dolfin team mailing list archive

["Garth N. Wells" <g.n.wells@xxxxxxxxxx>]

On Fri, 2006-03-31 at 06:16 -0600, Robert C Kirby wrote: 
> >
> > Small correction:
> > Blitz++: 1.7
> > ublas:   1.95
> >
> > ublas (http://www.boost.org/libs/numeric/ublas/doc/index.htm) looks
> > really nice and provides everything I would like to have for a dense
> > matrix. It's a template library, we don't have to worry about linking 
> > to
> > it. If we use it, we could avoid a dependency by including the header
> > files in DOLFIN somewhere, something like dolfin/contrib?
> >
> 
> Careful again -- blitz is templated but there is a small static library 
> you do have to link against.  Be careful about how many dependencies 
> there are hidden underneath uBlas (especially as it's part of a bigger 
> project).  This is not a statement of knowledge but suspicion.  I'm 
> happen to be proven wrong.

I downloaded boost, pulled the header files out and dropped them into my
directory and it worked fine, so it appears that no binary libraries are
involved. I did a grep on the header files, and the dependency on other
boost components is limited.

> Also, templates can create portability/compiler problems if used too 
> creatively (see blitz).
> I have not worked with uBlas, but suggest compiling with a couple of 
> different versions of g++ on mac/linux/cygwin and also the intel 
> compiler on some platforms before committing.
> 

Just checked it with the Linux Intel compiler, and indeed Blitz gives
error messages. uBlas worked great. Works fine under Cygwin too.

Garth

> That said, I'm very much in favor of doing less code instead of more, 
> but there can be corner cases in portability and other headaches.
> 
> Rob
> 
> 
> _______________________________________________
> DOLFIN-dev mailing list
> DOLFIN-dev@xxxxxxxxxx
> http://www.fenics.org/cgi-bin/mailman/listinfo/dolfin-dev