If we are going to be encouraging index downloading, we need to setup
some metadata to ensure that the index (file) and the indexers (code)
are kept in sync.
So if a 1.0 indexer code attempts to 'refresh' the index from a remote,
and sees that the stored index (file) is for indexer (code) v1.1+ then
it doesn't corrupt it's local index.
- Joakim
Brett Porter wrote:
Hi Milos,
I'm a bit lost as to what you need.
So, say you have a repository manager at http://localhost:8080/
Then you have managed repositories called "snapshots" and "releases".
You will be able to access those repositories at:
http://localhost:8080/snapshots/ and http://localhost:8080/releases/
respectively.
The index will just be one (say,
http://localhost:8080/retrieveMinimalIndex.action). Inside that, each
record has an ID that will be able to determine whether it is releases
or snapshots.
At present, we are not tracking where it came from if it was via a
proxy or a synced repo. I inted to add that, but I don't think it is
important for the purposes you are describing.
Once this works, we will need to look at other ways of making the
index updates more efficient (Eugene has had some thoughts on that and
has posted them to JIRA).
So, does this fit what you needed or am I still missing something?
- Brett
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
