db-derby-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stanley Bradbury <Stan.Bradb...@gmail.com>
Subject Re: RegressionSearchAndDestroy ruling: Re: [jira] Created: (DERBY-1633) Regression: The fields of views are not being calculated properly since 10.1.2.4
Date Fri, 04 Aug 2006 01:09:04 GMT
Rajesh Kartha wrote:
> John Embretsen wrote:
>
>> Kathey Marsden wrote:
>>
>>> Prasenjit Sarkar (JIRA) wrote:
>>>
>>>> Regression: The fields of views are not being calculated properly 
>>>> since 10.1.2.4
>>>> --------------------------------------------------------------------------------

>>>>
>>>>
>>>>  
>>>>
>>> Good find Prasenjit, Thank you!
>>>
>>> Judges for the RegressionSearchAndDestroy contest:  can you give me 
>>> a ruling.  Should this count toward the contest?
>>> I'd like to  include any regressions that are found during the 
>>> contest period even if they are  10.1.3 regressions.
>>> Can we count this one?
>>
>>
>> As I understand it, this is something that worked correctly in 
>> 10.1.2.4, but that does not work in 10.2 (nor in 10.1.3.x).
>>
>> The rules say:
>>
>> "Find a new product regression from a previous Derby release and log 
>> it in Jira."
>>
>> I would say DERBY-1633 qualifies for the contest, and that the 
>> reporter has contributed toward a seamless upgrade to 10.2. I have 
>> not discussed this privately with the other judges, though, so you 
>> should not consider this the official ruling until they have had time 
>> to comment.
>>
>>
>>> http://wiki.apache.org/db-derby/RegressionSearchAndDestroy
>>
>>
>>
> I agree, this DERBY-1633 indeed fits the rules of the contest and 
> hence qualifies.
> Having said that  I would also like to see the fix getting ported to 
> the 10.1 branch.
>
> Regards,
> Rajesh
>
>
>
So the question is whether the term regression applies to features / 
functions that were released for production use or if features added to 
a snapshot (e.g. 10.1.2.4) but somehow did not make it into the next 
official release should be considered a regression.  This is a gray area.

Going forward we need to be clear on this point (Official release 
version only or Snapshots as well) and we can debate that before the 
next contest and make it clear [I lean towards official releases only].

For this contest, on the other hand, I think we should allow the entry 
since
1) The statement '..new product regression from a previous Derby 
release..' does NOT specify Official Release
       and when rules are unclear I prefer to be inclusive rather than 
exclusive
2)  I assume the feature was stable and production quality in 10.1.2.4
3) The feature should have been in 10.1.3.1
        (odd it was 'regressed' before 10.1.3.1 -  between May 9 and 
June 30)

Here is my understanding of Derby Releases (see the download page) are:
 Latest Official Release
    * 10.1.3.1 (Jun 30, 2006 / SVN 417277)
Previous Official Releases
    * 10.1.2.1 (Nov 18, 2005 / SVN 330608)
    * 10.1.1.0 (Aug 3, 2005 / SVN 208786)

SNAPSHOT DESCRIPTION:  Snapshot jars provide early access to bug fixes 
and/or features but are not official release jars. Snapshots have 
usually passed functional tests on a few platforms.
  ... 10.1.2.4  is a snapshot in which we acknowledge is not production 
ready (but generally of very good quality).
  >> Derby 10.1.2.4 (May 9, 2006 / SVN 405564)




Mime
View raw message