Sourceware Survey 2025 Results
In the end we got 103 (!) responses with a nice mix of developers, users and maintainers from various hosted projects.
https://sourceware.org/survey-2025
#binutils #cygwin #dwarfstd #elfutils #gcc #gdb #glibc #libabigail #newlib #systemTap #valgrind #bzip2 #libffi #dwz #debugedit #gnupoke #bunsen #lvm2 #annobin #gnu-gabi #cgen #kawa #insight #pacme
#libarchive 3.7.9 has been released (#MultiFormatArchive / #CompressionLibrary / #FileArchiver / #DataCompression / #7Zip / #7z / #RAR / #ZIP / #GZip / #TAR / #XAR / #WARC / #BZIP2 / #XZ) https://www.libarchive.org/
we've just released `bzip2` version 0.5.2, which supports webassembly
https://crates.io/crates/bzip2
with the `libbz2-rs-sys` feature flag you won't even need a C toolchain
![[dependencies.bzip2]
version = "0.5.2"
default-features = false
features = ["libbz2-rs-sys"]](https://media.hachyderm.io/media_attachments/files/114/081/954/785/290/283/original/23d361dd51cecf0a.png "[dependencies.bzip2]
version = \"0.5.2\"
default-features = false
features = [\"libbz2-rs-sys\"]")
@kaasbaas #wikipedia (English language) is only ~22GB of #bzip2 compressed #xml (uncompressed size is ~86GB).
is it possible to access it without decompression? I guess #random #access to .xml.bz2 should be a solved problem, right?
we're routinely using gzip with random access in #bioinformatics ie via #samtools or #tabix
EDIT:
Wikipedia xml.bz2 does support random access for multistream version. does @kiwix or any other wiki reader support it? I couldn't find info on their website...
BZip3
在 Hacker News 上看到 BZip3 的連結：「Bzip3: A spiritual successor to BZip2 (github.com/kspalaiologos)」。
雖然名字看起來與 bzip2 有關，但看起來是不同的人弄出來的東西，不過有些經典的演算法有留下來用，像是 Burrows-Wheeler transform。
另外值得一提的是，bzip2 是 1996 年出的 (不過 1.0 大約是 2000 年時出的)，BZip3 的第一個 release 在 2022 年，這段時間也累積了不少有趣的演算法可以用。
無損壓縮中如果期望有比較的壓縮率，目前比較常用的應該是 LZMA 類的演算法 (差不多是 2001 年出現的)，用的工具通常會是 X
@fluoriteByte @cobra nodds in agreement Like #bzip2 and #lzma2...
Speaking of #bzip2 I wonder what happened to bzip1 or just #bzip. Was the sequel just too good? 
RE: https://social.sdf.org/users/MistressRemilia/statuses/111861667590062847
StarFive has donated 4 VisionFive-2 risc-v boards with 8GB, 4-core JH7110 supporting the RV64GC ISA for the CI running on https://builder.sourceware.org for Sourceware hosted projects.
They are running ubuntu server 23.10 https://wiki.ubuntu.com/RISC-V/StarFive%20VisionFive%202
Various projects are already running on risc-v: #annobin, #binutils, #bzip2, #debugedit, #dwz, #elfutils, #gnupoke and #libabigail.
https://builder.sourceware.org/buildbot/#/builders?tags=riscv
Please contact the builder project if you want to help out adding other Sourceware hosted projects.
@ASTAFATHERSATAN Granted, I'm not gonna implement a custom compression and decompression algo in the Linux kernel as I'm not only able to do so with my skills but literally doubt I could beat the #XZ numbers without sinking in decades of R&D - and even then that may be a dubious investment of time and resources...
I only used #bzip2 for said backups as it's convenient, fast and didn't require me to setup stuff to work.
@chakuari OFC.
In theory offering more than just one is "trivial" in the sense that for OS/1337 the idea is to have statical binaries (and maybe the few necessary configs) as "#packages" so it's just a download as an archive (#bzip2 because it's available in #toybox) and just pull that and place them in the system.
OFC absolute hardcore folks will literally do #eMail just with #curl, #cat, #sed & #awk I guess, but ideally offering a convenient alternative like #neomutt is better.
It's always sad when the dedicated #apt binding cannot handle #apt's own files (Packages, Sources, etc.), but only when they're compressed in a particular way.
With gzip on the way out, not being able to grok bz2 (~ unimportant) and xz (~ vital) isn't so good…
See https://bugs.debian.org/932491#20 for #xz test cases, and https://bugs.debian.org/932491#27 for #bzip2 test cases (which seem to confirm the problem isn't somewhere in the built-in #lzma module).
H/T @bremner who spotted this 4+ years ago…
Ah, so there was a bzip1 (technically, bzip 0.21):
https://web.archive.org/web/19980704181204/http://www.muraroa.demon.co.uk/
By way of https://unix.stackexchange.com/questions/125798/what-happened-to-bzip1
@girlfreddy @BleepingComputer
I bought my #WinRAR license before #7Zip existed. I needed to manage a bunch of log files and WinRAR had advanced command line options that #WinZip and Windows ported #gzip / #bzip2 could only dream of.
Edit: I use 7zip now, too. WinRAR still has some command line features that 7z.exe lacks.
The #gentoo #sparc builders are disabled for now because the worker had become a little too unstable.
https://builder.sourceware.org
We hope to get it back soon because it is doing a lot of builds for #Sourceware projects.
#elfutils, #libabigail, #bzip2, #debugedit, #gccrust, #binutils, #gdb, #gcc, #systemtap, #glibc, #gnupoke.
#dar 3/ Dar's homepage is http://dar.linux.free.fr/ . As a simple file archiver tool, dar is great. Random access is preserved, you can compress with #gzip #bzip2 #lzo #xz #zstd or #lz4. You can encrypt with #GnuPG or symmetric #AES. You can stream if you want. You can split ("slice") across multiple media, and dar will prompt you for the slice(s) you need and seek you right to them.
That's cool, but we're just getting started.
You're familiar with #bzip2, but do you know why we don't have a "bzip"/"bzip1"?
bzip-0.21, released in 1997, used arithmetic coding. Due to patent infringement fears, the author changed to Huffman coding and released bzip2:
https://web.archive.org/web/19980704181204/http://www.muraroa.demon.co.uk/
You can still get access to bzip-0.21 however:
Mima-sama 
:EA DATA. SF: 
