reintroducing root ids

[Jan Hudec]

On Wed, Feb 08, 2006 at 20:09:09 +0800, James Henstridge wrote:
> On 08/02/06, Jan Hudec <bulb at ucw.cz> wrote:
> > > The problem with this is if two people perform parallel conversions
> > > with different sets of reachable baz-1.x branches they will get
> > > different root IDs.

Um, I just realized the branches DON'T have to be ACCESSIBLE. The logs
are right there and all the information can be gathered from them
(provided they either were not pruned, or both conversions have pruned
from the same point).

> > Yes. But I don't see any way to fix it. Do you?
> 
> No, but if you are indexing some data by root ID, it would be a
> problem if you were to combine the two separate conversions.
> 
> That would make the root ID an inappropriate property to index on.

The data are indexed by file-id. Root-id is a special case of that. It
is the only property by which the data can be indexed.

Or from another POV. We need root-id, for many different reasons. The
question is how to generate it for conversions from Arch. If you can
come up with anything better than I did, please do so. But not
generating them and not using them is not an option.

-- 
						 Jan 'Bulb' Hudec <bulb at ucw.cz>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060208/c0746e8c/attachment.pgp