Log of the #nice channel on irc.freenode.net

Using timezone: Central European Time
* arjanb leaves03:54
* arjanb joins12:11
<DRMacIver>Hm.13:00
* DRMacIver wonders how hard it would be to add annotation functionality to Nice.
It might be a good first project for me to get stuck into if there are no objections. :)
<arjanb>you mean like the java 5 annotations?13:34
<DRMacIver>Yeah13:46
I'm contemplating some reflective magic libraries which really need annotations to be viable.13:47
I can still *write* them in Nice without reflections, but it will result in a very weird project with the core annotations written in Java, included in a Nice library for use exclusively with Java projects. :)13:48
(Which is the default plan - I don't intend to write them in Java if I can help it. Multiple dispatch would be far too useful)13:49
<arjanb>afaik it has been decided annotations should be added to Nice14:18
it would be quite a challenge for a first project as it goes from parsing all the way to bytecode generation14:19
<DRMacIver>Ah14:20
In which case I'll probably leave it to someone else to coordinate. :)
* DRMacIver is feeling very out of depth with the Nice compiler at the moment
Any idea if this is planned for doing pre-1.0?14:21
Politically speaking it might be advisable to have them in place for the 1.0. Annotations are huge at the moment, so their lack is probably a major barrier to adoption. :)14:22
<arjanb>i don't think there's plan for 1.0, but it's a bad idea to be behind on the good new java features14:25
<DRMacIver>They're not exactly new any more. ;)14:26
Hm. Does !T on a type parameter mean something like "T and not null"?15:08
<arjanb>right15:09
<DRMacIver>I've encountered another case where the nullness inference doesn't work, although I'd hesitate to call it a bug.
(It would require some fairly subtle automated theorem proving to catch it)
* DRMacIver will submit it anyway.Might be worth looking at at some point in the future.15:10