For those of you who don't follow perl5-porters, I had to share this bit of insight from Mark-Jason Dominus. You may argue it's only words, but words are all we have. -- Andy
I wish that people would stop using "intuitive" and "counterintuitive" as arguments for or against programming language features. I think it is one of those words that sounds meaningful but isn't particularly, and the judgements that come from it nearly always seem arbitrary to me. Nobody comes out of the womb with an intuition for programming languages, so in a sense all language features are counterintuitive.
Intuitions vary, and it may be possible to find someone for whom any
particular feature is "counterintuitive". When people say "X is
counterintuitive" they apear to mean that they find it
counterintuitive; this may not have any correlation with whether
anyone else feels the same way, so it's not clear that it has any
probative value at all. It would appear to be more accurate and less
prejudicial to make a personal statement, saying "I find this
counterintuitive", rather than an absolute statement "this is
counterintuitive."
I have just gathered a large collection of examples of Perl features
that have been deemed "counterintuitive" over the years, including the
=~ operator, garbage colection of anonymous subroutines, the vec()
function, the fact that defined("") and defined(0) are true, and
insertion of spaces when arrays are interpolated into strings.
People also said things like this:
Also remember, if all 256**16 addresses are possible, than any scheme which compresses some addresses to less than 16 bytes, will expand others to more than 16 bytes. This may be counter-intuitive, but is an inviolable propery of all compression schemes; it's called "The Pigeonhole Principle."and this:
Actually (though it's counter-intuitive), NFAs recognize the exact same set of languages as DFAs.and I don't find either of these things counter-intuitive. Just the opposite; I find them both manifestly obvious. That doesn't mean that they are manifestly obvious; like counterintuitivity, manifest obviousness is a property of the observer as much as it is of the thing observed.
Here's another nice example: Responding to one of the many articles in
which someone asked how to convert a string to a number, Jeff Pinyan
said:
Have you tried just USING the variable as a number?and Larry Rosler replied:
That concept is quite counter-intuitive to anyone who has used a typed language such as C and has had to face functions like atoi() and atof() all the time.
I think this sums up my point: when people use "counter-intuitive" they usually mean "unexpected", and what is unexpected to one programmer may be quite ordinary to another programmer with a different background.
I would like to see an end to the characterization of language
features as "intuitive" or "counter-intuitive", and even more, an end
to "elegant" and "inelegant". These certainly express something, but
whatever that is, it does not appear to me to be useful in discussions
about the desirability of these features.
In an interesting coincidence, I was just arguing the same thing today.
As for whether or not it's intuitive, there is very little in any language(computer or spoken) that's intuitive. I can't tell you how many times I've read in computer books something along the lines of "ternary (or unary) operators confuse some programmers so you may want to avoid them".
I do find the "inuitive/non-intuitive" arguments annoying. I used to use them until I had people start using them on me and I realized how silly they were.
There is a huge class of words I wish no one would ever again use to describe computer things. I've said before I don't care if a non-programmer can't read my code , and I still mean it.
"Consistent" and "simple" and "effective" are much better descriptions of languages and APIs and features. I prefer to talk in those terms, though they're also rather fuzzy.
Re:Exactly!
pudge on 2003-06-26T02:02:13
I would like to see an end to the characterization of language features as "intuitive" or "counter-intuitive", and even more, an end to "elegant" and "inelegant". These certainly express something, but whatever that is, it does not appear to me to be useful in discussions about the desirability of these features.
I almost entirely disagree. I agree that we should not use those words as though they are universals, absolute values shared by everyone, but I see nothing remotely damaging about saying I dislike something because it is counterintuitive to me, or that I think it would be counterinutitive to others. Much of the reason I liked perl from the beginning is because much of it was intuitive to me. I can't see how it is wrong to want to think this is a good quality in the language, and to want to preserve and enhance it.
When it all boils down to it, I could just say, "I dislike that," but saying why I dislike it -- that to me, it is counterintuitive -- seems more useful and productive. YMMV.
What I would like to see an end to is people trying to define the terms by which others should express their preferences. Again, YMMV.
