On 05/06, brian m. carlson wrote:

This is the eighth series of patches to convert unsigned char [20] to
struct object_id.  This series converts lookup_commit, lookup_blob,
lookup_tree, lookup_tag, and finally parse_object to struct object_id.

A small number of functions have temporaries inserted during the
conversion in order to allow conversion of functions that still need to
take unsigned char *; they are removed either later in the series or
will be in a future series.

This series can be fetched from the object-id-part8 branch from either
of the follwing:

https://github.com/bk2204/git
https://git.crustytoothpaste.net/git/bmc/git.git

Changes from v2:
* Remove spurious space after ampersand.
* Undo more needless line rewrapping.
* Expand computation for notes path.
* Remove check for line->len with parse_oid_hex.

Changes from v1:
* Rebase on master.  This led to a conflict with the ref-cache changes in patch
  39.  Extra-careful review here would be welcome.
* Undo the needless line rewrapping.
* Fix the commit message typo.
* Use GIT_MAX_RAWSZ instead of struct object_id for the pack checksum.

I wasn't able to find any issues with v3.  Looks good.

-- 
Brandon Williams