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

use Perl Log In

Log In

[ Create a new account ]

chromatic (983)

chromatic
  (email not shown publicly)
http://wgz.org/chromatic/

Blog Information [technorati.com] Profile for chr0matic [technorati.com]

Journal of chromatic (983)

Friday February 02, 2007
04:33 PM

Perl 6 Design Team Minutes for 31 January 2007

[ #32310 ]

The Perl 6 Design team met via phone on 31 January 2007. Larry, Patrick, and chromatic attended. These are the notes.

Patrick:

  • have reference working
  • assignment seems to be working
  • next step is for loops and END blocks

c:

  • that gives you sanity?

Patrick:

  • yes
  • have use statements mostly working
  • need a free full day to work on that
  • reminds me of a question from my talk last week
  • how difficult is it to use NCI to link to C++ libraries, not just C libraries?

c:

  • two answers
  • "I don't know" is the best answer
  • if they use "extern C" as most do, it's not too difficult
  • depends on the quality of the bindings
  • exceptions are difficult
  • C++ object support may be tricky

Larry:

  • it's ironic that the motivation for Perl 5 was to link to C++
  • I'm fleshing out the declarative versus procedural semantics
  • nothing you haven't seen before
  • started constructing the official grammar
  • trying to allow processing into whatever grammar people want to use
  • based in on your current grammar
  • trying to write it without the actions
  • it has placeholders with the actions
  • identifiers make it easy to post-process that into the grammar with actions

Patrick:

  • it doesn't have closures?

Larry:

  • they have nothing in them but star
  • you can use it bare
  • the closures will terminate longest-token matching without any other effect
  • basically no-ops
  • easy to identify visually

Patrick:

  • is it somewhere I can look at it?

Larry:

  • did a paste last night of the top end of the grammar
  • haven't checked it in anywhere yet
  • currently using hash style
  • may or may not be the right thing in the long run
  • I'll send it to you

c:

  • I've heard a few people ask for a tutorial on writing a language on Parrot
  • I might whack at that, unless it steps on anyone's toes

Patrick:

  • the more the merrier!
  • languages/abc is a tutorial-type language

c:

  • great, I'll take a look

Larry:

  • doing some documentation cleanup
  • filling in semantics to the operators
  • did the great S03 reorg, as I said I would last week
  • basically trying to fill in
  • instead of assuming "this'll be the same as Perl 5 mostly", I'm trying to be more explicit
  • also changed negative subscripts
  • they no longer DWIM for shaped arrays
  • only for unshaped arrays
  • we're negotiating the syntax
  • probably involving stars
  • seems to work out nicely
  • did a lot of cleanup on the syntax of context coercions, particularly the sigiled ones
  • unified precedence
  • involved making prefix list operators looser than list infix operators
  • separated comma operator into its own categories
  • have 21 categories now, which is still two less than Perl 5
  • didn't like that the hand-waving sophistry over comma being a list operator
  • because we have consistent contextualizer syntax with parens or colons, we can use the null one as the target of a feed operator
  • you can put @() in there
  • it'll know that that's where it installs a list
  • can use the other contextualizers to coerce this
  • it's all much more orthogonal
  • no longer a *** token
  • did a lot of gather/take semantics clarification
  • with consistent contextualizers, we can unify separate functions that were producing different kinds of lists
  • zip and each are now zip, depending on whether you want it to return subarrays
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More | Login | Reply
Loading... please wait.
  • Thanks for posting these. I don't read them, but someday I might want to, and I think that's the point!
    • I read every one. Mostly because it is interesting to see how "Perl research" is coming along or however you want to classify it. : )