I agree with dropping JDK 7 support and 2.0 would be a good flag to
users.  I opened a pull request to require Java 7 and give us some
tangible benefit, > 2 GB single-part payloads[1].  We should allow the
possibility of further 1.7.x releases but not commit to a schedule.
Waiting until September for 2.0 to address our Guava 16 incompatibility
seems like a long time given the increasing number of users encountering
this issue.  Can we continue the release schedule discussion as a
separate thread on jclouds-dev since we should drive the release date
based on which features scope into a convenient time-frame?

[1] https://github.com/jclouds/jclouds/pull/426

On Mon, Jun 30, 2014 at 08:12:18PM +0000, Everett Toews wrote:
> +1
> 
> Let’s go for it.
> 
> There are some very good reasons for leaving Java 6
> 
> 1. The Java 6 EOL was over a year ago (this alone is reason enough). [1]
> 2. Java 6 has been a security nightmare on the desktop and we shouldn’t 
> encourage its use anywhere. [2]
> 3. Java 7 has reached a critical mass. [3]
> 
> I’ve also emailed with several large Rackspace customers and gathered some 
> metrics too. There will be an impact to some customers but I’ve done my due 
> diligence and am confident we can and should move forward (I confess I’m a 
> bit nervous about it).
> 
> Switching to Java 7 is a HUGE backwards incompatible change and I think the 
> versioning must reflect that.
> 
> Here’s what I propose.
> 
> jclouds 1.7.4 (due mid-July) is the last version of jclouds to work with Java 
> 6.
> 
> jclouds 2.0  (due early Sept.) is the first version of jclouds to work with 
> Java 7.
> 
> As of jclouds 2.0 we adopt the major release cadence (which we’re converging 
> on).
> 
> WDYT?
> 
> Everett
> 
> [1] http://www.oracle.com/technetwork/java/eol-135779.html
> [2] 
> http://www.darkreading.com/vulnerabilities-and-threats/hackers-target-java-6-with-security-exploits/d/d-id/1111293?
> [3] http://pages.zeroturnaround.com/Java-Tools-Technologies.html
> 
> 
> On May 30, 2014, at 2:52 PM, Everett Toews 
> <[email protected]<mailto:[email protected]>> wrote:
> 
> First off, I’m totally on board with the move to Java 7. Beyond Java 6 being 
> EOL, Java 7 has finally reached some kind of critical mass [1].
> 
> My primary concern is around the kind of change it is and the timing of it. 
> This is a major backwards incompatible change. As soon as we write that first 
> line of code using a Java 7 feature, we’re leaving a lot of people behind.
> 
> If 1.8 is our next release after 1.7.3 then dropping support for it seems a 
> bit sudden for our users (both Rackspace and jclouds in general). However, 
> I’m not entirely certain of that. I would prefer this to be a data-driven 
> decision and I need some time to dig into the data.
> 
> Please give me until the end of June to gather some data and poll our 
> customers.
> 
> I also think we should be prepared to support the last version of jclouds 
> with Java 6 longer than our usual support (i.e. an exception to our N-1 
> policy). This is especially important for security issues. I’m sure we have 
> users from all across the jclouds community that will not be able to upgrade 
> from Java 6 for X amount of time for Y reason. We can’t abandon these 
> developers.
> 
> Everett
> 
> [1] http://jaxenter.com/what-s-cool-in-java-these-days-50419.html
> 
> 
> On May 30, 2014, at 1:52 AM, Ignasi Barrera 
> <[email protected]<mailto:[email protected]>> wrote:
> 
> 
> Totally agree with Jeremy and Chris.
> 
> I think it is reasonable to think that users that don't want to upgrade the 
> JVM won't want to upgrade to the latest jclouds version (major changes, 
> newest Guava and other deps, etc).
> 
> Keeping 1.7 compatible with 1.6 and dropping support for it in 1.8 sounds 
> like a reasonable way to move forward.
> 
> El 30/05/2014 00:46, "Chris Custine" 
> <[email protected]<mailto:[email protected]>> escribió:
> I agree with Jeremy, in that 1.8 seems like a reasonable point to start 
> requiring Java 7 as a minimum.  I know many people are still using jclouds 
> 1.6.x and Java 6 so I don't think you will cause too much grief for anyone 
> moving up to the eventual jclouds 1.8 release (meaning, they are likely to 
> stay at jclouds 1.6 or 1.7 for quite some time anyway).
> 
> Just my 2 cents.
> 
> Chris
> 
> --
> Chris Custine
> 
> 
> 
> On Thu, May 29, 2014 at 9:32 AM, Jeremy Daggett 
> <[email protected]<mailto:[email protected]>> wrote:
> I really appreciate that you brought this up Andrew!
> 
> I have been through every Java version transition since v1.0. I know the
> challenges and pain points this decision might cause some developers. It
> will most likely be an issue in enterprise environments where Java
> runtimes can only change every couple of years.
> 
> Technology changes. Java 6 is no longer supported. That alone should be
> the deciding factor for the project.
> 
> 
> In my experience, the industry best practice is to support the most
> current version and one previous. I actually really like what vert.x did
> in requiring a Java 7 (or newer) right from the start.
> 
> Here is an idea for a potential transition plan:
> jclouds 1.7.x -> Java 6+ (maintenance branch to support Java 6+)
> jclouds 1.8.x -> Java 7+
> 
> 
> In order to move jclouds forward, we need to drop support for Java 6. If
> it were totally up to me, I would vote to jump right to Java 8, but that
> is just my perspective. ;)
> 
> /jd
> 
> On 5/28/14, 11:57 AM, "Andrew Gaul" <[email protected]<mailto:[email protected]>> 
> wrote:
> 
> >jclouds presently supports Java 6, 7, and 8 which imposes extra
> >development costs and prevents uptake of new language and library
> >features including try-with-resources, NIO.2, and HTTP client
> >improvements.  Oracle ceased public updates to Java 6 in early 2013[1]
> >and jclouds could use this to guide its support strategy.  The jclouds
> >developers would like to understand how many users continue to use Java
> >6 and what prevents upgrading to newer versions.  Please respond to this
> >thread with any relevant information.  Thanks!
> >
> >--
> >Andrew Gaul
> >http://gaul.org/
> 
> 
> 
> 

-- 
Andrew Gaul
http://gaul.org/

Reply via email to