HNKit isn't in ARC, and does a myriad of things that I think are unnecessary and can be maintained cleaner (there's absolutely no documentation!!!). Those methods are actually very quick (imperceptibly fast) right now - but I'm not a computer scientist by any means, so I believe they can be made even faster by someone that knows a better data structure to use. Each comment comes back out of order with a parent ID and a comment ID. A comment that is a reply to Comment 1 will have the parent ID matching Comment 1's comment ID (hence my linked list structure right now). There's got to be a better way, and that's also why I open sourced the whole thing.
HNKit is also doing XML parsing over everything, and string comparisons can be fairly slow as well. That, in conjunction with not being ARC ready and a total lack of documentation made me choose something else entirely.
HNKit is also doing XML parsing over everything, and string comparisons can be fairly slow as well. That, in conjunction with not being ARC ready and a total lack of documentation made me choose something else entirely.