Brendan Eich wrote:

> Item 2, automated conversion to use C++ exceptions instead of nsresults,
> rotating out parameters back into return values, is also something we're
> about ready to experiment with, based on Taras's great elsa-based tool
> work. I hope to hear results soon, in this newsgroup even.

What I really want out of this is the ability to throw and catch the same
exception type(s) from tamarin (JS) that we do in binary code. This is a big
challenge, because the Tamarin engine doesn't actually use binary exceptions
currently.

> The bulleted detailed points after this list of three are less clearly
> desirable, or I think wrong. We don't want nsISupports <: GCRoot. Do we

I didn't understand what GCRoot was at that point ;-) But I think we've
covered that effectively in the other thread. If virtual inheritance weren't
so horribly messy, perhaps nsISupports should virtually inherit from
GCObject to make the contract clear, but I think that would bloat each
instance with an extra vtable.

> really want GetInterface from nsIInterfaceRequestor folded into
> nsISupports, along with QueryInterface?

Probably not... certainly not right away, let's focus on the big wins first.

> Likewise I'm skeptical about weak ref support -- I do not think that's
> common enough to warrant being in the top abstract interface, and it
> falls under the future-of-XPCOM-memory-management thread, where we're
> thinking in pretty concrete MMgc terms.

I think I disagree, but the move to MMGc makes it irrelevant because you can
hold a weakref to any object you wish.

--BDS
_______________________________________________
dev-tech-xpcom mailing list
[email protected]
https://lists.mozilla.org/listinfo/dev-tech-xpcom

Reply via email to