Perl 6 hackathon targets: Revision 19
These are topics and/or tasks that are likely candidates to be addressed or worked on at the [Oslo Perl 6 hackaton 2009]. They're here as a list to remind us of (1) important issues we may want to address when we have a gathering of the principals, and (2) ideas for things that we can recruit others to work on.
^^ Rakudo-specific targets
* Update the development roadmap
* Review RT tickets, verify that open bugs/tickets have corresponding spectests
* Ideas/planning for initial module implementations, where hosted, if/how distributed with implementations
* Update documentation on rakudo.org
* Create/update FAQ on rakudo.org
* Update "committer's guide", "release manager's guide", etc. documents
* Plan for handling .leave
* Spectest coverage review
* Fix 'make install' target
* Planning for laziness and lazy lists
* Identify critical stumbling blocks for newcomers to Rakudo / Perl 6
* Clean up handling of True, False, .true, .not, etc.
* Clean up handling of .Str for some builtin types (Int, Num)
* qx() workarounds?
* RT #64810?
* Fix up capture handling
* Container autovivification
^^ Perl 6 language targets
(Language design issues that may involve discussions with Larry or proposals to p6l)
* Handling of take() values, especially in loops
* Details of infix:<cmp>
* Do we want want() ?
* See also [perl6 tasks]
* String Range semantics
* Types nested inside roles
* access private members of sibling instances
^^ Rakudo things accomplished / decided
* Fixed some bugs with .hash and .list on match objects (pmichaud)
* Added IO::Socket (masak, mberends)
* Lots more parameterized type thingies (jonathan++)
* Syntax highlighting in Padre
^^ Perl 6 things accomplished / decided
* is augment lexically scoped? (answer: no, it's a form of monkey patching)
* review EXPORTALL semantics
* Clarify enum/boolean stringification semantics
* Type constraints in arrays [3, 4, 5] <-> Positional[Int] ?
** (answer -- generally an anonymous array/list doesn't promote to a given constraint)
** however, there is a "narrowness criteria" for parameterized roles, which jonathan++ has specced (S14) and implemented (Rakudo)
* Eliminated prefix:<=>, now use .get, .lines.
* Also eliminated fish term ( =<> ), it's now lines() .
* Binding doesn't have any "action at a distance" semantics -- tends to work locally
* "augment" should probably affect existing instances of a type