(no subject)

[Marco Marchioro-2]

Dear QuantLib user,
it is with great pleasure that I will report the conclusions of the First
QuantLib Conference in Milan, held on October 20th, 2002. Well, that's wishful
thinking, may be not a conference per se, let's say the First QuantLib
Afternoon. However, I still hope that there will be a real QuantLib conference
in the future.

There were participants from the major European countries: myself, Luigi, and
Nando from Italy, Sad from Switzerland, and Jens from Germany.
The meeting started with a visit to the facilities of RiskMap, exactly in the
place where QuantLib took shape a little more than two years ago. Shortly
after
the CEO of RiskMap, Dario Cinitioli, welcomed everybody, coffee and water were
served. Moments later the discussion begun in a single session with all the
participants very active in the debate. Here's a list of all the major
conclusions.

+Changes to the Web Site
   * Improve documentation, maybe add a weblog where QL developers can
     leave short progress notes
   * Update the "to do" list
   * Draw a road map of future works

+Importance of Unit Testing
   The idea is to certify the quality of QuantLib using a unit-test framework.
   The tests would also serve as concrete examples on how the library should
   be used.
   QuantLib would be defined by what is tested, all the major features should
   be tested. In this way it will also be possible to maintain the
synchronization
   of QuantLib with QuantLib.NET.
   All the tests done in python will be copied to corresponding tests
written in C++.

+Changes to the Library
   * Improve the documentation
   * Move to the Instrument/pricing engine framework
   * Refactor of the Finite-Difference framework merging with
     the Tree/Lattice framework
   * Develop a framework for stochastic processes
   * Build basic Bond classes

In particular the last point, build basic Bond classes, needs volunteers.

Afterwards, the meeting moved to a more mundane venue where pizza was
eventually served.
Distinguishing himself for his class, the representative of the Swiss QuantLib
community asked for a Martini Bianco to drink with his pizza. The other
developers
enjoyed their pizza with a beer.
Later in the evening the chief architect of QuantLib astonished everybody
winning the competition of eating the biggest(they say size doesn't matter)
slice of cake.

The evening ended late at night after many more discussion about life, the
universe, everything and, of course, QuantLib.

Marco Marchioro, Milan, October 21st, 2002.

[Luigi Ballabio-2]

At 10:16 AM 10/25/02 +0100, Marco Marchioro wrote:
>The evening ended late at night after many more discussion about life, the
>universe, everything and, of course, QuantLib.

Apparenty, the result was 42.
But as of today, it's not in the test suite.

Bye,
         Luigi

[Michele Ravani-2]

Hi

improving the documentation is a great idea.
IMHO, it would be extremely beneficial if it could be aimed also at
beginners in the quant field.
Particularly, it should simplify the learning and mastering of QL and
related themes. I do realize that there is a lot of literature about, but I
think that with QL one could provide a kind of practical guide to the
various topics.

Obviously, the experts should also find all they need, explained in
sufficient. This could be quite a challenge.

Did you already think what to use to write such a 'manual'?
Stick to doxygen (wouldn't math formulas be a problem?) or use lynx (html
for the website could be provided from the latex output)?

I would love to get involved in the writing, as it would get me into the
topic (a friend said 'if you want to learn something, give a course about
it') under the direction of the experts (you).
I guess I would be able to dedicate some time to the writing (not right
now, I am moving house) and being a beginner myself, I may be able to get
the right 'tone' for others in my situation.

Any thoughts?

--
Michele Ravani                  [hidden email]
"Those who live hoping, die singing" My Gran