I think Chimera goal is to enhance even for other use cases. For Hadoop,
CTR mode should be enough today, but when we want to support other modes
for other users(ex:While a lot of encryption such as network encryption or
data transfer encryption over the wire doesn't necessarily CTR, other
modes such as CBC and GCM are better modes for this use cases, hadoop can
use them as well in some form later), I think separate module and
independent release is good idea but I am not so strong on the point to
keep under Hadoop. It may be good to keep at generalized place(As in the
discussion, we thought that place could be Apache Commons). But lets see
what others feel. Getting attention to more Hadoop use cases will be
depending on community IMO. Irrespective of where we place, if community
is same then attention would be same right.
To summarize my points:
1. To add a subproject with independent release in Hadoop actually shows
it is something may be widely shared. May nicely and better fit in Apache
Commons
2. Sub projects with independent release cycles in Hadoop may complicate
the Hadoop releases and versions and there is no existing practice in
Hadoop yet. 

Others also please comment what you guys feel, to conclude further steps.

Regards,
Uma



On 1/26/16, 11:26 AM, "Owen O'Malley" <omal...@apache.org> wrote:

>Sorry to be coming in to this discussion late. Rather than pull the code
>out of Hadoop, may I suggest instead making it a separate subproject
>within
>Hadoop itself? I'd suggest letting it release independently of Hadoop,
>since it will need a much faster cadence that Hadoop proper does. It
>should
>also keep the number of dependencies to a very small set (empty?).
>
>Thoughts?
>   Owen

Reply via email to