[sakai2-tcc] Kernel 1.0 merges
David Horwitz
david.horwitz at uct.ac.za
Tue Jul 20 05:02:25 PDT 2010
See bellow
On 07/20/2010 01:05 PM, Seth Theriault wrote:
> Folks,
>
> Please take this discussion to the sakai2-tcc list.
>
> Seth
>
> Jean-Francois Leveque wrote:
>
>
>> Looks like some of the merges are also wanted in Kernel 1.1.
>>
>> It seems there are unresolved issues in the filter:
>> KNL-528
>> KNL-527
>> KNL-448
>> KNL-414
>>
>>
Yes I was focusing on 1.0 right now as that is pressing and the filter
lists all jira's marked "merge"
D
>> Matthew Jones a écrit :
>>
>>> KNL-524: This is a small change but is recent and changes behaviour (maybe
>>> for the better)
>>> MJ: Probably some other ways to rewrite this, but the fix as-is
>>> looks okay to me.
>>> +1 merge
>>>
>> JFL: I see no answer to Savitha's question and Alan's suggestion.
>> +1 merge unless the current solution isn't approved by someone
>>
>>
>>> KNL-512: Recent change. IMHO needs testing
>>> MJ: Agreed, would probably want to see this tested outside the unit
>>> test.
>>> +1 pending tests
>>>
>> JFL: Agree with MJ
>> +1 pending tests
>>
>>
>>> KNL-503: Security issue, needs testing and discussion as potential change of
>>> behaviour
>>> MJ: I believe if the default is set to be as it currently is, with
>>> documentation describing how to enable it, then this is good.
>>> +1 merge with documentation
>>>
>> JFL: Agree with MJ
>> +1 merge with documentation
>>
>>
>>> KNL-496: Security issue - doesn't merge cleanly
>>> MJ: It's possible that this fix 'as-is' might require getting in
>>> the KNL-341 changes, but there's also an api change. To checkValue.
>>> -1 merge as is because of the api change and dependencies, seems
>>> like a useful fix though.
>>>
>> JFL: Looking for Kernel Merge Guru input
>> -1 until things are more obvious for me
>>
>>
>>> KNL-477: conflict merging to 1.0 - only affect HSQL, is this realy critical?
>>> MJ: HSQL is mostly for testing anyone. In addtion I noticed hsql
>>> 2.0 was released in June that supports full transaction. This should make a
>>> lot of the tests we commented out pass if we upgraded. Could be worth an
>>> investigation if we can get to a newer version easily.
>>> Anyway are you asking of HSQL is really critical or the fix? The
>>> fix sounds important. the HSQL part less so if it's a big problem.
>>> +1 merge
>>>
>> JFL: Don't think 2.6 HSQL is so important now that 2.7 is out. Because
>> nightly 2.6.x is using HSQL, might still be worth it.
>> +0 merge
>>
>>
>>> KNL-455: security issue involves changing the User API: needs discussion
>>> MJ: We typically have said changes to the API are not allowed, but
>>> additions to the api are different than changes. Often these can allow for
>>> backward compatibility and tool implementers to do less work. I'd be more
>>> inclined to allow an api addition, especially for security, but still this
>>> seems like it would require some discussion.
>>> +1 pending further discussion
>>>
>> JFL: Logging unallowed tries will be a good addition, AFAICT.
>> +1 pending further discussion
>>
>>
>>> KNL-380: recent change needs testing
>>> MJ: Visually this looks like a good fix which is bug that should be
>>> merged back. The unit test also validates this. +1 merge
>>>
>> JFL: Agree with MJ
>> +1 merge
>>
>>
>>> KNL-358: needs review
>>> MJ: Don't know enough about this change. +0 merge
>>>
>> JFL: Agree with MJ
>> +0 merge
>>
>>
>>> KNL-341: this looks like a new feature to me
>>> MJ: This is a new feature primarily created so that the flash movie
>>> (and other) plugins could be inserted into FCK SAK-15523. Since
>>> this plugin isn't being backported. I would consider this something not
>>> needed to be merged for 2.6, and it was previously reverted. FYI: We have
>>> been running this change since it was released at Michigan as we need the
>>> movie plugin
>>> -1 merge, as it is mostly to support new features
>>>
>> JFL: Anthony commented a revert with "This fix is not intended for
>> kernel 1.0.x (Sakai 2.6). "
>> -1 merge and remove merge request
>>
>>
>>
>>> KNL-213: implications beyond the kernel - changes behaviour therefore needs
>>> discussion.
>>> MJ: Right, looks involved with SAK-17460 and SAK-17462. Doesn't
>>> look like an API change. Would have to be a lot of QA around this change.
>>> I'd saw get a plan to test all of this before merging it.
>>> +0 merge, seems like more of a is QA effort available to guarantee
>>> nothing else gets broken worse.
>>>
>> JFL: Agree with MJ
>> +0 merge
>>
>>
>>> On Sat, Jul 17, 2010 at 12:05 PM, David Horwitz <david.horwitz at uct.ac.za
>>> <mailto:david.horwitz at uct.ac.za>> wrote:
>>>
>>> Hi All
>>>
>>> I have gone through the kernel issues marked for merging to the 1.0
>>> kernel and merged a great many of them. There are now 10
>>> resolved/closed issues in that queue. That remain unmerged:
>>>
>>> http://jira.sakaiproject.org/secure/IssueNavigator.jspa?reset=true&&pid=10410&customfield_10250=Merge&sorter/field=issuekey&sorter/order=DESC
>>> <http://jira.sakaiproject.org/secure/IssueNavigator.jspa?reset=true&&pid=10410&customfield_10250=Merge&sorter/field=issuekey&sorter/order=DESC>
>>>
>>>
>>> They are:
>>>
>>> KNL-524: This is a small change but is recent and changes behaviour
>>> (maybe for the better)
>>>
>>> KNL-512: Recent change. IMHO needs testing
>>>
>>> KNL-503: Security issue, needs testing and discussion as potential
>>> change of behaviour
>>>
>>> KNL-496: Security issue - doesn't merge cleanly
>>>
>>>
>>> KNL-477: conflict merging to 1.0 - only affect HSQL, is this realy
>>> critical?
>>>
>>> KNL-455: security issue involves changing the User API: needs discussion
>>>
>>> KNL-380: recent change needs testing
>>>
>>> KNL-358: needs review
>>>
>>> KNL-341: this looks like a new feature to me
>>>
>>> KNL-213: implications beyond the kernel - changes behaviour
>>> therefore needs discussion.
>>>
>>>
>>> Outside of these there is also the issue of upgrading the libraries
>>> in the kernel (commons-lang, javax-mail etc). This are in 1.1 and
>>> at UCT we have been running them in a 1.0 based kernel for 6
>>> months. Last time this was discussed on list
>>> (http://thread.gmane.org/gmane.comp.cms.sakai.devel/36574/focus=36589)
>>> the Voting was:
>>>
>>> +1 David H, Nuno, Steve S, David R-M. Jean-Francois
>>> -1 Matthew B
>>>
>>> Should this be put to a new vote?
>>>
>>> D
>>>
>> From a TCC PoV, I think so.
>>
>> J-F
>>
>>
More information about the sakai2-tcc
mailing list