Dear Berserker, I was afraid that by uncatching the exceptions in the caller, moreover if in a multithreaded application, you're going to have for sure an abort, if not some other side effect in the worst case... (I told you that it was the stupid one, if you're not a newbie!). Btw, with boost 1.35.0 and 1.36.0, same stuff but with async methods works more than perfectly, both if you use client auth or not. Ciao, Carlo Medas On Fri, Sep 5, 2008 at 5:47 PM, Berserker <berserker_r@hotmail.com> wrote:

Thanks for the reply Carlo :)

...

* when you close and connect again, we use 'lowest_layer()' where you use 'next_layer()'

stream::lowest_layer() is only a wrapper around "next_layer_.lowest_layer()" where, in case of a "basic_socket" as template argument, is:

lowest_layer_type& basic_socket::lowest_layer() { return *this; }

So there is no difference with:

next_layer_type& stream::next_layer() { return next_layer_; }

...

* stupid one (you never know...): are you catching exceptions on the caller? :)

Sure but I don't think that a stack overflow is a "catchable exception" don't you think :) ?

_________________________________________________________________ Cerchi un locale per la serata? Chiedilo a Live Maps! http://maps.live.it/ _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost