On 2013-28-08 21:33, Andy Parker wrote:
So, short of swapping in the egrammar in as the default, I'm actually
not seeing any change that is requiring the bump to puppet 4. It would
be *nice* to remove deprecated functionality, and it would be really
nice to get the egrammar, but we can also continue working on that as
the feature flag, flesh it out some more, make it faster, even change
the semantics of parts of the puppet language to something sane, all in
the puppet 3.x series.

Fixing the semantics of undef is hard without breaking things.

One more release on 3.x with more deprecations could be nice anyway while also getting one more round of work on future parser out for beta testing.

- henrik

So...I'll open this up. Should we continue with Puppet 3..x? Is there a
reason that we absolutely need to break compatibility to move forward?


On Wed, Aug 28, 2013 at 12:22 PM, Rob Reynolds <[email protected]
<mailto:[email protected]>> wrote:




    On Wed, Aug 28, 2013 at 10:45 AM, Andy Parker <[email protected]
    <mailto:[email protected]>> wrote:

        On Wed, Aug 28, 2013 at 7:27 AM, Henrik Lindberg
        <[email protected]
        <mailto:[email protected]>> wrote:

            [snipped]


        I would have a slightly different list:

           * Split the codebase (master/agent/common?
        apps/types+providers/language? something else?). We need to
        decide where the right lines are to make the split and then
        figure out the work that it will take to get there. This may or
        may not change how puppet is deployed, but I think we can do it
        in a manner where we deploy in the same manner as right now.
           * Promote the egrammar to the default grammar. Do we keep the
        old grammar around for ease of migration for users (I think we
        shouldn't)?
           * Optimize the egrammar. It has to be faster than the
        existing grammar, otherwise I don't think we can promote it.
           * #8040 - anchor pattern. I think a solution is in sight, but
        it didn't make 3.3.0 and it is looking like it might be
        backwards incompatible.
           * Puppet Doc - I agree, this has sat and rotted for too long.
        It doesn't show up on any of our roadmaps :( It seems to have
        just been forgotten about.
           * Remove all of the currently deprecated functionality
        (activerecord store configs, many other things)
           * Reboot support - This is being worked on, but didn't make
        it in time for Puppet 3.3 (part of it did, but not all of it)


    Reboot support only needed small bits into puppet core. What we
    would look at for 4 would be integrating the core type into puppet
    proper and just having the provider in the puppetlabs-reboot module.
    Another thing we would add to puppet is having the ability to report
    on why resources get skipped
    (http://projects.puppetlabs.com/issues/15963)


        Something that we also need to work out is how much Puppet 4 can
        break compatibility. I've been aiming for the conservative
        approach were we try to only break compatibility, when possible,
        by first providing a deprecation warning for the old behavior,
        have the new behavior available at the same time and then remove
        the old behavior in a major release. This approach would mean
        that we can't change language semantics for the Puppet 4 release
        because we haven't deprecated a lot of the behavior and provided
        a new set of behaviors. I'm willing to be told that we can take
        a larger step than that, though.

        [snipped]



    --
    Rob Reynolds
    Developer, Puppet Labs

    Join us at PuppetConf 2014, September 23-24 in San Francisco

    --
    You received this message because you are subscribed to the Google
    Groups "Puppet Developers" group.
    To unsubscribe from this group and stop receiving emails from it,
    send an email to [email protected]
    <mailto:puppet-dev%[email protected]>.
    To post to this group, send email to [email protected]
    <mailto:[email protected]>.
    Visit this group at http://groups.google.com/group/puppet-dev.
    For more options, visit https://groups.google.com/groups/opt_out.




--
Andrew Parker
[email protected] <mailto:[email protected]>
Freenode: zaphod42
Twitter: @aparker42
Software Developer

*Join us at PuppetConf 2014, September 23-24 in San Francisco*

--
You received this message because you are subscribed to the Google
Groups "Puppet Developers" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/puppet-dev.
For more options, visit https://groups.google.com/groups/opt_out.


--
You received this message because you are subscribed to the Google Groups "Puppet 
Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/puppet-dev.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to