Cloudera Development Kit is Now “Kite SDK” by Ryan Blue.
From the post:
CDK has a new monicker, but the goals remain the same.
We are pleased to announce a new name for the Cloudera Development Kit (CDK): Kite. We’ve just released Kite version 0.10.0, which is purely a rename of CDK 0.9.0.
The new repository and documentation are here:
- Kite repository: https://github.com/kite-sdk/kite
- Kite documentation: http://kitesdk.org/
- Kite examples: https://github.com/kite-sdk/kite-examples
Why the rename?
The original goal of CDK was to increase accessibility to the Apache Hadoop platform by developers. That goal isn’t Cloudera-specific, and we want the name to more forcefully reflect the open, community-driven character of the project.
Will this change break anything?
The rename mainly affects dependencies and package names. Once imports and dependencies are updated, almost everything should work the same. However, there are a couple of configuration changes to make for anyone using Apache Flume or Morphlines. The changes are detailed on our migration page.
The continuation of the Kite SDK version 0.10.0 along side the Cloudera Development Kit 0.9.0, should make some aspects of the name transition easier.
However, when you search for CDK 0.9.0, are you going to get “hits” for the Kite SDK 0.10.0? Such as blog posts, tutorials, code, etc.
I suspect not. The reverse won’t work either.
So we have relevant material that is indexed under two different names, names a user will have to remember in order to get all the relevant results.
Defining a synonym table works for cases like this but does have one shortfall.
Will the synonym table make sense to us in ten (10) years? Or in twenty (20) years?
There is no guarantee that even a synonym mapping based on disclosed properties will remain intelligible for X number of years.
But if long term data access is mission critical, something more than blind synonym mappings needs to be done.