Luke:
- I started school today
- I have five hours in a row of classes
- started to learn Scala outright, as if I were to program in it
- I read three papers about its cool object model
- it seems that it has every cool idea I've already thought of
- the idea that a role isn't just an object really seems like a good idea
to me
- they're not just for composing into classes
- they're for composing into anything
- a collection of stuff
- I don't see any downsides except for the lack of strict subtyping
- but there's a neat result: parameterized classes
- they're the same as virtual classes
- just make roles purely compositional
- not subtyping things
- eventually I'll write a document on my thoughts
c:
- why do you lose subtyping?
Luke:
- if you can compose anything, you can compose class names
- you can compose those, you lose subtyping based on how that class is
used in the rest of that role
- if you use the virtual class for argument types, you have to have a
contravariant relation
Larry:
- busy with my day job
- just keeping up with the mailing lists
- I have another week or two of heavy duty work
- eventually I'll get back to the translator
- I have to prepare a talk for it for OSDC Israel
- also thinking through how to make a lower impedence mismatch connection
between my AST and some of the Pugs projects
Jesse:
- is it worth throwing random ASTs at them?
Larry:
- not yet
Jesse:
- did they convince you to visit Japan?
Larry:
- didn't take much convincing
Allison:
- haven't touched Punie in about three days
- I finished off comma lists since last week
- still haven't heard from Patrick
- have other things I need to fix too
- I might start on functions before operators
- besides control structures, that's about it
c:
- don't forget regexes
- they gave me trouble
Allison:
- I might just implement a compatibility layer
Jesse:
- how's the licensing going?
Allison:
- I'll release something within two weeks, even if I don't have comments
from everyone I asked to comment
Jesse:
- sounds good
Damian:
- working frantically to write a keynote for next week
- involved a lot of photoshopping Peter Jackson
- also negotiating with Japanese folks for the YAPC there
- preparing for the talk has thrown up a lot of questions for the mailing
list
- I'm still curious about a special way to declare class methods
- Larry seems to be keen not to tie them down
- there are a lot of people who'd like the option
Larry:
- it can always be the option
Damian:
- I'm trying to show syntax, though
- I could just have a blurry screen
- maybe I'll just show it with some fake syntax
Luke:
- I couldn't come up with a good single word for that
- Rob Kinyon kind of has a point with the polymorphism there
Larry:
- more of a cultural one, as far as I see it
Damian:
- no doubt about that
- there are plenty of people who want to leave Java without leaving the
Java mindset
Larry:
- it'd be interesting to see how much information the type inferencer
gives us for free
Luke:
- it would be wise not to rely on it
- there hasn't been much research into inferencing these type models
Damian:
- even inferrence doesn't mark them physically in the syntax
- people like that
Luke:
- it's like Haskell
- you don't have to put explicit types on your functions
- but people do -- for documentation and future-proofness
Damian:
- and checking your logic of the system
- you're showing what you believe about a system
Luke:
- Haskell will catch you eventually
- but it won't point you to the right place without prototypes
Damian:
- I almost had a bad moment and suggested "state"
Luke:
- I don't understand how partially undefinedness supports prototype-based
OO
- replying to Rob's thread might answer that
Larry:
- I'll think about that
Jesse:
- fighting with customers mostly this past week
- went to the GPL v3 thing Monday and Tuesday
- talking to potential Ponie contributors, but one fell through
- two people are interested in hacking
- they asked what there is to do
- Jerry Gay is poking at getting a Windows box for regular Parrot builds
