mirrors/git-annex
2
0
Fork 0
Code Issues Pull requests Projects Releases Packages 1 Wiki Activity Actions

comment and toddo

Browse source
This commit is contained in:
Joey Hess 2019-03-22 09:23:31 -04:00
parent 2e4fedf1a4
commit 5fea7efee7
No known key found for this signature in database
GPG key ID: DB12DB0FF05F8F38
2 changed files with 30 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
doc/bugs/__34__git_status__34___thread_blocked_indefinitely_in_an_MVar_operation/comment_4_df73b69e651c728943a404223f350ebf._comment Normal file
View file

@ -0,0 +1,15 @@
[[!comment format=mdwn
username="joey"
subject="""comment 4"""
date="2019-03-22T13:14:51Z"
content="""
Ok, it makes sense that sqlite would crash creating a DB if the disk was
full.
I hope that the earlier Mvar error was the same problem, just a different
handling of the exception in that version. Or something similar failing on
full disk.
annex.thin needing hard links is a limitation that may one day be lifted,
see [[todo/annex.thin_without_hardlinks]]
"""]]

15
doc/todo/annex.thin_without_hardlinks.mdwn Normal file
View file

@ -0,0 +1,15 @@
Currently annex.thin needs hard link support to be efficient;
it hard links the content from .git/annex/objects into the work tree.
When hard links are not supported, two copie of checked out files exist on
disk.
Would it be possible to make it work w/o hard links? Note that direct mode
does avoid two copies of files.
IIRC the main reason for the hard link is so, when git checkout deletes a
work tree file, the only copy of the file is not lost. Seems this would
need a git hook run before checkout to rescue such files.
Also some parts of git-annex's code, including `withObjectLoc`, assume
that the .annex/objects is present, and so it would need to be changed
to look at the work tree file. --[[Joey]]