From: Edward Diener (eldiener_at_[hidden])
Date: 2006-10-05 09:40:59
Jeff Garland wrote:
> Mateusz Loskot wrote:
>> Steve Hutton wrote:
>>> In summary I think that our 2.2 release with iterator support will be
>>> very close to what you have sketched out here, with the exception of
>>> boost::optional integraton, which I think we can address in a
>>> boostified version of SOCI, post 2.2.
>> Yes, I agree it's a very interesting idea, the boost::optional.
>> I second it to post 2.2.
> While I agree consideration of null handling using optional makes sense, there
> has been some controversy in the past about the optional interface...so be
> aware that this approach may bring criticism from some people.
What better way is there of specifying a database null value being
passed to or returned from a column of a particular row other than to
use boost::optional is the key question. I believe the criticism of
optional in the past regards using it with optional references, but when
passing and returning database null values, only values would normally
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk