Date: 2001-06-24 13:28:35
> > I am certainly coming to that conclusion. When we established a coding
> > standard at my job recently, we chose the leading-underscore convention,
> > in light of recent boost messages I am strongly advising a switch to "m_".
> > BTW, I am beginning to wonder if it's worth having a "suggested coding
> > standards" document. I have put considerable effort into the
> > boost-compatible coding standards we're using at work. Clearly it would be
> > overkill for boost to mandate some of the nitpicky things in our standard,
> > but it sometimes helps to have some guidelines when you're trying to
> > whether to write something one way or another.
> > -Dave
> I think a suggesting coding standards document is a great idea. Obviously,
> trying to direct a group like this is like herding cats, but if there is a
> style it would at least allow people that *want* to write code that is
> compatible to do so.
I believe that an accepted set of coding guidelines would be a huge 'boost' to
development teams around the globe. This comes up on literally every project,
and most companies don't want to bother so they end up saying something like:
use Hungarian notation (gagging heard in background) and follow the guidelines
in Effective C++ (cheers from the audience). Ideally, these guidelines would
be written like Scott's books: small concrete sections with short titles and
deep explanations of why something is a guideline (and hence when it can be
violated). They should be highly modular so that projects can make
customizations and removals.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk