You are viewing a plain text version of this content. The canonical link for it is here.
Posted to oak-dev@jackrabbit.apache.org by Angela Schreiber <an...@adobe.com> on 2013/04/03 10:15:18 UTC
BaseState of a Tree instance
hi
while taking a closer look at the base state associated with a TreeImpl
i got the impression at the same base NodeState is also available from
the NodeBuilder associated with the tree.
if this was really the case: what is the reason for keeping it twice?
regards
angela
Re: BaseState of a Tree instance
Posted by Michael Dürig <md...@apache.org>.
On 3.4.13 11:26, Michael Dürig wrote:
>
>
> On 3.4.13 9:15, Angela Schreiber wrote:
>> hi
>>
>> while taking a closer look at the base state associated with a TreeImpl
>> i got the impression at the same base NodeState is also available from
>> the NodeBuilder associated with the tree.
>> if this was really the case: what is the reason for keeping it twice?
>
> Yes this is something we need to clarify further. But AFAICS currently
> the purge logic is in the way since it causes the two base states to
> diverge. See also OAK-659 which is about moving that logic down below
> the NodeBuilder interface. It's not clear to me how to do that though
> since we don't have the information about the transient moves at that
> level and restoring it from looking at the content diffs doesn't work as
> discussed several times before.
See also https://issues.apache.org/jira/browse/OAK-744 for some follow up.
Michael
Re: BaseState of a Tree instance
Posted by Michael Dürig <md...@apache.org>.
On 3.4.13 9:15, Angela Schreiber wrote:
> hi
>
> while taking a closer look at the base state associated with a TreeImpl
> i got the impression at the same base NodeState is also available from
> the NodeBuilder associated with the tree.
> if this was really the case: what is the reason for keeping it twice?
Yes this is something we need to clarify further. But AFAICS currently
the purge logic is in the way since it causes the two base states to
diverge. See also OAK-659 which is about moving that logic down below
the NodeBuilder interface. It's not clear to me how to do that though
since we don't have the information about the transient moves at that
level and restoring it from looking at the content diffs doesn't work as
discussed several times before.
Michael
>
> regards
> angela