Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Accumulo >> mail # dev >> Backporting policy proposal

Copy link to this message
Re: Backporting policy proposal

On 6/18/13 11:26 AM, Adam Fuchs wrote:
> On Jun 18, 2013 11:05 AM, "Josh Elser" <[EMAIL PROTECTED]> wrote:
>> On 6/18/13 10:53 AM, Adam Fuchs wrote:
>>>>> 2. Back-porting performance improvements to a prior release line that
>>>>> is not EOL (end-of-life) is usually okay (subject to normal lazy
>>>>> consensus), so long as it does not change user-facing behavior or API.
>>>>> It is still strongly preferred to add such fixes in the older branch
>>>>> first, and merge forward to the newer one(s).
>>> Agree, although doesn't the transition to git alleviate the problems with
>>> order of operations?
>> I don't understand what you mean by "order of operations". If you push a
> commit in 1.6.0-SNAPSHOT that should really be in 1.4.4-SNAPSHOT, Git will
> not handle this well in terms of an accurate history.
> I don't really understand either -- that was mostly speculation. Doesn't
> the reliance on hashes of the patch rather than commit IDs make it
> immaterial as to which branch a patch was committed to first? Admitedly, I
> am not an expert on git.
> Adam

No, that's incorrect. A commit is uniquely identified by a set of
attributes which, most importantly, include the parent commit. As such,
what you would have typically (happily) done with subversion (`svn merge
-r 12345 .`) has implications to the Git history graph.

The only way to accomplish this in Git is to cherry-pick that commit
which causes multiple unique commits which apply the same changes to
content. This also makes using tools like `git-bisect` much less
effective as a bug introduced in a cherry-pick'ed commit now has
multiple points of introduction instead of a single commit.

In the downtime before we hear back from INFRA, it would be good to read
through the documentation: http://git-scm.com/documentation