git-annex/doc/devblog/day_40__another_fine_mess.mdwn
Joey Hess e213ef310f git-annex (5.20140717) unstable; urgency=high
* Fix minor FD leak in journal code. Closes: #754608
  * direct: Fix handling of case where a work tree subdirectory cannot
    be written to due to permissions.
  * migrate: Avoid re-checksumming when migrating from hashE to hash backend.
  * uninit: Avoid failing final removal in some direct mode repositories
    due to file modes.
  * S3: Deal with AWS ACL configurations that do not allow creating or
    checking the location of a bucket, but only reading and writing content to
    it.
  * resolvemerge: New plumbing command that runs the automatic merge conflict
    resolver.
  * Deal with change in git 2.0 that made indirect mode merge conflict
    resolution leave behind old files.
  * sync: Fix git sync with local git remotes even when they don't have an
    annex.uuid set. (The assistant already did so.)
  * Set gcrypt-publish-participants when setting up a gcrypt repository,
    to avoid unncessary passphrase prompts.
    This is a security/usability tradeoff. To avoid exposing the gpg key
    ids who can decrypt the repository, users can unset
    gcrypt-publish-participants.
  * Install nautilus hooks even when ~/.local/share/nautilus/ does not yet
    exist, since it is not automatically created for Gnome 3 users.
  * Windows: Move .vbs files out of git\bin, to avoid that being in the
    PATH, which caused some weird breakage. (Thanks, divB)
  * Windows: Fix locking issue that prevented the webapp starting
    (since 5.20140707).

# imported from the archive
2014-07-17 11:27:25 -04:00

15 lines
921 B
Markdown

Solid day of working on repository recovery. Got `git recover-repository
--force` working, which involves fixing up branches that refer to missing
objects. Mostly straightforward traversal of git commits, trees, blobs, to
find when a branch has a problem, and identify an old version of it that
predates the missing object. (Can also find them in the reflog.)
The main complication turned out to be that `git branch -D` and `git
show-ref` don't behave very well when the commit objects pointed to by refs
are themselves missing. And git has no low-level plumbing that avoids
falling over these problems, so I had to write it myself.
Testing has turned up one unexpected problem: Git's index can itself refer
to missing objects, and that will break future commits, etc. So I need to
find a way to validate the index, and when it's got problems,
either throw it out, or possibly recover some of the staged data from it.