|
From: | Adam Fedor |
Subject: | Re: GNUstep base/make pre-release |
Date: | Tue, 18 Feb 2003 21:08:17 -0700 |
User-agent: | Mozilla/5.0 (X11; U; Linux ppc; en-US; rv:1.1) Gecko/20020905 |
Stéphane Corthésy wrote:
Hi,Here's a last modification suggestion, already posted on bug-gnustep, but without any reply.I compared how GNUStep and Apple implement NSKeyValueCoding and saw that GNUStep's implementation raises a NSGenericException exception in -[NSObject(KeyValueCoding) handleQueryWithUnboundKey:] and -[NSObject(KeyValueCoding) handleTakeValue:forUnboundKey:], without any userInfo, whereas Apple's raises (it is undocumented) a @"NSUnknownKeyException" exception (no symbolic names though, I should bug Apple for that), with a userInfo containing keys @"NSTargetObjectUserInfoKey" and @"NSUnknownUserInfoKey"Stéphane
What's the chance that, if I put this in the freeze branch, it's going to mess someone up, because they are checking for this and it doesn't come out the way they expect?
Any thoughts?
[Prev in Thread] | Current Thread | [Next in Thread] |