Quote from: AlexGR on January 30, 2016, 06:26:02 PM

I guess it depends on the nature of data.

It does. I've recently had to learn something in regards to compression. With some types of data the savings can be great, but with random data they are negligible.

Quote from: AlexGR on January 30, 2016, 06:26:02 PM

If someone broadcasts a 1mb/2mb/4mb solved block, that can probably be compressed as it will have redundant bytes.

I don't think it would work though. This was discussed before and would have probably been implemented already(?).

Quote from: AlexGR on January 30, 2016, 06:26:02 PM

If someone is sending the blockchain to someone who is lagging 1-2-3 years, well, he could send it in compressed batches. Surely that takes compression (?)
134212277 - blk00000.dat
99375871 - blk00000.dat.gz (74.04%)
97307825 - blk00000.dat.bz2 (72.5%)

Just ran my own test with 7zip on Windows (Compression format Maximum).

~73%.

Some people had to be put on ignore due to trolling and idiocy.

Quote from: watashi-kokoto on January 30, 2016, 06:30:53 PM

I think the segwit pruning would help much more than this. The gains should be on the order of 50% reduction.

Segwit pruning?