Re: bzr commit into mysql-6.0 branch (Rafal.Somla:2734) Bug#40305

[Rafal Somla]

Hi,

An updated patch implementing the new solution is here 
http://lists.mysql.com/commits/61844. Please review.

Rafal

Chuck Bell wrote:
> Rafal,
> 
> I agree with you new stateless solution.
> 
> Chuck
> 
> Rafal Somla wrote:
>> Chuck Bell wrote:
>>> STATUS
>>> ------
>>> Patch not approved. Changes, explanations requested.
>>>
>>> REQUESTS
>>> --------
>>> 1. Fix various spelling errors.
>>>
>>
>> I will. My spell checker is crappy - sorry.
>>
>>> SUGGESTIONS
>>> -----------
>>> 2. I don't think we should add methods and attributes to the logger 
>>> that don't actually log anything or that store a state. I would 
>>> suggest moving the m_completed and m_data_changed elsewhere. I would 
>>> like to hear why you think recording of state belongs in the logger 
>>> class.
>>>
>>
>> This is a good point. I like the idea of making backup::Logger 
>> stateless (as much as possible). I have tried to go in that direction 
>> - see the attached patch and let me know if you think it is better.
>>
>>> 3. Are the new error messages generic?
>>>
>>
>> Yes, I think they can be useful in other contexts too.
>>
>> I'm waiting for your opinion on the new proposed patch and after that 
>> I'll commit a new one.
>>
>> Rafal
>>
>>
>> ------------------------------------------------------------------------
>>
>>
>