Can you translate all of this into user (or product) speak? Where does it leave us as far as any features we have already deployed? Features we have built but not deployed? Features we had hoped to build this quarter?


Kevin Smith
Agile Coach, Wikimedia Foundation

On Fri, May 6, 2016 at 7:46 AM, David Causse <> wrote:
I think this situation raises a problem related to feature planifications in our software based on upcoming features from upstream. In this case I assumed that the feature being merged in upstream it would have been available in elastic2.x that we plan to deploy very soon in our cluster.
The mistake I made here was to not actively monitor the elasticsearch github repo: the feature has been reverted few months ago, knowing this fact earlier would have certainly helped to review our plans accordingly.

Le 06/05/2016 10:43, David Causse a écrit :
Areek (the author of this new version) needed to add a breaking change which is only allowed during major version upgrades, the inter-node communication protocol would have been incompatible making rolling upgrades impossible.
Unfortunately there is nothing we can do except following more actively the elasticsearch github repo. In this case it was confusing since the feature was merged but then reverted.


Le 05/05/2016 20:45, Tomasz Finc a écrit :
That's disappointing but in no way diminishes your work David.

What do you need from the team to make sure we address the issue that
areek identifies?


On Thu, May 5, 2016 at 11:10 AM, David Causse <> wrote:

sad news, the completion suggester v2 was reverted from elastic 2.x
branch[1] and is now delayed to elasticsearch v5.
Unfortunately I was planning to use important features in this version to
add realtime support.
I'll start to think about it and see if we can workaround and still improve
the current implementation.

I'm really sorry about that, I should have monitored the elasticsearch repo
more closely...


discovery mailing list
discovery mailing list

discovery mailing list