> On Feb 12, 2018, at 9:14 PM, Yudhi Karunia Surtan <brainmaster...@gmail.com> 
> wrote:
> 
> Do I need to create an issue and create a new branch, or just commit the
> changes to the master branch?

Hi Yudhi,

Typically when we make a change, we’ll notify the project about the issue, the 
fix, etc, (which you have already done).  I always create an issue in JIRA, 
with a detailed description of the problem and the fix, and use the jira issue 
#, along with a short desc, in the commit comments.

e.g. FC-231 - DAO's should be package private

That way we can tie the change back to the issue, where longer description of 
the change may take place.

For something trivial, like a code comment, or some other self-explanatory 
change, I will skip the step to create the JIRA issue, and just commit, with 
the short comment.  

For really disruptive changes, i.e. changing the API, data structures, we 
discuss on this ML the pros/cons, reach consensus, before going forward.  In 
these cases we might create a branch, because of the duration it takes to 
complete the work.

For simple changes, like that which you’ve brought up now, simply committing 
directly to trunk is fine.  

One word of caution, I always make sure the junit tests pass successfully 
before checking in a change.  The idea if the tests pass, no regressions, or 
other known bugs have been inserted.  If the tests pass, and a bug is introduce 
anyway, we know its time to add a new test case.

Let me know if you have any questions about how to do this, which tests, etc.

Good questions btw, please feel welcomed to continue asking these questions 
here.  One of these days I’m going to document these procedures, for our coding 
handbook, in the meantime, apologize for the tribal knowledge.  :-)

Shawn

Reply via email to