Date: 2000-07-31 08:25:08
>I don't see how this changes it's usefulness, we're just talking about
>mangling the names of data which should really be private, it may protect
>people from using private data - but anyone who reads the header to fugure
>out what that data is should realise that it is supposed to be private?
This is true, but I have caught people simply using source auto-completion to
find member names, without looking at the source. Surely a method that makes
access to the member harder to do, making it closer to the private member that
it should be, is a good thing?
If people are not happy with this method, would it be acceptable instead to
name the members explicitly, to highlight their private-ness and make them easier
to search for, to see if anyone has used them illegally? For example: pRaw_PRIVATE
or something like that? No tricks here, just a bit of auto-documenting code.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk