bup 0.33.8 released

[Rob Browning]

https://bup.github.io/
https://github.com/bup/bup/tags
https://codeberg.org//bup/bup/tags

This is a bugfix release. Please give it a try and let us know what's
broken. If you're new to bup, start with the README, and then HACKING
if you'd like to help further:

https://github.com/bup/bup/blob/main/README.md
https://github.com/bup/bup/blob/main/DESIGN.md
https://github.com/bup/bup/blob/main/HACKING.md

https://codeberg.org/bup/bup/src/branch/main/README.md
https://codeberg.org/bup/bup/src/branch/main/DESIGN.md
https://codeberg.org/bup/bup/src/branch/main/HACKING.md

(As usual, while we expect bup to work fairly well, I still wouldn't
recommend it as your sole backup strategy. I'd suggest a periodic
tar/rsync/etc. backstop.)

Notable changes in 0.33.8 since 0.33.7
======================================

May require attention
---------------------

* Previously, `bup get --force-pick: SRC /.tag/DEST` created broken
commits if the `DEST` was not itself a commit (the parent would be
whatever `DEST` initially pointed to).

* `bup split --copy` now writes the split data to standard output
instead of Python memoryview representations like

<memory at 0x7f7a89358ac0><memory at 0x7f7a89358a00>...

General
-------

* The full git path is now reported for missing objects encountered
during a `gc` or `validate-object-refs` invocation,
e.g. `.../some/path/.bupm` instead of just `.../some/path/` when
only the `.bupm` is missing.

Bugs
----

* `bup on HOST get ...` should no longer hang in some situations where
the client could provoke duplicate index-cache suggestions from the
server, which the client then treated as an error.

* `par2` changed its behavior in 1.0 to be incompatible with `bup`'s
use of symlinks to mitigate a `par2` bug (see the [0.33.4 release
notes](0.33.4-from-0.33.3.md) for additional information. `bup` now
uses hardlinks instead.

* Saves with identical dates won't end up with the same name
(e.g. 2025-01-08-135615 in "bup ls BRANCH") when they're not
adjacent -- when one isn't the other's direct parent. Previously bup
appended an increasing "-N" to disambiguate duplicates, but only
when they were directly related. Now it appends across all
duplicates.

* `bup prune-older` should no longer be confused by duplicate save
names, i.e. commits with duplicate commit times (author
dates). Previously it would fail with a message like "error: cannot
access SAVE in SAVE".

* Data durability should be improved in some cases where `bup` wasn't
fully syncing the data, or wasn't syncing the parent directory after
a rename.

* Durability should also be improved on macos. There were some cases
where `bup` wasn't relying on `fcntl.F_FULLFSYNC` to avoid macos'
inadequate fsync/fdatasync implementation.

* `bup` should no longer hang on exit when trying to shut down its
internal output filtering.

Thanks to (at least)
====================

Anton Khirnov, Ed Maste, Greg Troxel, Johannes Berg, and Rob Browning

--
Rob Browning
rlb @defaultvalue.org and @debian.org
GPG as of 2011-07-10 E6A9 DA3C C9FD 1FF8 C676 D2C4 C0F0 39E9 ED1B 597A
GPG as of 2002-11-03 14DD 432F AE39 534D B592 F9A0 25C8 D377 8C7E 73A4