Boost logo

Boost :

From: Jeremy Siek (jsiek_at_[hidden])
Date: 2000-10-31 13:21:45

I agree that dependencies tables/diagrams are useful.

It was helpful to see the complete list of directly and indirectly
included headers for a particular file, however it is also important to
see the whole graph... not just the "flattened" version. Would
it be possible to output the header dependencies in some sort
of hierarchical view that shows this information? Perhaps
just run BFS over the graph and output the resulting tree
like this:


Just add an indenting level for the distance the node is from
the root...



 Jeremy Siek www:
 Ph.D. Candidate email: jsiek_at_[hidden]
 Univ. of Notre Dame work phone: (219) 631-3906

On Tue, 31 Oct 2000, Beman Dawes wrote:

> At the Boost meeting in Toronto, Nico Josuttis asked that we publish header
> dependencies. His motivation was wanting to know what other headers have
> to be available to use one of the Boost headers. I thought it would also
> be useful as a way of judging the degree of coupling.
> I've written a little program to which turns the GCC -M dependency output
> into an HTML table. See
> I originally choose a table rather than a Doxygen style chart because it
> was more compact. But also look at
> (thanks
> to Jeff Garland). Doxygen isn't picking up the indirect includes, and so
> gives a less useful view, IMO. I don't know if it has an indirect include
> option.
> Please take a look at my HTML table in the files/vault, and then comment:
> * Is the table useful and worth including on the boost web site?
> * Should implementation detail headers be included or excluded?
> * Should the 2nd column entries be sorted?
> I'm not including Standard Library headers; exactly which ones get included
> is very Standard Library implementation dependent.
> It might be nice to include Standard Library headers called directly from
> Boost headers, but I don't offhand see an easy way to do that.
> --Beman

Boost list run by bdawes at, gregod at, cpdaniel at, john at