|
Boost : |
From: Jesse Jones (jesjones_at_[hidden])
Date: 2000-11-24 17:09:52
>At 03:42 PM 11/23/2000 -0800, Jesse Jones wrote:
>
> >>>6) Replaced NULL with 0 (Gregor says NULL was causing problems with
>GCC).
> >
> >>>This change stinks. :-)
> >>
> >>Couldn't disagree more :-> NULL is, depending on your sensibilities,
> >>either a C-ism or an MS-ism, but it's not generally considered part of
> >>the received C++ style.
> >
> >You're painting with an awfully broad brush here. :-) Plenty of people
> >think the way C overloads 0 impairs readibility.
>
>You may be right about overloading, but nonetheless Kevlin is correct that
>"NULL [is] not generally considered part of the received C++ style."
Readibility is, to me, one of the most important attributes of code and
there's no question in my mind that "p = NULL" is more readable than "p =
0". I don't care of every member of the comittee frowns on the practice.
I'll continue to use NULL outside of boost.
> >>Speaking of stylistic elements, is it possible to move the code towards
> >>more conventional C++ naming? Eg function objects instead of functors (a
> >>bit early 90s, that ;->) and member functions instead of methods? Not a
> >>big issue, but it affects my comfort level :-)
> >
> >Well the current names are just as clear as what you're proposing and
> >they happen to be shorter...
>
>"Methods" is only clear to a portion of the C++ community, and on top of
>that sounds pretty dated to my ear.
>
>Both of those issues are yesterday's battles. Those who happened to be on
>the losing side just need to move on, rather than fighting the same old
>fights over and over again.
This isn't a big deal to me.
-- Jesse
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk