On 29 May 2011, at 06:06, Richard Frith-Macdonald wrote: > AGSParser.m: In function ‘-[AGSParser parseDeclaration]’: > AGSParser.m:1747:9: error: ‘arp’ undeclared (first use in this function) > AGSParser.m:1747:9: note: each undeclared identifier is reported only once > for each function it appears in > AGSParser.m: In function ‘-[AGSParser parseImplementation]’: > AGSParser.m:2209:8: error: ‘arp’ undeclared (first use in this function) > AGSParser.m: In function ‘-[AGSParser parseInterface]’: > AGSParser.m:2404:4: error: ‘arp’ undeclared (first use in this function) > AGSParser.m: In function ‘-[AGSParser parseMethodIsDeclaration:]’: > AGSParser.m:2953:4: error: ‘arp’ undeclared (first use in this function) > AGSParser.m: In function ‘-[AGSParser parseProtocol]’: > AGSParser.m:3664:4: error: ‘arp’ undeclared (first use in this function) > AGSParser.m: In function ‘-[AGSParser setupBuffer]’: > AGSParser.m:4111:4: error: ‘arp’ undeclared (first use in this function)
These are all related to the autorelease pool changes. Moving this expansion of the CREATE_AUTORELEASE_POOL macro out of the conditional will fix it: NSAutoreleasePool *X = [NSAutoreleasePool new] Since we're sending +new, instead of +alloc / -init, this means that we're getting the singleton object in a single message send in GC mode, so I don't think the overhead is excessive... David -- Sent from my Apple II _______________________________________________ Discuss-gnustep mailing list [email protected] https://lists.gnu.org/mailman/listinfo/discuss-gnustep
