Re: feasibility of numerical lib on the top of glib

[Byron Alley <liondios uvic ca>]

> The only disadvantage I see is exactly what you stated -- the
> dependence on glib. Some might not like it, saying "well, if I
> want to use your, say, matrix multiplication, I'd have to go
> and install glib." For me, personally, this is not an argument but it
> might be for someone else. I think that there is no need to
> reinvent the wheel, and if there is a good code already that can do
> much of what you need -- use it.

What about offering statically linked binaries for those people?