>Yes yes. I've contributed code and more to the latest release (and >previous ones). I was just very amused about the "Guys, it works!" >statement. As if QuantLib would be a dubious medicine
It is not a dubious medicine, it is just a gun, heavy enough to shoot your own foot
I think no one will doubt the power of QuantLib. But the usability...
I talked to many C++ Quant-Developers, from novices to experts with 10+ years experience, none of them finds QuantLib user-friendly.
Some companies tried it but dropped.
As I, myself, started with QuantLib, I could relatively easy set up boost via the boostpro's installer. In such a way QuantLib can be compiled but the unit-test-suite not!
(Later I have found howto setup the boost via bjam, so that everything works).
Another problem: before the version 1.1 there was no straightforward way to use QuantLib with VS2010. A naive way just to convert vs9-files didn't work.
One more remark: an excellent tutorial by Dimitri Reiswich has become just recently available.
Implementing Quantlib by Luigi Ballabio, which was longer time there, does not bring much (well, ihmo).
So just recently one had to start without a good starting point.
And the last but not the least: VS has a powerful debugger but the data structures in QuantLib are very debugging-unfriendly.
Even tricking with autoexp.dat (s. https://svn.boost.org/svn/boost/sandbox/boost_docs/subprojects/DebuggerVisualizers/) does not help much.
And even such a simple class as Date cannot be overridden in autoexp.dat to dd:mm:yyyy format!
I think now my excitement about "it works!" becomes clear
P.S.
I know that one shouldn't look a gift horse in the mouth