mpdiffs, bundles and ghosts

[Robert Collins]

On Mon, 2008-06-02 at 09:05 -0400, Aaron Bentley wrote:
> 
> 
> I think it's reasonable for mpdiffs to specify parents by position.
> This decouples them from any particular way of naming parents.  This
> is
> also what knit deltas do, though with knit deltas, it's implicitly
> always parent 0.

I hadn't quite got the layering and data present in a bundle straight.
Now that I do I think its simply a bug in the generation logic.

That said, its a defect in knits that the parent isn't explicitly
listed; and to have regeneratable indices we need a single parseable
block of data that lists the parents - you are right that we can nest
things:
parents list
delta using parents list indices
to get a composite which has both things, without forcing the lower
layer to know about parents-as-tuples or parents-as-strings etc.

-Rob
-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20080603/77825fb9/attachment.pgp