" [...] wrote:

>Peter Michaux wrote:

>>If I was writing JavaScript to last 100 years and knew
JavaScript would be backwards compatible with the exception
of new keywords, I would probably start all my identifiers
with a dollar sign (eg. $foo, $bar) to avoid clashes with
new keywords.

Doesn't that imply that given a sufficiently long period of time a
programming language will grow an infinite number of keywords?

I don't think I implied that. By using $ at the start of every one of
my identifiers, I'm protected from that one new keyword that clashes.

Generally, you are not. There are built-in properties that begin with `$'.
Hence "machine-generated".
PointedEars
--
realism: HTML 4.01 Strict
evangelism: XHTML 1.0 Strict
madness: XHTML 1.1 as application/xhtml+xml
-- Bjoern Hoehrmann

[Randy Webb]

Thomas 'PointedEars' Lahn said the following on 9/26/2007 5:40 PM:

Peter Michaux wrote:

>On Sep 26, 1:02 pm, David Mark <dmark.cins...@ gmail.comwrote:

>>jQuery uses dollar sign variables, which I don't think are allowed
under the specification.

Dollar signs are allowed. See sec 7.6. The spec then says "The dollar
sign is intended for use only in mechanically generated code."

Which is exactly why jQuery should not.

But there is no reason why it *can not* other than people's pedantic
desires to whine about it. Had it been S that was used instead, people
would shut up about it and move on. Except that if it had been S instead
of $ then people would be whining about how the name wasn't an
indication of what it did.

--
Randy
Chance Favors The Prepared Mind
comp.lang.javas cript FAQ - http://jibbering.com/faq/index.html
Javascript Best Practices - http://www.JavascriptToolbox.com/bestpractices/

[Peter Michaux]

On Sep 26, 5:39 pm, "Richard Cornford" <Rich...@litote s.demon.co.uk>
wrote:

Generally, the place for one letter Identifiers is as local variable
names in functions that are sufficiently small that their declarations
and all of their uses can be simultaneously surveyed. And maybe not even
then.

I find it's odd that the Prototype folks like very short API
identifiers but use very long local variable names inside Prototype.

Peter

[Peter Michaux]

On Sep 26, 4:38 pm, "Richard Cornford" <Rich...@litote s.demon.co.uk>
wrote:

Peter Michaux wrote:

On Sep 26, 1:02 pm, David Mark wrote:

jQuery uses dollar sign variables, which I don't think are
allowed under the specification.

Dollar signs are allowed. See sec 7.6. The spec then says
"The dollar sign is intended for use only in mechanically
generated code."

The same is true for Java (indeed it appears that the idea of making
provision for a conventional form to exist for machine generated code
was inspired by the similar convention in Java). You never see Java
programmers using $ symbols in their code, even though not doing so is
no more than a convention. A Java programmer who disregarded that
convention would be derided, but unfortunately a sufficiently large
proportion of javascript 'programmers' are unprofessional (in every
sense of the world) that these faults can seem tolerable.

After reading all the uproar about dollar signs in variables
or even one character dollar sign variable, I think it would
have been better if they had left this statement of intention
out of the spec.

In that case it would have been better to leave the $ symbol out of the
set of acceptable characters for an Identifier.

A dollar sign can or cannot be in an identifier.

Just as it can in Java.

If it can then it is up to the programmer to decide how
to use it.

Just as it is for Java programmers, but they have more understanding of
their specification, and more appreciation for the value of conventions
in programming.

So Java programmers are all instances of the same class?

Peter

[John G Harris]

On Wed, 26 Sep 2007 at 20:21:44, in comp.lang.javas cript, Randy Webb
wrote:

>Thomas 'PointedEars' Lahn said the following on 9/26/2007 5:40 PM:

>Peter Michaux wrote:

>>On Sep 26, 1:02 pm, David Mark <dmark.cins...@ gmail.comwrote:
jQuery uses dollar sign variables, which I don't think are allowed
under the specification.
Dollar signs are allowed. See sec 7.6. The spec then says "The dollar
sign is intended for use only in mechanically generated code."

Which is exactly why jQuery should not.

But there is no reason why it *can not* other than people's pedantic
desires to whine about it. Had it been S that was used instead, people
would shut up about it and move on. Except that if it had been S
instead of $ then people would be whining about how the name wasn't an
indication of what it did.

But it's already happened :

Thou shalt not start a global name with "MM_".

Because otherwise there will come a day when you get an obscure bug that
will drive you crazy.

There's a similar rule in C++ about names starting with leading
underlines. The difference is that the rule has been biting people for
years, so competent programmers don't dismiss it as pedantic standards
mania.

John
--
John Harris