> On Aug 8, 2013, at 17:52, Ryan Baxter <rbaxte...@apache.org> wrote:
> 
> Does anyone think it is work going to the OpenSocial list and asking
> if post 2.5.1 we stick to using only major.minor versioning?  This
> allow Shindig to stick to use the revision field for updates that are
> not necessarily spec related.

+1

> 
>> On Wed, Aug 7, 2013 at 5:02 PM, Henry Saputra <henry.sapu...@gmail.com> 
>> wrote:
>> The OpenSocial specs have major.minor.revision versioning scheme, so the
>> discussion is either we follow OpenSocial versioning up to the "revision"
>> or just major.minor.
>> 
>> If we want to follow up to the "revision" part then we need to add
>> "-updateXXX" or "rXXX" tail in Apache Shindig versioning which I think a
>> bit ugly (see Hadoop release versions, ugh)
>> 
>> - Henry
>> 
>> 
>> On Tue, Aug 6, 2013 at 10:30 PM, Marcel Offermans <
>> marcel.offerm...@luminis.nl> wrote:
>> 
>>> Maybe it makes sense to use semantic versioning, as described here:
>>> http://semver.org/
>>> 
>>> If indeed the specs are major.minor then you can implement with
>>> major.minor.patch
>>> 
>>> Greetings, Marcel
>>> 
>>>> On Aug 7, 2013, at 7:14 AM, Craig McClanahan <craig...@gmail.com> wrote:
>>>> 
>>>> FWIW an approach to version numbers I have seen a lot is that the
>>>> major.minor version numbers match the spec being implemented (2.5 in this
>>>> case), but anything after that is totally up to the implementation.  So,
>>>> 2.5.1 ... 2.5.2 ... etc. would be fine for improved implementations of
>>> the
>>>> 2.5 spec.
>>>> 
>>>> It's also perfectly reasonable to think about 2.5.1-rc1 and 2.5.1-rc2
>>> (and
>>>> so on) for release candidates of 2.5.1 before a final 2.5.1 release.
>>>> 
>>>> Craig
>>>> 
>>>> 
>>>>> On Tue, Aug 6, 2013 at 2:03 PM, Ryan Baxter <rbaxte...@gmail.com> wrote:
>>>>> 
>>>>> Henry does JSF use revision numbers or do they stick to major and
>>>>> minor version only?  If the revision numbers are not going to match
>>>>> anyways than maybe we shouldn't worry about calling the next Shindig
>>>>> release 2.5.1.
>>>>> 
>>>>> On Mon, Aug 5, 2013 at 9:41 PM, Henry Saputra <henry.sapu...@gmail.com>
>>>>> wrote:
>>>>>> The next version of Shindig will contain the first release without PHP.
>>>>>> I believe that's worth a bump in revision part of the version to 2.5.1.
>>>>>> 
>>>>>> @Ryan, I would love to keep it in-sync but Shindig will move faster
>>> than
>>>>>> spec development and it would be hard to keep up with the exact
>>> version.
>>>>>> I believe other projects like Apache Tomcat and Apache MyFaces (latest
>>>>>> release is 2.1.12 that implement JSF 2.1) that also implement open
>>> specs
>>>>>> follow its release version up to certain levels.
>>>>>> 
>>>>>> I like @Matt's idea, we could release  2.5.1-alphaX releases but final
>>>>>> release should be 2.5.1 but branching early to add minor emergency
>>> fixes
>>>>>> with 2.5.0-update1.
>>>>>> 
>>>>>> - Henry
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> On Mon, Aug 5, 2013 at 5:08 PM, Matt Franklin <
>>> m.ben.frank...@gmail.com
>>>>>> wrote:
>>>>>> 
>>>>>>> 
>>>>>>>> On Aug 5, 2013, at 20:03, Henry Saputra <henry.sapu...@gmail.com>
>>>>> wrote:
>>>>>>>> 
>>>>>>>> I am actually still +1 for just 2.5.1. We agreed that Shindig version
>>>>>>> will
>>>>>>>> adhere to OpenSocial specs up to minor version which in this case is
>>>>>>> 2.5.x
>>>>>>> 
>>>>>>> What about developing in trunk at 2.5.1-alphaX and branching for fixes
>>>>> in
>>>>>>> 2.5.0-update1?
>>>>>>> 
>>>>>>> I also think 2.5.1 should be relatively minor in changes to the
>>>>> software
>>>>>>> itself.  Ideally, only additions and no breaking changes to existing
>>>>>>> interfaces, etc.
>>>>>>> 
>>>>>>> 
>>>>>>>> 
>>>>>>>> - Henry
>>>>>>>> 
>>>>>>>> 
>>>>>>>>> On Mon, Aug 5, 2013 at 4:50 PM, Ryan Baxter <rbaxte...@gmail.com>
>>>>>>> wrote:
>>>>>>>>> 
>>>>>>>>> Here is what I found on version numbers [1].  From what I gather
>>>>> after
>>>>>>>>> reading that using 2.5.0.1 would be considered "non-standard".  The
>>>>>>>>> only downside to this would be the version numbers would be compared
>>>>>>>>> as strings.  We could use 2.5.0-fix1 which would be considered
>>>>>>>>> standard, but I don't think that buys us anything with regards to
>>>>>>>>> version comparison.  I could pose a question to the Maven users list
>>>>>>>>> and see if they have any advice.
>>>>>>>>> 
>>>>>>>>> [1]
>>> http://books.sonatype.com/mvnref-book/reference/pom-relationships-sect-pom-syntax.html
>>>>>>>>> 
>>>>>>>>> On Mon, Aug 5, 2013 at 7:34 PM, Stanton Sievers <
>>> ssiev...@apache.org
>>>>>> 
>>>>>>>>> wrote:
>>>>>>>>>> +1. Shindig-1924 is one such cleanup. I also agree with staying in
>>>>> line
>>>>>>>>>> with the spec.
>>>>>>>>>> 
>>>>>>>>>> I would just want to make sure we have no technical or process
>>>>> issues
>>>>>>>>> with
>>>>>>>>>> maven artifacts (or the like) with 4 numbers in the version.
>>>>>>>>>>> On Aug 5, 2013 7:29 PM, "Ryan Baxter" <rbaxte...@apache.org>
>>>>> wrote:
>>>>>>>>>>> 
>>>>>>>>>>> The current version of trunk is set to 2.5.1.  I am wondering what
>>>>>>>>>>> people think of changing that to 2.5.0.1?  There are a few cleanup
>>>>>>>>>>> changes that have already been identified that would be good to
>>> get
>>>>>>>>>>> out there.  At same time we want to stay in sync with the spec
>>>>> version
>>>>>>>>>>> so I don't think we want to release 2.5.1 yet.  What does everyone
>>>>>>>>>>> think?
>>> 
>>> 

Reply via email to