The Perl 6 design team met by phone on 07 March 2007. Larry, Damian, Allison, Nicholas, Jesse, and chromatic attended. These are the minutes.
Damian:
- feeling under the weather and unproductive
- tried coding the other day, but it was like those pictures of spiders on acid
- I backed that out
- decided to read through the Synopses instead
- sending Larry feedback
- very minor feedback
- just about finished Synopses 4
- Synopsis 3 is surprisingly long
- should finish within the week
c:
- finished updating the Perl 6 design docs published on Perl.com
- every page now links to the relevant Synopsis on dev.perl.org
Allison:
- had a great Hackathon in the Netherlands this past week
- I hacked with Jonathan and Liz on the Parrot object metamodel
- pushed tewk's experiments in
- not sure Jonathan has checked in everything
- very concrete, much less theoretical
Nicholas:
- Rick Delaney asked p5p what the MAD stuff is
- I answered what I could
Larry:
Nicholas:
- he said that he might start doing things, if it's interesting
- he wants to manipulate the optree
- so he probably will use the
B:: modules
Larry:
- fiddled around with the semantics of ranges
- their use as intervals does not conflict with their use as beginning and ending points
- settling in on the syntax for assigning rules for symbolic operators
- generalizing that identifiers can be qualified with adverbs in various situations
- not just like infix operators
- specifying long names where you want to have control over name mangling
- turning into strings internally
- the user gets explicit delimiters and canonicalization
- not only for the names of rules
- going back to the names of modules
- there's a three-part hyphenated special case
- the real name of the module is probably an identifer with various adverbs
- trying to see where that sort of thing comes through
- fielding Damian's Synopses feedback
- lots of Pugs activity going on
- converting internal data types to the new meta-object model
- many of the regression tests are red at the moment
- improving from here on in
- continuing to look for various simplifications
- did delete the short form of
.assuming
- that went unmourned
Jesse:
- getting close to being able to do Perl 6 microgrants
- that's most of it for me
Damian:
- I'm watching a p6l thread with horror
- trying to nail down the star as an array index
- I want to straighten this out within my own head
- we want a syntax that allows us to use ordinals, instead of indices
- in case an array doesn't start from zero
- if you have an offset from whatever, that might be useful
- the current mutation seems very different
- was that the goal?
- should there be an ordinal way to specify array elements, not just a cardinal way?
c:
- I could have used the ordinal syntax yesterday, actually
- we had that in Perl 5, for negative indices
Damian:
- only from the wrong end
- it would be nice to have it for both ends
Larry:
- I do like the notion that square brackets for your subscript always assumes that your indices always start from zero
- great benefit to clarity
- I do like the idea that the curlies indicate something fancier
- we can certainly figure out where the ordinal syntax is going
- what exactly does the whatever-star mean?
- does it have a consistent gloss that the user can read?
- if you see a star in a subscripty position, can it always mean "the keys in this spot"
Damian:
- I like that
- good idea
- makes it easier to teach
Larry:
- the basic idea is sound here
Damian:
- I'm nervous about using curly braces on arrays
- eliminates almost the last syntactic hint differentiating them from hashes
- maybe the fixed sigils always helps there
- I'm worried about accesses through references though
Larry:
- it sort of is an ordered hash
- you can view it that way
Damian:
- maybe the
*-1 and *+0 needs to disappear
- I'll look more into it
Larry:
- regarding string auto-increment
- you gave a more generalized regex
- your example though, would end up incrementing the
jpg not the filename
Damian:
- I wanted to achieve a more sophisticated regex
- not sure I did that
- I can almost see the regex that does it
- want me to have another run at it?
- that particular case has bitten me a few times
- if we have a chance to fix it, maybe we should
- another question from S04
- we returned to a named iterator in a
for loop aliasing $_ at some point
- that makes it impossible to put a
for loop inside a given
Larry:
- we'll have to weigh that up again
