4 ms·
The derivations are auto generated for easier access by various clients. In the metadata, they are marked as original versus not original for later discernment
by textfiles 7y ago
The derivations are auto generated for easier access by various clients. In the metadata, they are marked as original versus not original for later discernment and perhaps rerendering.
- PostOnce 7y agoSo, then, the convenience outweighs the disk space cost? And also, I guess, transcoding the least-popular format on-the-fly is still too CPU intensive for large files (zips of hundreds of jp2 images, etc) Guess I'm looking for a magic bullet where there might not be one, I just want to see Archive.org keep doing what it's doing far into the future. Who knows, maybe we will stumble on some magic bullet, new compression algos (Zstandard? AVIF? AV1?), user clouds for compression (like boinc; archive already lets users assist via bittorrent for bandwidth costs). Thinking out loud. Anyway, keep doin' what you're doing. And podcasting. That's good too :)