Peter TB Brett

Discussion with @runrevmark today (offline). About the unresolved issues: the thing that determines a decoding error is if we encounter something while decoding that we can't represent internally. That means:...

@runrevmark Running the tests for this results in SIGSEGV somewhere in libscript. I think it's due to the conflict with the `is a` type comparison operator.

Obviously, this is a very minimal trait system, and omits neat things like: - Traits with type parameters - Associated types - Generic handlers bounded by traits etc.

@runrevmark I've been thinking about the issues we discussed last night over breakfast. I think that the way traits are proposed to work is going to have to change in...

I've updated and greatly revised the traits spec to reflect today's very interesting discussions!

@runrevmark @livecodestephen @livecodefraser: I've updated the spec again to reflect some discussions over the last couple of weeks. For example, I've removed `derived` traits (for now), removed explicit `me` arguments,...

@montegoulding I'm not sure this is a good idea. Several places in this patch are exposing places where APIs should take `const char*` argument rather than `char *` arguments, and...

Looks great! My only comments would be: - avoid redundancy between the filename and the handler name, because both get logged by the test framework - Feel free to use...

@montegoulding As long as it's building locally, don't worry about it for now.

Per [RFC 6838](http://tools.ietf.org/html/rfc6838) I propose `application/vnd.livecode.builder-package`.