Re: Carbon CFM callbacks for CFNetServices
Re: Carbon CFM callbacks for CFNetServices
- Subject: Re: Carbon CFM callbacks for CFNetServices
- From: Chris Silverberg <email@hidden>
- Date: Tue, 21 May 2002 14:41:59 -0700
- Organization: Silverberg.Net
on 5/21/02 1:59 PM, Dean Dauger at email@hidden wrote:
>
As for my being able to move to Mach-O, it's "not yet". Many of my users
>
use OS 9 or mixed computing clusters of Macs running OS 9 and OS X, so I
>
clearly need my code to support both (I've been using NSL/SLP for a year and
>
a half now).
Just to add to this a bit...
I am in the same boat in that I need to continue to support MacOS 9 for at
least a while... (perhaps a year). However, I have chosen a different
route. I am now compiling my application as Mach-O... and as such, I will
produce two separate binaries for my customers.... one for MacOS 8.6 & 9 and
one for MacOS X.
The 8.6/9 version is a single-file CFM application, while the X version is a
bundled Mach-O application. This is actually not a big deal. I'm using a
pre-release of CodeWarrior 8... which compiles both targets like a champ.
To me, having two separate binaries the way to go. I'm in the process of
rewriting my networking code to use CFNetwork, CFSocket, and CFStream. (On
MacOS 9 networking is based on the PowerPlant networking classes). This
certainly does have its disadvantages... I will have two base networking
implementations to maintain and it increases the testing workload. But
adopting the native networking classes should payoff big with much better
performance on MacOS X.
Obviously there are methods of using those API's within a Carbon application
as you are doing with ZeroConfig. And for many folks that will be a viable
solution. It just wasn't the ideal solution for me.
-Chris
_______________________________________________
macnetworkprog mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/macnetworkprog
Do not post admin requests to the list. They will be ignored.