I don't think we have a document on that anywhere, I assume here you are
talking about open/unassigned JIRAs possibly with patch(es) attached?

Generally speaking if it's unassigned on JIRA feel free to take it and
start looking at it, if someone is assigned but there's no apparent
activity you can always leave a comment asking if they'd like you to look
instead.  I know I like most of the other committers have some assigned
JIRAs which we started on and then never finished because of
work/time/personal constraints.

In terms of reviewing patches the general guidelines I tend to follow
based on what I and other people have expected to see in the past are as
follows:

Includes tests - If a bug report then you need test cases which
demonstrate this.  If it is missing tests generally ask the proposer of
the patch to update their patch to add some tests.

Copyright Header - For patches to existing code this shouldn't be an
issue, for new code make sure the Apache header is present.  As Andy
continues to remind me when I accidentally check in stuff with my
companies header (which Eclipse auto-inserts for me) only the original
author should change the copyright headers.  If the headers are wrong ask
the proposer to correct them.

No @author tags - We've agreed in the past not to use these, these you can
just remove yourself as and when you integrate patches if they are present.

Contribution to Apache - Apache policy is that any patch submitted via
JIRA or email is assumed to be a contribution by the author to the ASF
unless the author explicitly states otherwise.  Typically this should not
be an issue.

Contribution Size and Impact - Small patches can generally be incorporated
by yourself, larger patches - particularly those adding significant
features - should usually be discussed on the dev list prior to acceptance.

Depending on where a large patch comes from there may also be IP clearance
issues.  Use your judgement here, a few hundred lines of code may be
considered small if it isn't changing/extending functionality
significantly.  Conversely a small patch that changes a core behavior
should be more widely discussed.

If in doubt start a thread on dev or comment on the JIRA issue, JIRA
comments get copied to the dev list so all developers should see the
comments even if they aren't explicitly watching the issue.

Rob

PS. I will volunteer to write up a wiki/website page on this once others
have had chance to chime in if that would be helpful for future reference?



On 2/5/13 10:35 AM, "Claude Warren" <[email protected]> wrote:

>Is there a document that describes the approver process?
>
>How do you select which changes to review/accept?
>
>I have some time that I can spend reviewing changes but need to know
>where to start.
>
>thanks,
>Claude
>-- 
>I like: Like Like - The likeliest place on the web
>Identity: https://www.identify.nu/[email protected]
>LinkedIn: http://www.linkedin.com/in/claudewarren

Reply via email to