>>
>>48977
1. That's correct. Actually, they're downloaded in order of the id of the upload, but of course that is the same thing as chronological order. So, 2007, then 2008, and so on.
2. Actually, there wasn't supposed to be a data2.txt file in /any/ of the directories... I was sure I had deleted them all... anyway, as you can see, I originally had data2 files listing all relevant metadata, but when I realized that approach wouldn't scale well (considering gelbooru's 1.75m+ database), I abandoned it in favor storing all metadata in one centralized file. Basically, it's leftovers from an obsolete approach.
I know it's all rather dull; to be fair, the first several thousand (???) posts are copied from Danbooru and from 2007 (or before). The real magic comes once I get a search interface up and running, but that's involved than a few torrents of raw images... For future releases, should I include data2 files? I had abandoned it in favor of the 'centralized' approach, but if I'm going to be releasing raws for a while, /some/ source of metadata might be useful... the only problem is that data2's will become outdated overtime, and not able to be updated. It would still be better than nothing, I guess. Argh, I talk too much.