-1
There is concern that has been voiced by users in which this may
alienate them from moving to Accumulo 1.6. Many arguments for moving to
Java 1.7 are based on wanting to use some new feature or syntactic
sugar. Also, not building artifacts against Java 1.7 would not keep
people from running with Java 1.7 (even though Accumulo was built
against 1.6).
On the other side, Oracle has stated that no further security-related
issues will be patched on Java 1.6. Given that most people using
Accumulo are security minded, this is important.
Forcing Java 1.7 alienates a group of users. Allowing users to run with
Java 1.6 or 1.7 virtual machines satisfies all parties. As such, I don't
believe this is best.
Like John said, I agree that adequate discussion hasn't been had here to
justify forcing a change. Accumulo is not that popular that we can force
people to do what we think is best. I would be happy to continue to
participate in discussions as to the concrete benefits forcing Java 1.7
provides.
On 06/05/2013 04:42 PM, Christopher wrote:
The vote was already called, and it was a vote on whether we should do
it now (now, as in for Accumulo 1.6.0 development). If you think more
time is needed, then your vote should be no. I don't think it's
productive to continue to have a meta-discussion about whether or not
a discussion/vote should occur. Just vote "-1", with a reason "not
enough time to address potential concerns".
--
Christopher L Tubbs II
http://gravatar.com/ctubbsii
On Wed, Jun 5, 2013 at 2:14 PM, John Vines <vi...@apache.org> wrote:
Given this thread, I think more discussion is necessary before a vote.
On Wed, Jun 5, 2013 at 1:54 PM, Christopher <ctubb...@apache.org> wrote:
All-
Please explicitly vote in favor or against changing the java
dependency to >=1.7.
Parsing vague "may cause..." or "might be..." concerns throughout the
text of the thread is tedious, and does not help me know what the
consensus of the group is, so we can move forward. If there's a
specific issue that is informing your vote, that's great, feel free to
state it, but I don't want this issue to drag out for the duration of
the the Accumulo 1.6.0 development cycle because people are reluctant
to come to a concrete opinion.
If it fails a vote, we'll revisit for Accumulo 1.7.0.
I'm personally in favor of the change (+1), but it's not a big deal to
me. I just want a concrete resolution.
--
Christopher L Tubbs II
http://gravatar.com/ctubbsii
On Tue, Jun 4, 2013 at 10:51 AM, John Vines <vi...@apache.org> wrote:
I have also heard mulling about issues with the way Kerberos
authentication
behaves with JDK1.7 for hadoop. This may also have implications on the
Accumulo implementation as well.
On Tue, Jun 4, 2013 at 8:21 AM, Sean Busbey <bus...@cloudera.com> wrote:
On Mon, Jun 3, 2013 at 10:37 PM, Ben Popp <b...@sqrrl.com> wrote:
<snip>
CDH4 claims JDK 1.6 and 1.7 support:
http://www.cloudera.com/content/cloudera-content/cloudera-docs/CDH4/latest/CDH4-Requirements-and-Supported-Versions/cdhrsv_topic_3.html
<snip>
CDH4 comes with some additional caveats about 1.7:
http://www.cloudera.com/content/cloudera-content/cloudera-docs/CDH4/latest/CDH4-Release-Notes/cdh4rn_topic_2_2.html?scroll=concept_c1n_bln_tj_unique_1
The biggest one being the disclaimer about 1.7 compiled code.
--
Sean
