Patches for packager tweaks

[Kevin Rushforth]

Right. I was answering the general question.

For the specific question, I will defer to Mark Howe, who is working on 
the packager.

-- Kevin


Daniel Zwolenski wrote:
> I'm guessing Danno would like to know how long he should expect to 
> wait for the patches he kindly contributed and linked to in that email 
> to get included. Seems like a fair and reasonable question and one I'd 
> also like to know the answer to. 
>
> Perhaps a linked question that I'd also like to know: is anyone 
> actually allocated to any work on the packager at the moment, and if 
> not when are they next going to be?  
>
>
> On Fri, May 31, 2013 at 11:44 AM, Kevin Rushforth 
> <kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com>> wrote:
>
>
>         How long is it taking for community patches to get into a
>         build these days?
>
>
>     Hi Danno,
>
>     There are two parts to the answer:
>
>     1) How long does it take for a proposed fix (patch) to be reviewed
>     and accepted?
>
>     2) Once your patch is accepted and the changeset is pushed to the
>     repo, how long before it shows up in an EA build?
>
>     #1 depends on what area it is, what is the scale of the proposed
>     change: is it a simple bug fix, or a new feature with API or
>     documentation implications, are there compatibility concerns, how
>     risky is it, etc.
>
>     #2 is typically between 0.5 and 1.5 weeks depending when it is pushed.
>
>     As a reminder (Richard may have recently posted something about
>     this, so my apologies if this is a duplicate reminder), anyone
>     submitting a patch must first sign the Oracle Contributor
>     Agreement (OCA) before we can consider taking it.
>
>     -- Kevin
>
>
>
>
>     Danno Ferrin wrote:
>
>         Just posted to bugs with patches for some tweaks I'de like to
>         see to the
>         packager.
>
>         https://javafx-jira.kenai.com/browse/RT-30792
>         https://javafx-jira.kenai.com/browse/RT-30793
>
>         The first one is to allow for a comma separated list of enumerated
>         packagers, so it's not one or all.
>
>         The second one is more relevant, it moves the discovery of the
>         bundlers
>         from being hard coded in the class file to loaded off of the
>         META-INF/services directory.  This allows a bunlder to be added at
>         "runtime" when the build is being done.  For example, a
>         bundler that would
>         bundle RoboVM or APK files provided at runtime rather than
>         having to
>         package it's reference into the source code itself.
>
>         How long is it taking for community patches to get into a
>         build these days?
>          
>
>