From: Peter Dimov (pdimov_at_[hidden])
Date: 2002-02-22 08:04:03
From: "Herb Sutter" <hsutter_at_[hidden]>
> > But it's off topic to have a general "make suggestions for the new
> > language" discussion here.
> Indeed, and thanks to you, Dave, and Beman for repeating this. My question
> for a list of features "most needed" for writing C++ "libraries ... and
> in particular."
> So far, there have been three such suspects discussed: typedef templates,
> auto, and typeof.
I have to disagree with this. 'auto' is a nice feature that definitely gets
publicity but it's _not_ needed for writing libraries (except for the return
type autodeduction which is solved by typeof.)
In contrast, the 'argument forwarding' problem definitely needs a solution.
As has been repeatedly pointed out, currently there is no practical way to
make a function "forward" to another function its arguments unmodified (i.e.
f(a1, ..., aN) to be equivalent to g(a1, ..., aN).)
In particular, Boost.Bind and Lambda suffer a lot from this C++ limitation.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk