Hi all
Ilja Chagalov wrote:
>I get the following exception:
>"QL exception: stub date (November 30th, 2002) is holiday and end of month
>for TARGET calendar"
>This exception is generated by the Scheduler.
>What is the reason for restrictiing the choice of the stub date?
and off-list he added:
>Why is the stub date rolled with the default rolling convention (which is
>"Following")
>
>Line 85:        ... calendar_.roll(stubDate_) ...
>
>and all other dates with the rolling convention given bei the user:
>
>i.e. Line 76: calendar_.roll(endDate_,rollingConvention_)
>       Line 94: temp = calendar_.roll(temp,rollingConvention_);
I think this is just a legacy, isn't it Luigi? When we were in 
CabotoIntesaBCI we were told
you cannot have a stub date that is holiday and end of month.
I don't remember the rational behind, maybe I never got it, but it was 
related to the fact that we were used to roll the stub date with the 
"Following" convention in any case.
It took me a while to "answer" this question since I planned to take a look 
at the FpML specifications about date/schedule generation. Unfortunately I 
didn't have enough time and I won't have that much free time in the next weeks.
Anyone able to help here? Any FpML expert?
FYI the QuantLib scheduler files are scheduler.hpp 
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/quantlib/QuantLib/ql/scheduler.hpp?rev=HEAD&content-type=text/vnd.viewcvs-markup 
and scheduler.cpp 
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/quantlib/QuantLib/ql/scheduler.cpp?rev=HEAD&content-type=text/vnd.viewcvs-markupany help _appreciated_
ciao -- Nando