Joey Hess [Tue, 21 Nov 2023 19:57:10 +0000 (15:57 -0400)]
move a comment that is a bug report
Joey Hess [Tue, 21 Nov 2023 16:05:47 +0000 (12:05 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Tue, 21 Nov 2023 15:34:21 +0000 (11:34 -0400)]
improve attribution armoring
Split out an author parameter, will make it easier to add authors and
reads better.
Got rid of the function without the copyright year, because an adversary
could have mechanically changed the function with a copyright year to
the one without, and so bypassed the protection of LLM copyright
year hallucination.
Sponsored-by: Luke T. Shumaker on Patreon
Added a comment
Joey Hess [Tue, 21 Nov 2023 01:34:29 +0000 (21:34 -0400)]
exhaustiveness check fix
Joey Hess [Tue, 21 Nov 2023 01:33:09 +0000 (21:33 -0400)]
golfing
Joey Hess [Tue, 21 Nov 2023 01:29:19 +0000 (21:29 -0400)]
avoid displicate of year range
Joey Hess [Tue, 21 Nov 2023 01:20:37 +0000 (21:20 -0400)]
improve attribution armoring
Joey Hess [Tue, 21 Nov 2023 00:25:51 +0000 (20:25 -0400)]
avoid replacing otherwise
While authorJoeyHess is True same as otherwise, ghc's exhastiveness
checker turns out to special case otherwise. So this avoids warnings.
Joey Hess [Tue, 21 Nov 2023 00:08:35 +0000 (20:08 -0400)]
fix build
Joey Hess [Mon, 20 Nov 2023 16:29:44 +0000 (12:29 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Mon, 20 Nov 2023 16:07:07 +0000 (12:07 -0400)]
make my authorship explicit in the code
This is intended to guard against LLM code theft, which is the current
bubble technology de jour.
Note that authorJoeyHess' with a year older than the year I began
developing git-annex will behave badly, by intention. Eg, it will spin
and eventually crash.
This is not the first anti-LLM protection in git-annex. For example see
9562da790fece82d6dfa756b571c67d0fdf57468. That method, while much harder
for an adversary to detect and remove, also complicates code somewhat
significantly, and needs extensions to be enabled. There are also
probably significantly fewer ways to implement that method in Haskell.
This new approach, by contrast, will be easy to add throughout the code
base, with very little effort, and without complicating reading or
maintaining it any more than noticing that yes, I am the author of this
code.
An adversary could of course remove all calls to these functions
before feeding code into their LLM-based laundry facility. I think this
would need to be done manually, or with the help of some fairly advanced
Haskell parsing though. In some cases, authorJoeyHess needs to be
removed, while in other places it needs to be replaced with a value.
Also a monadic use of authorJoeyHess' may involve other added monadic
machinery which would need to be eliminated to keep the code compiling.
Alternatively, an adversary could replace my name with something
innocuous. This would be clear intent to remove author attribution
from my code, even more than running it through an LLM laundry is.
If you work for a large company that is laundering my code through an
LLM, please do us a favor and use your immense privilege to quit and go
do something socially beneficial. I will not explain further
developments of this code in such detail, and you have better things to
do than playing cat and mouse with me as I explore directions such as
extending this approach to the type level.
Sponsored-by: k0ld on Patreon
NewUser [Mon, 20 Nov 2023 13:17:03 +0000 (13:17 +0000)]
Added a comment
nobodyinperson [Mon, 20 Nov 2023 10:25:52 +0000 (10:25 +0000)]
Added a comment: forgot to add the new ignored link + shortcoming of 'ln' command
NewUser [Mon, 20 Nov 2023 04:24:35 +0000 (04:24 +0000)]
Added a comment: Is `annex.tune.objecthashlower=true` recommended for interop with windows?
NewUser [Mon, 20 Nov 2023 04:22:02 +0000 (04:22 +0000)]
Added a comment: `git annex uninit` fails when I symlink your symlink
NewUser [Mon, 20 Nov 2023 02:09:42 +0000 (02:09 +0000)]
NewUser [Sun, 19 Nov 2023 18:20:03 +0000 (18:20 +0000)]
Added a comment: Thanks Yann!
nobodyinperson [Sun, 19 Nov 2023 16:27:35 +0000 (16:27 +0000)]
Added a comment
nobodyinperson [Sun, 19 Nov 2023 16:26:44 +0000 (16:26 +0000)]
Make it more clear that `git annex webapp` will add files and sync via the assistant.
nobodyinperson [Sun, 19 Nov 2023 16:16:15 +0000 (16:16 +0000)]
Added a comment
NewUser [Sat, 18 Nov 2023 20:08:18 +0000 (20:08 +0000)]
Added a comment: launching `git annex webapp` starts adding to the annex which is surprising
NewUser [Sat, 18 Nov 2023 20:03:06 +0000 (20:03 +0000)]
Added a comment: launching `git annex webapp` starts adding to the annex which is surprising
yarikoptic [Sat, 18 Nov 2023 01:35:35 +0000 (01:35 +0000)]
Added a comment
Joey Hess [Fri, 17 Nov 2023 21:27:58 +0000 (17:27 -0400)]
Make git-annex copy --from --to --fast actually fast
Eg when the destination is logged as containing a file, skip
actively checking that it does contain it.
