Exa, lost&foundmight invent a new archive format
just zstd blobs in an nbt file or some shit
all i really want is 7z semisolid archives with zstd and good encryption
maybe with an autogenerated dictionary in the head of the file or something
how has nobody used the name "YAR" for Yet Another aRchive yet
(at least not that i can find)
ARSAB, A Really Simple And Bad (Archive Format)
squashfs meets all my needs really
i forgot about my .sqfs.luks thing just long enough to get upset about 7z again
i have contrived an archive format that is an uncompressed zip that contains a zstd dictionary trained on every file in the input as a specially-named file, and every non-special file in the zip is compressed with zstd
using the linux kernel as a corpus and compressing a bunch of other formats for comparison
zsa is the aforementioned contrived archive format
it's better than zip but that is saying very little
here's the dictionary file zstd generated as text http://unascribed.com/f/linux-dict.txt
The encoding for paths within the archive is Windows-1252 (case-insensitive)
i strongly disagree with this
but seems good otherwise and is in the arch repo
the reference implementation command line tool is super barebones and crashes when encountering symlinks
so i can't get it to compress the linux kernel tree
...i dereferenced all the links and now it explodes because it's designed by a windows user and treats
include/uapi/linux/netfilter_ipv6/ip6t_hl.h
and
include/uapi/linux/netfilter_ipv6/ip6t_HL.h
as the same path
i think i hate zarchiver actually
vriska, thief of light 
@unascribed per an issue comment from the dev the readme is incorrect, paths are opaque binary blobs (which will conventionally be UTF-8 on *nix)
@leo i guessss that's fine
Kat
ajft [SEC=yferc ooxie]@unascribed the one thing I'd like to be able to do with #squashfs is generate a multi-part fs, with each chunk a max of, say, 100GB, then be able to mount them later without needing to manually cat them all back together. maybe I haven't looked hard enough, but I can't see a way to do that
@unascribed yarrrrrrr
Scott Williams 🐧@Kat @unascribed The preferred archive format for pirated content.
Joe Cooper 🇺🇦 🍉@unascribed perfect for compressing warez. Because they're pirated.
aburka 🫣@unascribed this is all I could find https://web.archive.org/web/20081005025557/http://wiki.tukaani.org/yar:comparison
@unascribed How about yartar or twar?
em@unascribed what about rawr
sure it’s confusing but it’s also fun
sure it’s confusing but it’s also fun
Richard Barrell@unascribed I suggest it is because for historical reasons RAR already sounds pirate-y. 🦜