On 05/13/2011 03:10 PM, Vincent van Ravesteijn wrote:
>>
>>> 0002 seems correct.
> If you really want comments:
>
> It seems you've done something strange to the whitespacing.
>
Indeed. I'll fix that.
> I would prefer to rename 'tmp' to 'params_copy'. Then it is
> clearer what it is.
>
Of course.
> I'm not sure about this, but the following does seem like it follows
> from some good principle: If code can go into BufferParams instead of
> Buffer, it should.
Hmm.. I think there should be more to it.
I've been thinking about the Buffer class before. I think it should
only care about manag
On 05/13/2011 03:24 AM, Abdelrazak Younes wrote:
> On 05/12/2011 11:36 PM, Richard Heck wrote:
>> Before I commit these---especially to branch---I thought I'd post them
>> for comment.
>>
>> The 0002 patch shows the real point of this: We do not need to clone the
>> whole Buffer here, but only need
On 05/12/2011 11:36 PM, Richard Heck wrote:
Before I commit these---especially to branch---I thought I'd post them
for comment.
The 0002 patch shows the real point of this: We do not need to clone the
whole Buffer here, but only need access to a copy of the BufferParams.
The 0001 patch does wha
Before I commit these---especially to branch---I thought I'd post them
for comment.
The 0002 patch shows the real point of this: We do not need to clone the
whole Buffer here, but only need access to a copy of the BufferParams.
The 0001 patch does what's necessary to make that happen, and then a
