there are pro's and con's to using a clustered or non clustered index for a
primary key.... the one that will perform better basically depends on your
specific scenario

however, there is a limit in SQL server which only allows one clustered
index per table... if you need a clustered index on anything other than the
primary key, you've gotta deal with a lot of pain to make that change.
That's why i think having the default set to non clustered is better than
defaulting to the only clustered index you're allowed to use per table.

On Sat, Mar 14, 2009 at 3:27 PM, Ayende Rahien <[email protected]> wrote:

> How it is an improvement?I _want_ my PK to be clustered. I make query
> assumptions based on that.
>
>
> On Sat, Mar 14, 2009 at 4:19 PM, Davy Brion <[email protected]> wrote:
>
>> but what if the change is an _improvement_ ?
>>
>> if you feel it should've been discussed on the list first, then fine,
>> you're right
>>
>> by the same token, reverting the change would've warranted discussion as
>> well
>>
>>
>> On Sat, Mar 14, 2009 at 3:16 PM, Dario Quintana <
>> [email protected]> wrote:
>>
>>> It's a breaking change becuase by default NH is generating a different
>>> code that the users were use to it, and change the behavior of the system.
>>> The change could be small, but still being a breaking change.
>>>
>>> On Sat, Mar 14, 2009 at 12:11 PM, Davy Brion <[email protected]>wrote:
>>>
>>>> how exactly is that a breaking change?
>>>
>>>
>>>
>>>
>>> --
>>> Dario Quintana
>>> http://darioquintana.com.ar
>>>
>>
>>
>

Reply via email to