Stories
Slash Boxes
Comments
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.

Realistic, Down to Earth, and Swarming with Magic Robots 15 Comments More | Login | Reply /

 Full
 Abbreviated
 Hidden
More | Login | Reply
Loading... please wait.
  • I've never quite understood the RSpec crowd's obsession with renaming things, but as a library RSpec kicks the default ruby Test::Unit's arse so comprehensively otherwise.

    Nor do I understand the style that has them use do ... end style blocks to wrap code that, at it's best looks declarative. It's not impossible to write:


    describe SomeClass {
        it "should be instantiable" {
            SomeClass.new.should_not be_nil
        }
    }


    which seems a good deal cleaner. Many ruby programmers appear to be blind to ugliness of do declaration end though.

    As for _why's stuff - I think he used to be a Perl hacker and hasn't forgotten the good bits.
    • I really, really, really hate that extraneous fscking do. If only the language would let me declare things so I could write:


      describe Something
          it "should do stuff"
              do_stuff_with_it
          end
      end

      I'd be a good deal happier. Yes, I could write a source filter type thing in the runner, but I don't want the pain of having to parse the whole of the rest of bloody ruby just to get rid of that stupid do

      But, hey, ruby pays my mortgage and I'll swap the aesthetic pain
      • So that means you'll be back to Perl 5 if 5.12 has named parameters to subs and/or methods? ;)
        • Probably not. See the comment about the mortgage.

          I'm holding out for malleable syntax, real objects, proper garbage collection and decent function/method prototypes.

          And a fanatical devotion to the Pope.

          It's more likely to be things like DateTime that lure me back - Ruby's date handling is woeful.
          • Care to be more specific about Ruby's date handling issues?
            • Go and read the documentation for the DateTime suite of modules. Now point me to anything as comprehensive and well thought out on rubyforge. At least Ruby on Rails is starting to address some of the problems in its earlier implementations of things like 1.month.from_now by inventing/rediscovering the Duration class.

              http://www.perl.com/pub/a/2003/03/13/datetime.html [perl.com] is a good read too, in which Dave Rolsky explains the issues with the ad hoc nature of most Perl calendrical modules before the DateTime better

      • Have you seen Devel::Declare? It’ll provide half a macro system – for Perl 5.

    • Many ruby programmers appear to be blind to ugliness of do declaration end though.

      Goes with the territory. Lisp programmers don’t even see the parentheses any more, nor do XML people consciously notice all the angle brackets. I used to be blind to the iffiness of Perl’s deref syntax.

    • I quite agree about Ruby's Test::Unit being barely functional. I very nearly ported Test::Builder and Test::Class to Ruby a year and a half ago, when I was doing more Ruby hacking. Then I realized that instead of spending the next several months reinventing all of the tools I like and use, I could spend the next several months building useful things that don't already exist, and stopped writing Ruby code.

    • I've never quite understood the RSpec crowd's obsession with renaming things...

      This is what grates me. It's clear that RSpec is a tremendous boon to Ruby testing. I'm all for that! I just wish that this goodness weren't couched in such tiresome wankery over decades-old concepts repackaged bizarrely.

      There's no shame in borrowing good ideas! Just please stop renaming them, explaining them badly, slapping bright yellow racing fins on them, and pretending that they're completely new and different. Th

      • This is where we differ, I think. I don't really give a bugger if people want to give things fancy new names, so long as the thing itself is useful. The RSpec crew have taken the view that the reason many people don't do Test Driven Design is because it seems arse about face to them. Testing is what you do after you've done the programming after all. (I remember once having a rather voluble argument with someone about the benefits of running your test before implementing anything to make sure that it failed

        • I don't really give a bugger if people want to give things fancy new names, so long as the thing itself is useful.

          I care, if it confuses the innocent. Count the number of people who say "We use BDD and TDD" or "I'm learning BDD and TDD". The number is non-trivial.

          Also, ISTR we all called the practice "Test First Development" until around the time Beck published "Test Driven Development", which is definitely a way better name for the process.

          Agreed. Also, I prefer the term Test-Driven Design, but

    • Many ruby programmers appear to be blind to ugliness of do declaration end though.
      That's why I use brackets instead. ;)
Stories, comments, journals, and other submissions on use Perl; are Copyright 1998-2006, their respective owners.