[sakai2-tcc] Sakai 2.7.2 and Sakai 2.8.1
Anthony Whyte
arwhyte at umich.edu
Thu Nov 3 06:32:47 PDT 2011
I recommend when reviewing tickets to be wary of taking issue attributes at their face value. Look at the actual code commits as well.
The following tickets affects/fix versions are/were incorrect. In each case it looks like a branch manager failed to ensure that the affects/fix version were properly updated after performing a merge to the relevant *.x branch.
https://jira.sakaiproject.org/browse/SAK-21071
https://jira.sakaiproject.org/browse/SAK-21115 (related to SAK-21071)
https://jira.sakaiproject.org/browse/SAK-21015
https://jira.sakaiproject.org/browse/SAK-21078 (I updated the fix version last night)
https://jira.sakaiproject.org/browse/SAK-20913 (I updated the fix version last night)
This ticket is listed as open when the comments indicate it has been addressed by SAK-21015 et al.
https://jira.sakaiproject.org/browse/SAK-21109
There may be good reasons to recommend that a tag be rejected for public consumption but the rationale needs to be grounded in specifics rather than generalities and the specifics need to square with code commit history. It's not helpful that Jira ticket attributes can mislead but basing assertions on a cursory review of ticket attributes alone risks distorting the actual state of a tag.
The kernel issues and new chat issue filed by Chuck Hedrick are rather more troubling but all are currently open tickets that have yet to be resolved.
https://jira.sakaiproject.org/browse/KNL-433
https://jira.sakaiproject.org/browse/KNL-815
https://jira.sakaiproject.org/browse/KNL-797
https://jira.sakaiproject.org/browse/KNL-799 (add this as well)
https://jira.sakaiproject.org/browse/SAK-21353
Perhaps these are sufficient to recommend rejecting 2.8.1. Is there consensus on how to resolve the issues and is there an expectation that these tickets can be addressed and tested quickly? If not, introducing further delays by rejecting 2.8.1 may not be in the community interest.
As for the following Jiras noted by Seth, none should block 2.8.1 in my opinion since deployment to Tomcat by a overlay zip file in the case of the kernel, emailtemplateservice and Job Scheduler is the default (or standard) deployment strategy utilized by the kernel since 2.6 and the other tools since 2.7. While I recommend that the SAK fixes be back-ported to 2.8.x (KNL-818 has already been merged to kernel-1.2.x) I don't think they should be cited as a reason to invalidate 2.8.1.
spring-webmvc missing from Kernel deploy
https://jira.sakaiproject.org/browse/KNL-818
Create emailtemplateservice deploy pom in order to deploy the simple-xml jar to $CATALINA_HOME/shared-lib
https://jira.sakaiproject.org/browse/SAK-20570
Create job scheduler deploy pom in order to deploy the quartz jar to $CATALINA_HOME/shared-lib
https://jira.sakaiproject.org/browse/SAK-20572
I stand ready to generate a follow up a 2.8.2 release if 2.8.1 is judged deficient on reasonable criteria. The rationale at present appears problematic to me since it relies on a miscellany of tickets, the fixes for some of which are actually in 2.8.1, others require confirmation, still others are open with no estimate as to when they will be resolved, tested and verified, while others appear to me inadmissible as evidence to condemn a release.
But do not misconstrue what I write above. It takes only one serious regression in my view to sink a tag. If 2.8.1 is jact lets regroup and produce another one.
Cheers,
Anth
On Nov 3, 2011, at 12:31 AM, Steve Swinsburg wrote:
>
> For https://jira.sakaiproject.org/browse/SAK-21071 the Jira says 2.8.2, unless the version was set incorrectly.
>
> Also, this issue: https://jira.sakaiproject.org/browse/SAK-21015 has both an affects AND fix version of 2.8.1. This is impossible. If it was present in the 2.8.1 tag then it is fixed in 2.8.2.
>
> For the others that are fixed in 2.8.1 I simply listed those as serious issues that need to be put out there. We know there is a 2.8.1 release, but no one else does.
>
> In any case, I am of the opinion that we need a 2.8.2 release so as to at least alert people that there are issues that are fixed, and to upgrade. We missed the boat on 2.8.1 unfortunately.
>
> I will be focusing on our own 2.8 rollout in the next few days so I'll be merging anything that I can.
>
> cheers,
> s
>
>
>
>
> On 03/11/2011, at 3:19 PM, Anthony Whyte wrote:
>
>> Focusing on 2.8.1, of the 11 tickets you cite:
>>
>> Ticket resolved and fix included in 2.8.1 = 4
>> Tickets that can and should be addressed by off-cycle releases of emailtemplateservice and search = 3
>> Tickets that require confirmation that bug is actually resident in 2.8.1 = 3
>> Tickets not addressed by 2.8.1 = 1
>>
>> Cheers,
>>
>> Anth
>>
>>
>>
>> On Nov 2, 2011, at 11:11 PM, Steve Swinsburg wrote:
>>
>>> Here are some:
>>>
>>> Email fails if there is no en_US template even though a "no locale" template exists
>>> https://jira.sakaiproject.org/browse/SAK-21295
>>
>> This bug can be addressed by performing an off-cycle emailtemplateservice release. It does not require a 2.8.2 release to resolve.
