|
Boost : |
From: Rani Sharoni (rani_sharoni_at_[hidden])
Date: 2004-01-21 12:05:05
David Abrahams wrote:
> Samuel Krempp <krempp_at_[hidden]> writes:
>
>> On Sat, 2004-01-03 at 02:49, David Abrahams wrote:
>>>>>> I can't really understand the invariants of the basic_format
>>>>>> class...
>
> That wasn't me.
I'm the one to blame but mainly because you have significant influence on my
exceptional mindset ;-)
Anyway, I agree that such function should supply the basic guaranty since in
such cases it's not justified to have such expensive penalty (e.g. vector
copying) in the ordinary non-exceptional code path. Nevertheless maintaining
the invariants in the exceptional code path might be tricky. I'm still
looking for some guideline that simplifies the effort to achieve correctness
when performance is in mind (i.e. basic guarantee) but, as my friend told
me, this is (also) what makes our job difficult.
Rani
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk