Re: [boost] Re: Re: Sockets!

From: Marshall Clow (marshall_at_[hidden])
Date: 2003-10-31 21:33:14

At 3:54 PM -0500 10/30/03, Beman Dawes wrote:
>At 09:00 AM 10/30/2003, Peter Dimov wrote:
>
>>Bronek Kozicki wrote:
>>>  Wynand Winterbach <wynand_at_[hidden]> wrote:
>>>
>>>  why don't we just make *small* hierarchy of exceptions, all derived
>>>  from "socket_error" derived from "std::runtime_error" ?
>>
>>What does the user gain from this?
>
>Users who care about the exact reason can catch the more detailed 
>types, uses who just care that the exception is sockets related can 
>catch socket_error?

I've been using Nitrogen <http://sourceforge.net/projects/nitric/> 
lately, which has
a large exception hierarchy (mapping MacOS error numbers --> classes), and
it seems to work quite well.

You can either (a) catch the base class and recover the error from that, or
(b) catch individual errors seperately.
-- 
-- Marshall
Marshall Clow     Idio Software   <mailto:marshall_at_[hidden]>
Hey! Who messed with my anti-paranoia shot?