>>
>>>
>>> Regression: delayed announcements are emailed immediately
>>> https://jira.sakaiproject.org/browse/SAK-21071
>>
>> All merges to 2.8.x included in 2.8.1. No fixes included in 2.7.x/2.7.2 due to merge conflicts as noted in the ticket.
>>
>>>
>>> 2.8.0 to 2.8.1 conversion loses mail messages
>>> https://jira.sakaiproject.org/browse/SAK-21305
>>
>> This issue needs to be confirmed for 2.8.1. Charles Hedrick notes: "This needs to be checked at some site other than Rutgers. I'm concerned that we may have done something after doing 2.8.0 that added the message_order="0"."
>>
>>>
>>> wrong default sort order for announcements
>>> https://jira.sakaiproject.org/browse/SAK-21308
>>
>> This issue needs to be confirmed for 2.8.1. Charles Hedrick notes: "I don't know how to duplicate a wrong sort order on the QA server, so I can't show any actual problem for this. But I note that initially no column is shown as selected for sort, whereas after posting an announcement, sort by date is selected. So I think the problem is present even though I can't demonstrate it."
>>
>>>
>>> 2.8.1 upgrade for announcements needs to renumber all announcements
>>> https://jira.sakaiproject.org/browse/SAK-21306
>>
>> This issue needs to be confirmed for 2.8.1. As Charles Hedrick notes: "message_order is really buggy in 2.8.0 (and in the 2.8.1 that we're using, but I think this is not true of the 2.8.1 tag)."
>>
>>> Copyright controls are not working
>>> https://jira.sakaiproject.org/browse/SAK-21178
>>
>> 2.8.x, r100255 (30 Oct 2011).
>>
>>>
>>> Regresion: exception on shared merge (This is in 2.8.1 so needs to be put out there)
>>> https://jira.sakaiproject.org/browse/SRCH-59
>>
>> If still an issue, this can be addressed by an off-cycle release of search. It does not require a 2.8.2 release to rectify.
>>
>>>
>>> "node already merging shared segments" on lone instance
>>> https://jira.sakaiproject.org/browse/SRCH-38
>>
>> If still an issue, this can be addressed by an off-cycle release of search. It does not require a 2.8.2 release to rectify.
>>
>>>
>>> Announcement reordering should be enabled by default
>>> https://jira.sakaiproject.org/browse/SAK-21109
>>
>> Fix in 2.8.1 tag. See SAK-21015.
>>
>>>
>>> Preferences -> Nofications descriptions missing for Announcements and Email Archive
>>> https://jira.sakaiproject.org/browse/SAK-21078
>>
>> Fix in 2.8.1 tag
>>
>>>
>>> Site Info: Manage Access interface allows all sites to be made joinable
>>> https://jira.sakaiproject.org/browse/SAK-20913
>>
>> Fix in 2.8.1 tag
>>
>>
>>>
>>> cheers,
>>> S
>>>
>>>
>>> On 03/11/2011, at 1:17 PM, Anthony Whyte wrote:
>>>
>>>> Can you list the tickets to which you allude to below.
>>>>
>>>> Cheers,
>>>>
>>>> Anth
>>>>
>>>>
>>>>
>>>> On Nov 2, 2011, at 6:13 PM, Steve Swinsburg wrote:
>>>>
>>>>> Maybe we should just push out 2.7.3 and 2.8.2 instead, then announce those.
>>>>>
>>>>> There have been a few major issues fixed, some of them serious regressions that are important to get out there.
>>>>>
>>>>> cheers,
>>>>> Steve
>>>>>
>>>>> On 03/11/2011, at 4:44 AM, Beth Kirschner wrote:
>>>>>
>>>>>> Hi everyone,
>>>>>>
>>>>>> This past September we released quietly 2.7.2 and this past October we quietly released 2.8.1, both with Anthony's help. It's been pointed out to me that we never made a formal announcement of either (feel free to correct me if I'm wrong), but I think that given the change in Anthony's responsibilities, it's likely that this was missed.
>>>>>>
>>>>>> I'd like to compose an email announcement that retroactively announces these two releases (wording to be vetted with this group), but first thought it worthwhile to put to a vote whether or not we're comfortable formally announcing these releases. Please let me know by the end of this week of any issues that you think should be addressed prior to a formal, retroactive announcement. Following this discussion, I'll put it to a vote.
>>>>>>
>>>>>> Thanks,
>>>>>> - Beth
>>>>>>
>>>>>> _______________________________________________
>>>>>> sakai2-tcc mailing list
>>>>>> sakai2-tcc at collab.sakaiproject.org
>>>>>> http://collab.sakaiproject.org/mailman/listinfo/sakai2-tcc
>>>>>
>>>>> _______________________________________________
>>>>> sakai2-tcc mailing list
>>>>> sakai2-tcc at collab.sakaiproject.org
>>>>> http://collab.sakaiproject.org/mailman/listinfo/sakai2-tcc
>>>>>
>>>>>
>>>>
>>>
>>>
>>>
>>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://collab.sakaiproject.org/pipermail/sakai2-tcc/attachments/20111103/cce6fc64/attachment-0001.html
More information about the sakai2-tcc
mailing list