Andrew Borley wrote:
Hi,
I've updated the BigBank sample today to get it working on windows
based on the new directory and deployment structure. One thing I had
to do to get it all working was change all the component interfaces
from using char* strings to std::string, as the changes to the
binding.ws service and reference stuff that went in for r449433 means
that using char* strings causes a crash.
Was this the desired effect of the change? Do we want users to only
use std::string in their component interfaces? Is there a way to
accomodate both std::string and char*?
Cheers
Andy
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
I guess r449433 introduced that bug, that was not the intent. I didn't
see the bug on Linux but that may have been an oversight. I'm trying to
wrap-up the changes to remove the requirement for the "packages/" and
"configuration/" folders (and it's taking a little longer because my
initial stab at this didn't go quite right). I'll take a look just after
that.
Independent of this bug, I like how you changed Bigbank to use
std::strings (more convenient than char* to work with) so what do you
guys think about keeping the std::strings in Bigbank even after the bug
is fixed?
--
Jean-Sebastien
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
