I've taught beginners before. I know that beginners often get confused by the use of $ to denote single array elements (e.g. $array[0]). I know why they get confused but I think I'm pretty good at getting them to understand it - this is, at least, partly because I think that the rule makes sense.
Yesterday I gave my first beginners course for a while. More importantly it was the the first beginners course I'd given since we've known a lot about the shape of Perl 6.
As always, there was someone who didn't like array (and hash) elements using $. "Oh, I understand exactly why it's there", he said, "but I just don't think it's the most logical way to do it." Of course I disagreed with him and tried to bring him round to my way of thinking. I found this was harder than usual because I now know that in Perl 6 array elements will be accessed using @. It's a lot harder trying to explain away something as logical when you know that it's going to change in the near future.
Have any other trainers come across this problem?
I've found it to be rather common to need to apologize for the state of the technology being presented. This could be something as simple as using $array[0] vs. @array[0] in Perl, or much more unfortunate like the state of XML. Most recently, I had to teach DTD syntax and usage to a bunch of programmers who were going to be using XML Schema, and had to defend DTDs, only to illustrate the failings of DTDs now that XML Schema exists -- and then to backpedal on XML Schema because it's not widely liked or implemented out in the field yet (and the syntax is kinda backwards as well).
I've heard Python trainers talk about similar issues, like meaningful whitespace. XP presenters also plea to suspend disbelief about pair programming, or spend a lot of time defending pair programming (or test-first programming, or other facets of XP). Thinking back to when I learned C <mumble> years ago, I remember the professor took a stand on learning how CPUs work, as if to apologize for the incredibly low-level nature of the C language.
Re:Technology apologists
Matts on 2001-11-15T09:49:40
Solution: Teach schematron for validating XML instead. Actually, I'd kill for a decent RELAX implementation any day over a Schemas implementation!
Re:You make no apologies and just deal
ziggy on 2001-11-15T14:29:54
You can make that statement today with a straight face. How about 2 years ago?
Jason
Re:Why is $array[0] such a bad thing?
Matts on 2001-11-15T09:47:51
Unfortunately it's worse than you think.
@$array[0] binds the @ more tightly than the [0], so it expects $array to be an array ref, which you're de-referencing, rather than $array[0] being the array ref. Except it doesn't work, because as you know, you don't say @x[0], you say $x[0], so when you say @$array[0], it thinks you're dereferencing $array, not @array, and $array doesn't exist (well it might, but it's not what you meant).
Now @{$x[0]} will give you the de-referenced array in the first entry in the @x array. Because here we're being explicit about the precedence using curlies to bind it. But it can get a bit more confusing, because say you wanted the second entry in the first array? i.e. given:
my @x = ([1,2], [3,4]);
So you want the "2" above. Now that's:
my $two = @{$x[0]}[1];
But we got a scalar! Now what happened to that "symbol means context" thing here? Screwey, huh?
Fixing this is a Good Thing (tm).
