[Laszlo-dev] potential merges from legals to 4.0.3: ptw

P T Withington ptw at openlaszlo.org
Mon Jun 25 12:39:12 PDT 2007


On 2007-06-25, at 14:18 EDT, Jim Grandy wrote:

>
> On Jun 25, 2007, at 11:13 AM, P T Withington wrote:
>
>> On 2007-06-25, at 14:03 EDT, Jim Grandy wrote:
>>
>>>
>>> On Jun 22, 2007, at 7:59 PM, P T Withington wrote:
>>>
>>>>>     r5226 | ptw | 2007-05-25 14:36:19 -0700 (Fri, 25 May 2007)  
>>>>> | 130 lines
>>>>>     Summary: Step 1 of integrating binary libraries from trunk  
>>>>> to legal's
>>>>
>>>> Propose taking
>>>> Risk: high -- this is a complex change and has had little use  
>>>> testing
>>>> Reward: high -- needed by laszlo, interest from community
>>>> Note: Must also take 5423, 5252
>>>
>>> If we can avoid it, I don't think we should take this. It's a  
>>> major feature, and we try to keep these dot-dot releases as  
>>> feature-free as possible to reduce risk and QA overhead. Remember  
>>> the model: full QA pass on the dot release (e.g. 3.4.0 or 4.0.0),  
>>> then less QA scrutiny for the subsequent minor releases (e.g.  
>>> 3.4.1, 4.0.1).
>>
>> Understood, although I was hoping for some community QA.  Perhaps  
>> the community would like to weigh in on whether they would like  
>> this feature in 4.0.3 or not?
>
> I'd rather do the community QA as part of the 4.1.0 pre-release  
> cycle. If we take this change we either pay for a full QA pass or  
> we lose the ability to claim that 4.0.x is about shoring up  
> stability. Perhaps we should be planning for a 4.1.0 beta release  
> soon?

My fear is that because of the size of this change, it could be risky  
to not take it.  If subsequent changes intersected with it at all,  
they will have to be hand merged (or re-created by the original  
developer and applied to the 4.0.x line.  Perhaps the only way to  
tell is for Ben to try making a merge with these 3 changes blocked  
and see what happens.

As to a 4.1 beta, I don't know what features were targeted for that  
other than binary libraries and... stability.  Are there others that  
we need to wait for?  And QA resources are a gating factor.




More information about the Laszlo-dev mailing list