Hacker News Reader

Comment by jhgb | Hacker News Reader

> 1. Its variables start with $ and that is ugly:

At the very least it seems nonsensical and cargo-cultish: As a non PHP-er, what is the actual purpose of $? In Perl, it indicates variable context (for better or worse, there's more than one, so it has to be indicated somehow). In shell, it indicates the substitution of variable name by its value. In PHP...I draw a blank. It really seems like an "I wanted my language to look like some other language but I had no idea what it should be doing" kind of thing.

View on HN

$ was taken from Perl

https://stackoverflow.com/a/3073818

but because early PHP was more simplistic than Perl it only has $.

Powershell also uses $ for variables.

Regardless of etymology of the $ sign and the usefulness of it, personally I like it because it makes it easier for me when I'm reading code, especially if you are scanning fast, to differentiate variables from symbols. Yes, you can get that with your editor too, but for me it is easier to associate the $ sign with a variable rather than a specific color, sometimes you don't have coloring available like when in command line going thru diffs, cat, nano etc.

It seemed kind of obvious to me even at that time that it was taken from Perl. I just didn't get why. PHP could have lifted IMPLEMENTATION DIVISION from COBOL, but likewise, there would have been no point either (fortunately it didn't do that!).

Early web was written in Perl, I guess it was the classic approach of lets borrow what is already successful and in end PHP won over Perl so it worked too. Bill Gates would have been proud.

It's actually slightly wrong to just say "variables start with $". It's more like "$ dereferences a string":

  $foo = "I am foo\n";
  $bar = "foo";
  echo $$bar;

This is in the documentation under "variable variables" [0]. This also lets you create dynamic function names:

  function run_foo() {
      echo "I'm a foo\n";
  }

  $method = 'foo';  
  $picked = "run_$method";
  $picked();

[0] https://www.php.net/manual/en/language.variables.variable.ph...

It differentiates variables from global constant things (constants, class names, function names, ...) for which the scoping is different. So you can have $strlen = strlen(...);

Off the top of my head, I struggle to come up with situations where you'd need an $ to distinguish class names or function names syntactically. Hell, in Lisp I can do (let ((length (length seq-1))) ... (length seq-2) ... ) just fine and it still works as expected, so no special characters necessary either.

Plus, to have different scoping rules was considered a good idea? Weird scoping issues were why I gave up on Ruby fifteen years ago; perhaps it was a good idea that I got never that far with PHP. But I'm intrigued now; what exactly are the scoping differences in question?

TBH, I don't know lisp and I have no clue what you just wrote. Someone new to PHP can just be told $something is how variables look like. And they will recognize variables forever from then on. What's a variable in that lisp code, or in any general lisp code? I have no clue at first glance.

I didn't say it's necessary for the parser, but it's useful for the learner/user.

You can do:

$strlen = 'strlen';

echo $strlen('test');

I’ve also seen things like this:

$return = ...

You can’t use a keyword as a variable in most other languages.

It means the start of a variable name, as simple as that, it isn't anything significant. Maybe it was made to make parsing the language easier at the time, who cares?

You can get used to that in 10 min and move on with your life, it is not a reason to not pick up the language, it's bike shedding.

There was a number of other reasons why I didn't end up using PHP, so it was a bit like death by a thousand cuts. It was not the reason for me, just a small showcase of the weirdness for a newcomer.

Reply on news.ycombinator.com