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)

Thursday March 29, 2007
03:43 PM

Perl 6 Design Notes For 28 March 2007

[ #32837 ]

The Perl 6 design team met by phone on 28 March 2007. Larry, Allison, Patrick, Nicholas, and chromatic attended. These are the minutes.

Larry:

  • madly simplifying things
  • made the big shift over to consider all punctuation in regexes as metasyntax
  • modifying the test suites and standard grammar to look for that
  • figuring out how to do what Damian asked for
  • simplifying the binding of topicalizers
  • there's no longer any implicit binding of the first arrow of a pointy
  • you must bind $_ either explicitly or implicitly, in the case of no arguments at all
  • thought I might not be able to exit the current topicalizer
  • those semantics are unclear
  • I have a checkin with a new tweak
  • it exits the innermost block which uses $_ as one of its formal parameters
  • that captures the notion of what a user expects to be treated as the topicalizer
  • ignores the fact that you can change the value of $_ on the fly
  • no one will expect a when statement to exit from an assignment statement
  • Damian should be happy
  • finally decided I liked Juerd's double-pointy block
  • it makes it easy to write a for loop with a bunch of variables and have them all default to read-write
  • treats the default like a my declaration
  • flips the default back over to the my view of reality
  • it's visually mnemonic
  • Damian's changes to simplify the semantics of arrow also helped make that reasonable
  • part of my test changes meant that I changed the regex tester to use a platform-neutral mechanism of specifying TODOs
  • just put a comment before the line in the data file there
  • adverbs let you specify what the status on each platform is
  • keeps them more-or-less inline and still easy to ignore
  • we'll have to line up the tests from the Pugs directory to get it and Parrot in line

Patrick:

  • my immediate plan is to implement the regex changes and migrate things over
  • I really want to get the sanity test done
  • that appears to be this weekend's test

c:

  • fixing minor memory leaks and frees at Parrot interpreter end
  • this makes Valgrind reports much more useful for tracing actual leaks
  • willing to work on just about anything else though

Patrick:

  • how about the SDL bindings?
  • could use those within Perl 6 soon

c:

  • happy to work on that

Allison:

  • trying to finish the Objects PDD this week
  • Jonathan is implementing it this week

Nicholas:

  • the regex expression talks about metacharacters in terms of graphemes
  • does a grapheme contain combining characters?

Larry:

  • yes
  • it's alphanumeric if its base character is alphanumeric
  • an accented a is alphanumeric because a is alphanumeric

Nicholas:

  • the accent isn't though
  • a backslashed, accented a is a metacharacter, right?

Larry:

  • yes
  • a grapheme is a continuous unit, and we treat them that way
  • at some point, we have to say "don't be stupid"

Nicholas:

  • when the GC collects, is there an option to have it scribble over memory?
  • 0xFEFEFEFE?
  • could be a nice way to make it segfault much easier
  • Perl 5 uses something called PERL_POISON when you call free() or realloc()
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.