Note that it is best to compile any unloadable bundles with the flag
-fno-constant-cfstrings—see Bundles for more details
On 3/15/07, David Blake <email@hidden> wrote:
Where might I find this rule documented?
If this is the case, then all calls to the Carbon Framework would fall
I'll work with this as the rule and replace my use of CFTSR() with
dynamic memory when calling anything outside my bundle.
On 3/15/07, Eric Schlegel <email@hidden> wrote:
> On Mar 14, 2007, at 8:10 PM, David Blake wrote:
> > I agree with the analysis. I more or less implied this when I
> > wrote this up.
> > My question is more whether there is a rule that loaded bundles should
> > be following, or is there a bug in OS X where it shouldn't assume
> > application IDs will always be around and should therefore be creating
> > a copy of the string rather than doing a CFRetain.
> I think it's a rule that an unloadable bundle should not use constant
> CFStrings that could be passed out to code outside of the bundle.
Do not post admin requests to the list. They will be ignored.
Carbon-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden