git-annex.git
4 years agoCommit Debian 3.0 (quilt) metadata
Sean Whitton [Mon, 18 Oct 2021 21:47:41 +0000 (14:47 -0700)]
Commit Debian 3.0 (quilt) metadata

[dgit (9.14) quilt-fixup]

4 years agochangelog
Sean Whitton [Mon, 18 Oct 2021 21:46:39 +0000 (14:46 -0700)]
changelog

4 years agoMerge tag '8.20211011'
Sean Whitton [Mon, 18 Oct 2021 21:44:20 +0000 (14:44 -0700)]
Merge tag '8.20211011'

tagging package git-annex version 8.20211011

# gpg: Signature made Mon 11 Oct 2021 09:53:17 AM MST
# gpg:                using RSA key 28A500C35207EAB72F6C0F25DB12DB0FF05F8F38
# gpg: Good signature from "Joey Hess <joeyh@joeyh.name>" [full]
# Primary key fingerprint: E85A 5F63 B31D 24C1 EBF0  D81C C910 D922 2512 E3C7
#      Subkey fingerprint: 28A5 00C3 5207 EAB7 2F6C  0F25 DB12 DB0F F05F 8F38

4 years agoreleasing package git-annex version 8.20211011
Joey Hess [Mon, 11 Oct 2021 16:53:17 +0000 (12:53 -0400)]
releasing package git-annex version 8.20211011

4 years agodocument how to resume downloads
Joey Hess [Mon, 11 Oct 2021 16:40:16 +0000 (12:40 -0400)]
document how to resume downloads

4 years agocomment
Joey Hess [Mon, 11 Oct 2021 16:22:38 +0000 (12:22 -0400)]
comment

4 years agoupdate
Joey Hess [Mon, 11 Oct 2021 14:06:36 +0000 (10:06 -0400)]
update

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Fri, 8 Oct 2021 17:27:23 +0000 (13:27 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoMerge branch 'borgchunks'
Joey Hess [Fri, 8 Oct 2021 17:26:45 +0000 (13:26 -0400)]
Merge branch 'borgchunks'

4 years agotodo
Joey Hess [Fri, 8 Oct 2021 17:26:40 +0000 (13:26 -0400)]
todo

4 years agoImportableContentsChunkable
Joey Hess [Wed, 6 Oct 2021 21:05:32 +0000 (17:05 -0400)]
ImportableContentsChunkable

This improves the borg special remote memory usage, by
letting it only load one archive's worth of filenames into memory at a
time, and building up a larger tree out of the chunks.

When a borg repository has many archives, git-annex could easily OOM
before. Now, it will use only memory proportional to the number of
annexed keys in an archive.

Minor implementation wart: Each new chunk re-opens the content
identifier database, and also a new vector clock is used for each chunk.
This is a minor innefficiency only; the use of continuations makes
it hard to avoid, although putting the database handle into a Reader
monad would be one way to fix it.

It may later be possible to extend the ImportableContentsChunkable
interface to remotes that are not third-party populated. However, that
would perhaps need an interface that does not use continuations.

The ImportableContentsChunkable interface currently does not allow
populating the top of the tree with anything other than subtrees. It
would be easy to extend it to allow putting files in that tree, but borg
doesn't need that so I left it out for now.

Sponsored-by: Noam Kremen on Patreon
4 years agoAdded a comment: Re: Resuming an interrupted download
alex@f04d0d3c452a2a99b27ccc93c1543bee4a1bf5be [Fri, 8 Oct 2021 03:06:57 +0000 (03:06 +0000)]
Added a comment: Re: Resuming an interrupted download

4 years agoAdded a comment
jkniiv [Thu, 7 Oct 2021 18:27:19 +0000 (18:27 +0000)]
Added a comment

4 years agoAdded a comment
asakurareiko@f3d908c71c009580228b264f63f21c7274df7476 [Thu, 7 Oct 2021 16:27:39 +0000 (16:27 +0000)]
Added a comment

4 years agoadd footnote about the permissions required and Microsoft blog post
jkniiv [Thu, 7 Oct 2021 16:25:37 +0000 (16:25 +0000)]
add footnote about the permissions required and Microsoft blog post

4 years agoavoid cursor jitter when updating progress display
Joey Hess [Thu, 7 Oct 2021 14:58:49 +0000 (10:58 -0400)]
avoid cursor jitter when updating progress display

When the progress display gets longer, and then shorter again, it causes
the cursor to jitter back and forth. Somehow I never noticed this until
this morning, but then it became intolerable to watch.

To fix it, pad the progress display to the maximum length it's occupied.

Sponsored-by: Svenne Krap on Patreon
4 years agoremove mention of mount option `case=dir` (turned out to be unnecessary)
jkniiv [Thu, 7 Oct 2021 11:40:25 +0000 (11:40 +0000)]
remove mention of mount option `case=dir` (turned out to be unnecessary)

4 years agoAdded a comment
jkniiv [Thu, 7 Oct 2021 11:36:49 +0000 (11:36 +0000)]
Added a comment

4 years agoAdded a comment
asakurareiko@f3d908c71c009580228b264f63f21c7274df7476 [Thu, 7 Oct 2021 06:21:25 +0000 (06:21 +0000)]
Added a comment

4 years agoAdded a comment
asakurareiko@f3d908c71c009580228b264f63f21c7274df7476 [Thu, 7 Oct 2021 06:17:30 +0000 (06:17 +0000)]
Added a comment

4 years ago(no commit message)
asakurareiko@f3d908c71c009580228b264f63f21c7274df7476 [Thu, 7 Oct 2021 05:53:23 +0000 (05:53 +0000)]

4 years agoAdded a comment: the WSL1 use case
jkniiv [Thu, 7 Oct 2021 04:12:48 +0000 (04:12 +0000)]
Added a comment: the WSL1 use case

4 years agorename WSL1 section to highlight date, add wording about being experimental, reword...
jkniiv [Thu, 7 Oct 2021 03:56:37 +0000 (03:56 +0000)]
rename WSL1 section to highlight date, add wording about being experimental, reword some awkwardness, add further directions

4 years agoAdd steps for WSL1
asakurareiko@f3d908c71c009580228b264f63f21c7274df7476 [Wed, 6 Oct 2021 22:35:27 +0000 (22:35 +0000)]
Add steps for WSL1

4 years agoAdded a comment
asakurareiko@f3d908c71c009580228b264f63f21c7274df7476 [Wed, 6 Oct 2021 21:10:34 +0000 (21:10 +0000)]
Added a comment

4 years agobranch
Joey Hess [Wed, 6 Oct 2021 21:08:57 +0000 (17:08 -0400)]
branch

4 years agoprogress in my head
Joey Hess [Wed, 6 Oct 2021 18:45:12 +0000 (14:45 -0400)]
progress in my head

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Wed, 6 Oct 2021 13:04:12 +0000 (09:04 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoAdded a comment
falsifian [Wed, 6 Oct 2021 03:36:02 +0000 (03:36 +0000)]
Added a comment

4 years agoremove unused imports
Joey Hess [Wed, 6 Oct 2021 01:27:06 +0000 (21:27 -0400)]
remove unused imports

4 years agominor optimisation
Joey Hess [Wed, 6 Oct 2021 01:26:11 +0000 (21:26 -0400)]
minor optimisation

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Wed, 6 Oct 2021 00:28:26 +0000 (20:28 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoremove old and uncessary sanity check
Joey Hess [Wed, 6 Oct 2021 00:25:51 +0000 (20:25 -0400)]
remove old and uncessary sanity check

All backends use genKeyName, which uses preSanitizeKeyName, which will
escape newlines.

I suspect this newline prevention was added back when git-annex was
reading the output of sha1sum..

Sponsored-by: Kevin Mueller on Patreon
4 years agoconvert Key to ShortByteString
Joey Hess [Wed, 6 Oct 2021 00:20:08 +0000 (20:20 -0400)]
convert Key to ShortByteString

This adds the overhead of a copy when serializing and deserializing keys.
I have not benchmarked much, but runtimes seem barely changed at all by that.

When a lot of keys are in memory, it improves memory use.

And, it prevents keys sometimes getting PINNED in memory and failing to GC,
which is a problem ByteString has sometimes. In particular, git-annex sync
from a borg special remote had that problem and this improved its memory
use by a large amount.

Sponsored-by: Shae Erisson on Patreon
4 years agocomment and correct incorrect info in previous comment
Joey Hess [Tue, 5 Oct 2021 23:05:20 +0000 (19:05 -0400)]
comment and correct incorrect info in previous comment

4 years agoAdded a comment
tomdhunt [Tue, 5 Oct 2021 22:07:45 +0000 (22:07 +0000)]
Added a comment

4 years agoAdded a comment
tomdhunt [Tue, 5 Oct 2021 21:30:03 +0000 (21:30 +0000)]
Added a comment

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Tue, 5 Oct 2021 21:20:53 +0000 (17:20 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agocomment
Joey Hess [Tue, 5 Oct 2021 21:20:32 +0000 (17:20 -0400)]
comment

4 years agoAdded a comment: Thanks!
bjornw@6a7d7d0413efc7ed3bb44922586f040bb768b71c [Tue, 5 Oct 2021 21:09:44 +0000 (21:09 +0000)]
Added a comment: Thanks!

4 years agoAdded a comment
falsifian [Tue, 5 Oct 2021 20:55:32 +0000 (20:55 +0000)]
Added a comment

4 years agoupdate
Joey Hess [Tue, 5 Oct 2021 20:51:54 +0000 (16:51 -0400)]
update

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Tue, 5 Oct 2021 20:33:44 +0000 (16:33 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agocomment
Joey Hess [Tue, 5 Oct 2021 20:32:10 +0000 (16:32 -0400)]
comment

4 years agouse bytestring filepaths more
Joey Hess [Tue, 5 Oct 2021 19:42:29 +0000 (15:42 -0400)]
use bytestring filepaths more

This should be more efficient, and allocate less.

Sponsored-by: Graham Spencer on Patreon
4 years agoAdded a comment
tomdhunt [Tue, 5 Oct 2021 19:08:23 +0000 (19:08 +0000)]
Added a comment

4 years agoconvert ExportLocation to ShortByteString to avoid PINNED memory fragmentation
Joey Hess [Tue, 5 Oct 2021 18:51:55 +0000 (14:51 -0400)]
convert ExportLocation to ShortByteString to avoid PINNED memory fragmentation

This adds the overhead of a copy whenever converting to/from ExportLocation and
ImportLocation.

borg: Some improvements to memory use when importing a lot of archives.
(It's still pretty bad.)

Sponsored-by: Mark Reidenbach on Patreon
4 years agoupdate
Joey Hess [Tue, 5 Oct 2021 17:24:31 +0000 (13:24 -0400)]
update

4 years agocomment
Joey Hess [Tue, 5 Oct 2021 17:11:20 +0000 (13:11 -0400)]
comment

4 years agoresponse
Joey Hess [Tue, 5 Oct 2021 16:16:18 +0000 (12:16 -0400)]
response

4 years agopromote comment to bug
Joey Hess [Tue, 5 Oct 2021 15:55:33 +0000 (11:55 -0400)]
promote comment to bug

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Tue, 5 Oct 2021 15:11:44 +0000 (11:11 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoupdate
Joey Hess [Tue, 5 Oct 2021 14:43:27 +0000 (10:43 -0400)]
update

4 years agocomment
Joey Hess [Sun, 3 Oct 2021 22:45:37 +0000 (18:45 -0400)]
comment

4 years agoAdded a comment: Memory issues
tomdhunt [Sun, 3 Oct 2021 22:30:04 +0000 (22:30 +0000)]
Added a comment: Memory issues

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Sun, 3 Oct 2021 22:18:30 +0000 (18:18 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoremove errant print debug
Joey Hess [Sun, 3 Oct 2021 22:18:04 +0000 (18:18 -0400)]
remove errant print debug

4 years ago(no commit message)
falsifian [Sun, 3 Oct 2021 17:06:21 +0000 (17:06 +0000)]

4 years agoAdded a comment
jkniiv [Sat, 2 Oct 2021 19:58:17 +0000 (19:58 +0000)]
Added a comment

4 years agoAdded a comment
spwhitton [Sat, 2 Oct 2021 17:04:02 +0000 (17:04 +0000)]
Added a comment

4 years agoremove information about old version of debian
Joey Hess [Sat, 2 Oct 2021 16:32:00 +0000 (12:32 -0400)]
remove information about old version of debian

4 years agoAdded a comment: Resuming an interrupted download
alex@f04d0d3c452a2a99b27ccc93c1543bee4a1bf5be [Fri, 1 Oct 2021 23:27:31 +0000 (23:27 +0000)]
Added a comment: Resuming an interrupted download

4 years agoremoved
alex@f04d0d3c452a2a99b27ccc93c1543bee4a1bf5be [Fri, 1 Oct 2021 23:22:19 +0000 (23:22 +0000)]
removed

4 years agoAdded a comment: Resuming an interrupted download
alex@f04d0d3c452a2a99b27ccc93c1543bee4a1bf5be [Fri, 1 Oct 2021 23:18:33 +0000 (23:18 +0000)]
Added a comment: Resuming an interrupted download

4 years agoreinject: Fix crash when reinjecting a file from outside the repository
Joey Hess [Fri, 1 Oct 2021 18:04:18 +0000 (14:04 -0400)]
reinject: Fix crash when reinjecting a file from outside the repository

Commit 4bf7940d6b912fbf692b268f621ebd41ed871125 introduced this
problem, but was otherwise doing a good thing. Problem being
that fileRef "/foo" used to return ":./foo", which was actually wrong,
but as long as there was no foo in the local repository, catKey
could operate on it without crashing. After that fix though, fileRef
would return eg "../../foo", resulting in fileRef returning
":./../../foo", which will make git cat-file crash since that's
not a valid path in the repo.

Fix is simply to make fileRef detect paths outside the repo and return
Nothing. Then catKey can be skipped. This needed several bugfixes to
dirContains as well, in previous commits.

In Command.Smudge, this led to needing to check for Nothing. That case
should actually never happen, because the fileoutsiderepo check will
detect it earlier.

Sponsored-by: Brock Spratlen on Patreon
4 years agofix bug in dirContains
Joey Hess [Fri, 1 Oct 2021 17:42:15 +0000 (13:42 -0400)]
fix bug in dirContains

dirContains "." ".." was incorrectly true
because normalize ("." </> "..") = ".."

Sponsored-by: Jochen Bartl on Patreon
4 years agofix bug in dirContains
Joey Hess [Fri, 1 Oct 2021 17:15:52 +0000 (13:15 -0400)]
fix bug in dirContains

dirContains ".." "../.." was incorrectly True.

This does not seem to be an exploitable security hole, at least
as dirContains is used in git-annex.

Sponsored-by: Jochen Bartl on Patreon
4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Fri, 1 Oct 2021 16:35:29 +0000 (12:35 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoavoid uncessary call to inAnnex
Joey Hess [Fri, 1 Oct 2021 16:02:35 +0000 (12:02 -0400)]
avoid uncessary call to inAnnex

sync --content: Avoid a redundant checksum of a file that was
incrementally verified, when used on NTFS and perhaps other filesystems.

When sync has just gotten the content, it does not need to check inAnnex a
second time. On NTFS, for some reason the write of the inode cache after
it gets the content is not immediately able to be read, and with an
empty/non-matching inode cache due to that stale data, inAnnex falls back
to hashing the whole object to determine if it's present.

Sponsored-by: Brock Spratlen on Patreon
4 years agoanalysis
Joey Hess [Fri, 1 Oct 2021 15:49:12 +0000 (11:49 -0400)]
analysis

4 years agoadd sign off
spwhitton [Thu, 30 Sep 2021 21:19:39 +0000 (21:19 +0000)]
add sign off

4 years agopost bug
spwhitton [Thu, 30 Sep 2021 21:15:04 +0000 (21:15 +0000)]
post bug

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Thu, 30 Sep 2021 19:20:59 +0000 (15:20 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agostarted analysis
Joey Hess [Thu, 30 Sep 2021 19:20:44 +0000 (15:20 -0400)]
started analysis

4 years agoAdded a comment: Email resent from my personal domain jibun.eu
jkniiv [Thu, 30 Sep 2021 18:22:44 +0000 (18:22 +0000)]
Added a comment: Email resent from my personal domain jibun.eu

4 years agonote about coreutils 9.0 supporting CoW by default
Joey Hess [Thu, 30 Sep 2021 18:12:58 +0000 (14:12 -0400)]
note about coreutils 9.0 supporting CoW by default

4 years agocomment
Joey Hess [Thu, 30 Sep 2021 16:52:02 +0000 (12:52 -0400)]
comment

4 years agoAdded a comment
jkniiv [Wed, 29 Sep 2021 04:38:08 +0000 (04:38 +0000)]
Added a comment

4 years agoAdded a comment
username [Tue, 28 Sep 2021 22:56:17 +0000 (22:56 +0000)]
Added a comment

4 years agocomment
Joey Hess [Tue, 28 Sep 2021 21:58:10 +0000 (17:58 -0400)]
comment

4 years agocomment
Joey Hess [Tue, 28 Sep 2021 20:33:21 +0000 (16:33 -0400)]
comment

4 years agoAdded a comment
adina.wagner@2a4cac6443aada2bd2a329b8a33f4a7b87cc8eff [Tue, 28 Sep 2021 19:19:20 +0000 (19:19 +0000)]
Added a comment

4 years agocomment
Joey Hess [Tue, 28 Sep 2021 18:14:37 +0000 (14:14 -0400)]
comment

4 years agoresponse
Joey Hess [Tue, 28 Sep 2021 18:09:56 +0000 (14:09 -0400)]
response

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Tue, 28 Sep 2021 18:00:44 +0000 (14:00 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agofollowup
Joey Hess [Tue, 28 Sep 2021 18:00:10 +0000 (14:00 -0400)]
followup

4 years agoAdded a comment
adina.wagner@2a4cac6443aada2bd2a329b8a33f4a7b87cc8eff [Tue, 28 Sep 2021 12:53:14 +0000 (12:53 +0000)]
Added a comment

4 years agoresume properly when copying a file to/from a local git remote is interrupted (take 2)
Joey Hess [Mon, 27 Sep 2021 20:03:01 +0000 (16:03 -0400)]
resume properly when copying a file to/from a local git remote is interrupted (take 2)

This method avoids breaking test_readonly. Just check if the dest file
exists, and avoid CoW probing when it does, so when CoW probing fails,
it can resume where the previous non-CoW copy left off.

If CoW has been probed already to work, delete the dest file
since a CoW copy will presumably work. It seems like it would be almost
as good to just skip CoW copying in this case too, but consider that the
dest file might have started to be copied from some other remote, not
using CoW, but CoW has been probed to work to copy from the current
place.

Sponsored-by: Dartmouth College's Datalad project
4 years agorevert change that broke test_readonly
Joey Hess [Mon, 27 Sep 2021 19:53:12 +0000 (15:53 -0400)]
revert change that broke test_readonly

commit 63d508e8855b2e61a725c906ea17d1c7f4a2e125 broke test_readonly.
When a local git remote is readonly, tryCopyCoW run to copy a file
from it failed at withOtherTmp.

Sponsored-by: Dartmouth College's Datalad project
4 years agocomment
Joey Hess [Mon, 27 Sep 2021 18:14:52 +0000 (14:14 -0400)]
comment

4 years agocomment
Joey Hess [Mon, 27 Sep 2021 17:59:54 +0000 (13:59 -0400)]
comment

4 years agocomment
Joey Hess [Mon, 27 Sep 2021 16:53:35 +0000 (12:53 -0400)]
comment

4 years agoimprove function name
Joey Hess [Mon, 27 Sep 2021 16:52:30 +0000 (12:52 -0400)]
improve function name

onAddUnlocked suggested that it had to add the file locked, but it
doesn't, it only queues the add of the file, and it's up to the
committer what to do -- which is currently indeed to always add the file
locked.

4 years agocomment
Joey Hess [Mon, 27 Sep 2021 16:35:15 +0000 (12:35 -0400)]
comment

4 years agocomment
Joey Hess [Mon, 27 Sep 2021 16:18:30 +0000 (12:18 -0400)]
comment

4 years agoresponse
Joey Hess [Mon, 27 Sep 2021 16:14:25 +0000 (12:14 -0400)]
response

4 years agoMerge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Sat, 25 Sep 2021 15:17:45 +0000 (11:17 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com

4 years agoAdded a comment
jkniiv [Sat, 25 Sep 2021 06:05:40 +0000 (06:05 +0000)]
Added a comment

4 years agoAdded a comment
jkniiv [Sat, 25 Sep 2021 05:47:16 +0000 (05:47 +0000)]
Added a comment

4 years agocomment
Joey Hess [Fri, 24 Sep 2021 22:33:34 +0000 (18:33 -0400)]
comment