About the ServiceIPC key
Hi, I'm trying to figure out when the ServiceIPC key should be used. This key is not documented but it is used in the various sample codes. My understanding is that this key is required on 10.4 to check-in with launchd but its use is deprecated on 10.5 and later. Am I right? Thanks, Thomas
At 13:42 +0200 30/6/09, Thomas Clement wrote:
I'm trying to figure out when the ServiceIPC key should be used.
You and me both (-: "ServiceIPC" is definitely not used on 10.5 and later. You can have it in your plist if you want, but it does not affect the operation of launchd. I believe that, on 10.4.x, if you want to check in with launchd you have to set "ServiceIPC" but my memory of 10.4.x is kinda hazy. I took a quick look at the code and it certainly does /something/, which is a change from 10.5.x (-: In summary, I think your analysis is correct. S+E -- Quinn "The Eskimo!" <http://www.apple.com/developer/> Apple Developer Relations, Developer Technical Support, Core OS/Hardware
On Jun 30, 2009, at 4:42 AM, Thomas Clement wrote:
Hi,
I'm trying to figure out when the ServiceIPC key should be used. This key is not documented but it is used in the various sample codes.
My understanding is that this key is required on 10.4 to check-in with launchd but its use is deprecated on 10.5 and later. Am I right?
Correct. On Leopard and later, you can treat the ServiceIPC key as being inferred through the presence of a Sockets or MachServices dictionary. There's no need to explicitly specify it. -- Damien Sorresso BSD Engineering Apple Inc.
On Jun 30, 2009, at 5:56 PM, Damien Sorresso wrote:
On Jun 30, 2009, at 4:42 AM, Thomas Clement wrote:
Hi,
I'm trying to figure out when the ServiceIPC key should be used. This key is not documented but it is used in the various sample codes.
My understanding is that this key is required on 10.4 to check-in with launchd but its use is deprecated on 10.5 and later. Am I right?
Correct. On Leopard and later, you can treat the ServiceIPC key as being inferred through the presence of a Sockets or MachServices dictionary. There's no need to explicitly specify it.
Thanks for the confirmation. I couldn't get an answer from my bug report (asking documentation for the key). Regards, Thomas
On Jun 30, 2009, at 10:15 AM, Thomas Clement wrote:
On Jun 30, 2009, at 5:56 PM, Damien Sorresso wrote:
On Jun 30, 2009, at 4:42 AM, Thomas Clement wrote:
Hi,
I'm trying to figure out when the ServiceIPC key should be used. This key is not documented but it is used in the various sample codes.
My understanding is that this key is required on 10.4 to check-in with launchd but its use is deprecated on 10.5 and later. Am I right?
Correct. On Leopard and later, you can treat the ServiceIPC key as being inferred through the presence of a Sockets or MachServices dictionary. There's no need to explicitly specify it.
Thanks for the confirmation. I couldn't get an answer from my bug report (asking documentation for the key).
You can treat the absence of keys in the documentation as meaning they're deprecated or no longer useful (or they're private and not part of a public API contract). -- Damien Sorresso BSD Engineering Apple Inc.
On Jun 30, 2009, at 7:35 PM, Damien Sorresso wrote:
On Jun 30, 2009, at 10:15 AM, Thomas Clement wrote:
On Jun 30, 2009, at 5:56 PM, Damien Sorresso wrote:
On Jun 30, 2009, at 4:42 AM, Thomas Clement wrote:
Hi,
I'm trying to figure out when the ServiceIPC key should be used. This key is not documented but it is used in the various sample codes.
My understanding is that this key is required on 10.4 to check-in with launchd but its use is deprecated on 10.5 and later. Am I right?
Correct. On Leopard and later, you can treat the ServiceIPC key as being inferred through the presence of a Sockets or MachServices dictionary. There's no need to explicitly specify it.
Thanks for the confirmation. I couldn't get an answer from my bug report (asking documentation for the key).
You can treat the absence of keys in the documentation as meaning they're deprecated or no longer useful (or they're private and not part of a public API contract).
Alright, thanks for the tip. That's good to know. Thomas
participants (3)
-
Damien Sorresso
-
Quinn
-
Thomas Clement