Re: [boost] [thread] is this reverse_lock class a source of errors?

From: Johan Torp (johan.torp_at_[hidden])
Date: 2008-05-11 17:23:17

Phil Endecott-48 wrote:
> 
> I wrote something like this to use inside my condition variable 
> implementation and called it "unlock".  I think I prefer the name 
> "unlock" to "reverse_lock", personally.  
> 

unlock_guard would be even better, IMO.

I would love if all "call pairs" in boost had supplied RAII wrappers. When I
write my own code I usually force client code to use RAII objects. For
instance, I've written my own async signal and slots implementation. It
consists of two classes AsyncSignal<> and AsyncSignalConnection. These
classes are supplied together and the _only_ way to listen to an AsyncSignal
is to create an AsynSignalConnection. If the user really needs to call just
disconnect earlier on some runtime condition, they can use scoped_ptr+reset.

AsyncSignal<void()> sig;
boost::scoped_ptr<AsyncSignalConnection> con(new AsyncSignalConnection(sig,
foo_call, ...));

if (...) con.reset(); // disconnect

Best Regards, Johan
-- 
View this message in context: http://www.nabble.com/-thread--is-this--reverse_lock-class-a-source-of-errors--tp17145123p17178157.html
Sent from the Boost - Dev mailing list archive at Nabble.com.