runrevmark

@peter-b: Can you clarify what you mean? The unpacked representation is just the unpacked zip...

It would be interesting to find out precisely what the bottleneck is on long scripts - e.g. when you add /\* to the start. Part of the problem, I suspect,...

@montegoulding: I think this can be happily removed - as long as the IDE handles extra (stackfile) parameters and opens them (on Windows if you double-click a stackfile, then you...

I don't think that's been done yet - this is currently a low-level internal API which was to be used by the internal implementation of stackdir IIRC. To leverage it...

@mwieder: If finished, this would end up only going into LC8. LC6.7 will be EOL'd in due course, and all being well there will be little need to support LC7...

@montegoulding: A more accurate statement would be to say that: LCB does not currently have direct 'supported' support for manipulating the LC object tree. However, that is not to say...

@montegoulding: If you don't want to hurt, don't do that :) I think the use-cases there need to be analyzed more as I'm not sure code which does such a...

@montegoulding: Indeed, but in that situation the application is basically in an error state. Remember that we can restrict the need for UUID references by ensuring that better referencing is...

@montegoulding: My main problem with UUIDs is that they offer no useful information at all. As soon as you have one which cannot be resolved it is useless. How do...

Btw, I should say that as a mechanism UUIDs are fine - but I'm not sure we've entirely got to the bottom of the problem exposing them is meant to...