8 ms·
right now, most of it is coming in as ZSTD with a central index. older stuff was gzip
by jrwr 5y ago
right now, most of it is coming in as ZSTD with a central index. older stuff was gzip
- AlexanderTheGr8 5y agoI am working on a ML-based compression which will be slow to compress/decompress but will give much better compression than zstd/gzip (even 2-3x better). Do you think that it's a useful algorithm for archival reasons or for storing huge amounts of data in cloud?