[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]

Re: shower strikes again (whap!)

> Yes the SnarfPacker handles it and could keep track of these objects,
> but then it has to keep track of approximately 4 times as many
> objects.  Therefore I've been assuming that stubs are extremely
> lightweight and don't get indexed in the SnarfPacker.

On the other hand, I'd been assuming that the SnarfWhatsit would have a
list containing an entry for every in-RAM shepherd stub or shepherd whose
flock has a place in a snarf, just as a com-handler has a list of every
object with a proxy across its link.

>    Great!  Are there any things that mutate while being pointed at by half
>    the world?  (Changing the annotation on the inter-snarf pointers that
>    point at them could require pulling in lots of snarfs, which would be bad.
>    On the other hand, I suspect we don't now have any ent-objects that change
>    type in-place, or we wouldn't be having this conversation, right?  B-) )
> The annotations must be only compile-time type information, and that
> can't change, so we shouldn't have to change annotations anywhere.

In other words, ent objects, once they're pointed at, don't change type.
So everything works.