|
Boost : |
From: John Max Skaller (skaller_at_[hidden])
Date: 2001-08-20 16:48:48
Greg Colvin wrote:
>
> Right. And though I highly respect Max's devotion to consistency,
> I also dread the politics of introducing yet another version of
> auto_ptr with yet another name.
Don't. Introduce 'array<T>' and 'fixed_array<T,n>'
instead. They're both needed anyhow. If I remember rightly,
Bjarne has said their omission from the C++98 is
an oversight.
> It just seems easier to introduce
> auto_ptr<T[n]> as a small delta to auto_ptr<T>, and it seems that
> shared_ptr<T[n]> and scoped_ptr<T[n]> would be needed to fill in
> the picture.
It always seems easier to hack and kludge than design.
And you always pay in the long run. Bjarne has 'officially'
advised that C arrays are broken. I agree. Don't try to
fix them, they can't be fixed. Just make sure that there's
something you can use in their place (an 'onstack'
or 'inobject' fixed_array is therefore mandatory).
Worrying about C arrays is a CWG problem :-)
> I'll also note that I dread the politics of re-introducing
> shared_ptr and scoped_ptr.
So don't. Let Beman do it. Check with the UK delegation.
There is an outanding requirement for these two classes
from Australia and UK. Australia is no longer an SC22 participant
(only an Observer), but the UK surely is. That puts the
two classes squarely on the agenda. You'll be surprised:
I think most people will vote for it because they're
sick of the outstanding issue (and their clients want it too).
> And finally, the Core group "broke" auto_ptr in Tokyo, so we also
> need a core language change to make it work as designed.
What happened?
> Fear and loathing in Redmond?
Seattle is far too beautiful, and there are lots
of great places to eat. Just make sure to take waterproof
clothing :-)
-- John (Max) Skaller, mailto:skaller_at_[hidden] 10/1 Toxteth Rd Glebe NSW 2037 Australia voice: 61-2-9660-0850 New generation programming language Felix http://felix.sourceforge.net Literate Programming tool Interscript http://Interscript.sourceforge.net
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk