From freedesktop-dbus  Tue Nov 13 17:01:22 2007
From: Havoc Pennington <hp () redhat ! com>
Date: Tue, 13 Nov 2007 17:01:22 +0000
To: freedesktop-dbus
Subject: Re: about the D-Bus C API.
Message-Id: <4739D862.20202 () redhat ! com>
X-MARC-Message: https://marc.info/?l=freedesktop-dbus&m=119497332915225

Hi,

Thiago Macieira wrote:
> Em Tuesday 13 November 2007 17:39:18 Havoc Pennington escreveu:
>> IIRC signals can set the destination, in which case they'd be sent to
>> that destination without a match rule.
> 
> This is actually something I was made aware of a month ago or so.
> 
> Signals can specify a destination. That seems wrong to me, if a signal is 
> supposed to be a broadcast.
> 

I think it's a reasonably common pattern to have "clients" that register 
with some "server," in that case having the server send client-specific 
signals seems OK to me; otherwise the client would end up registering a 
callback object with the server. The callback object pattern is fine 
too, but in some cases the signal seems useful.

I don't know if anyone is using this feature or not, though.

When I've talked about having a new C binding library, support for this 
"register client with server" pattern is one thing I wanted to have in 
it. I don't know if "directed signals" would fit in to an API like that 
or not, but it might be an interesting thing to look at.

Havoc
_______________________________________________
dbus mailing list
dbus@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/dbus