Question

This is considered typical

- (id)init {
    self = [super init];
    if (self) {
        // <#initializations#>
    }
    return self;
}

but wouldn’t it be better to go with something like this which actually responds appropriately?

- (id)init {
    self = [super init];
    if (self) {
        // <#initializations#>
    } else {
       @throw [NSException exceptionWithName:NSInternalInconsistencyException reason:@"you think your constructor is executing, but it's not"] userInfo:nil]
    }
    return self;
}

The corollary to this question is, “under what conditions would [super init] return nil and shouldn’t you handle that in the init method?“

Answer 1

One reason, why you should do, what JustSid is saying:

In object-orientated design you should always code, as if you maybe will hand your class over to another project by another developer. So you can’t assume, that a failure in initialization in his project may be as bad as it is probably in yours. Maybe this developer is you in 5 years. Imagine your hassle to fix your 200 classes, you want to reuse.