[
Lists Home |
Date Index |
Thread Index
]
> Hah! I didn't know that, but I suspected that it couldn't actually be
> useful. Thank you. -Tim
Well, one word in favor.
David points out that the implementation cannot stop the implementor
from doing a cast and breaking the guarantee, so he doesn't like const.
On the other hand, he expects the implementor to do the right thing
without for putatively-const values.
I don't get it. Why not at least set things up so that (a) those
who get their documentation from the header file know what's going
on; and (b) those who want to break the rules will at least be forced
to "admit" (via const-cast) that they are doing so.
But hey, do what you want. Just know that I think David's opinion
here is the most incorrect thing I've heard him say. :)
/r$
--
Rich Salz Chief Security Architect
DataPower Technology http://www.datapower.com
XS40 XML Security Gateway http://www.datapower.com/products/xs40.html
XML Security Overview http://www.datapower.com/xmldev/xmlsecurity.html
|
