RE: [boost] Re: Reference documentation: one approach

From: Jeff Garland (jeff_at_[hidden])
Date: 2002-11-01 22:36:12

> I've got to admit, it appears to have a reasonable grasp of what's
> happening in your code. I'm not fond of the way things are presented,
> but then I've got a lot of complaints with Synopsis in that area as
> well.

Well I agree about the presentation, but there are only so
many hours in a day.  And it is impossible to be get
a form that all will like.  However, I think that you have 
your eye on the big issue here.  I would like to be able 
to generate all the methods for gregorian::date as if it 
were a simple concrete type -- not some thing recursively 
derived from a bunch of template classes.  Most user's
don't need to know all that stuff, they just want to know 
what operators and methods the class supports.  Another 
example would be looking at one of Andrei's Abstract
Factories configured with a typelist.  The inheritance
hierarchy doesn't really capture the essence of the 
design -- just the mechanism of implementation.  I don't
really know much about boost::python, but I'm guessing it
has these same sort of issues with generated docs...

> It would be interesting to throw Doxygen at Boost.Python and Synopsis
> at Boost.DateTime for comparison.

That would be fun...

> > Let me know when I can download an RPM for Linux and an exe
> > installer for windows.
> 
> Hey! No fair playing the installation card! ;-)

Whew, finally got one point :-)

Jeff