Peter TB Brett

I suggest using "Whole-archive signature" vs. "Per-file signature".

@runrevmark I agree. Let's defer MIMEtypes to a separate discussion. - I'd like to suggest using [RFC 2119](https://tools.ietf.org/html/rfc2119) in this spec. - You asked me to remind you to add...

@runrevmark Can you add a bit about the unpacked representation? What about unpacking the `${LC_MODULE_DIR}/org.example.foo.lce` package into the `${LC_MODULE_DIR}/org.example.foo.lce/` directory?

Suppose we have a directory `/path/to/dir/` in the module search path. This directory is permitted to contain: - bare modules (`.lcm` files) - extension archives (`.lce` files) - presumably, unpacked...

@montegoulding I'm unclear whether this patch is ready for merge or not.

@montegoulding Is this relevant to your interests?

@montegoulding: To clarify, whenever you get a new `MCObject` instance -- no matter how -- it starts off with no UUID. See https://github.com/peter-b/livecode/commit/3cf5f8e71e39b09f0d19228369d951d233015aeb. Mutation operators would only need to know...

- I expect this will be LiveCode 8-only, if/when it gets merged. - An object gets a UUID the first time you ask it for one, or when you explicitly...

@montegoulding: Well done, you have identified one of the big sticking points I ran into. I couldn't find a solution to this problem that would be always unambiguously correct and...

@montegoulding LCB code is at a lower level than the LiveCode engine and has no visibility of LiveCode objects, so no. What about using script-only stacks? What I really wanted...