RFR[L]: 8237767 Field layout computation overhaul

[Daniel D. Daugherty]

On 1/29/20 8:44 AM, coleen.phillimore at oracle.com wrote:
> On 1/29/20 8:31 AM, David Holmes wrote:
>> PS. I've now seen the CSR and have commented there. I don't agree 
>> with the use of a diagnostic flag. It doesn't buy us anything other 
>> than not needing a CSR request when we decide to deprecate it. But if 
>> we're going to tell people to use these flags to revert behaviour 
>> then they will need to follow the deprecation process regardless - 
>> and a product flag is better from an end user perspective.
>
> Obviously I disagree.  We don't want to publicise this flag *unless* 
> someone has a problem, which makes it a diagnostic flag.   We don't 
> want to have to carry the old code 2 releases when there is honestly 
> no reason to suspect it is problematic.
>
> Coleen

I'm only chiming in on this one snippet from the review.

I'm obviously a big fan of diagnostic flags instead of product
flags. Based on what I've read about this new flag, I would vote
that it should be a diagnostic flag and not a product flag. To
be a product flag, you should have to justify the expense of
keeping/maintaining the alternate code for the two releases that
it would take to remove it.

My $0.02...

Dan