Note that --fast does not prevent other verifications of content
location that are done in a copy --from --to. Perhaps it could, but this
change will already avoid the real unnecessary work of operating on
files that are already in the remote.
And avoiding other verifications
might cause it to fail if the location log thinks that --to does not
contain the content but does. Such complications with `git-annex copy
--to remote --fast` led to commit
d006586cd0b706c9cc92b2747b2ba3487f52c04a
which added a note that gets displayed when that fails, mentioning it
might be due to --fast being enabled.
copy --from --to is already complicated enough without needing to worry
about such edge cases, so continuing to doing some verification of
content location after the initial --fast filtering seems ok.
Sponsored-by: Dartmouth College's DANDI project
Joey Hess [Fri, 17 Nov 2023 21:10:24 +0000 (17:10 -0400)]
small refactoring
Sponsored-by: Dartmouth College's DANDI project
Joey Hess [Fri, 17 Nov 2023 20:46:43 +0000 (16:46 -0400)]
fixed
Joey Hess [Fri, 17 Nov 2023 20:30:20 +0000 (16:30 -0400)]
Fix bug in git-annex copy --from --to
Caused it to skip files that were locally present.
Sponsored-by: Dartmouth College's DANDI project
Joey Hess [Fri, 17 Nov 2023 20:17:44 +0000 (16:17 -0400)]
comment
nobodyinperson [Fri, 17 Nov 2023 11:47:53 +0000 (11:47 +0000)]
imlew [Fri, 17 Nov 2023 11:43:39 +0000 (11:43 +0000)]
Added a comment: thank you
nobodyinperson [Fri, 17 Nov 2023 09:49:02 +0000 (09:49 +0000)]
Added a comment
imlew [Fri, 17 Nov 2023 09:41:19 +0000 (09:41 +0000)]
Added a comment
nobodyinperson [Fri, 17 Nov 2023 08:45:56 +0000 (08:45 +0000)]
Added a comment
imlew [Fri, 17 Nov 2023 08:23:32 +0000 (08:23 +0000)]
Added a comment: one more question
imlew [Fri, 17 Nov 2023 08:09:23 +0000 (08:09 +0000)]
Added a comment: not about backends after all
yarikoptic [Fri, 17 Nov 2023 02:23:17 +0000 (02:23 +0000)]
remove content of now independent issue https://git-annex.branchable.com/bugs/copy_--from_--to_does_not_copy_if_present_locally/
yarikoptic [Fri, 17 Nov 2023 02:07:35 +0000 (02:07 +0000)]
initial dedicated issue on copy --from --to not copying if present locally
yarikoptic [Thu, 16 Nov 2023 17:54:49 +0000 (17:54 +0000)]
initial report on --fast of being no effect for copy --from --to
nobodyinperson [Thu, 16 Nov 2023 16:34:13 +0000 (16:34 +0000)]
Added a comment
nobodyinperson [Thu, 16 Nov 2023 15:55:49 +0000 (15:55 +0000)]
Added a comment
nobodyinperson [Thu, 16 Nov 2023 15:54:08 +0000 (15:54 +0000)]
Added a comment
imlew [Thu, 16 Nov 2023 12:51:59 +0000 (12:51 +0000)]
Added a comment: I guess this question is about backends
Ilya_Shlyakhter [Wed, 15 Nov 2023 18:45:52 +0000 (18:45 +0000)]
added bug report for git init on a worktree checked out for a submodule.
imlew [Wed, 15 Nov 2023 16:36:13 +0000 (16:36 +0000)]
Added a comment
jcjgraf [Tue, 14 Nov 2023 21:25:24 +0000 (21:25 +0000)]
Added a comment
Joey Hess [Tue, 14 Nov 2023 19:02:06 +0000 (15:02 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Tue, 14 Nov 2023 18:56:58 +0000 (14:56 -0400)]
git-annex log --gnuplot
The gnuplot output is pretty good, but could still be improved with:
* more colors (repeating colors is confusing with a lot of repos)
* better positioning of the legend, making the plot wider and moving it
from over top of the graph
Sponsored-by: Kevin Mueller on Patreon
Joey Hess [Tue, 14 Nov 2023 18:00:06 +0000 (14:00 -0400)]
git-annex log --received modifier option
Only counting received and not dropped makes this show the bandwidth of
data coming into the repository, although only in a sense. Since
git-annex branch updates only happen at the end of a command, and we
don't know when a command started, it's only an approximation of the
actual bandwidth. (A previous git-annex branch update made have
happened in a different repository.)
It would be possible to also add a --dropped option, but I don't know
how useful that would be?
Sponsored-by: Nicholas Golder-Manning on Patreon
Joey Hess [Tue, 14 Nov 2023 15:45:16 +0000 (11:45 -0400)]
rename --when to --interval
More accurately describes its behavior.
grond66@79ca29ba964cd0d8e2f352871d54452e4a9dad88 [Tue, 14 Nov 2023 04:32:15 +0000 (04:32 +0000)]
Joey Hess [Mon, 13 Nov 2023 21:45:37 +0000 (17:45 -0400)]
close
Joey Hess [Mon, 13 Nov 2023 19:06:04 +0000 (15:06 -0400)]
guard against wrong timestamps in git log
For example, my sound repo has in the git-annex branch a commit from
2036, which is followed by one from 2034, in amoung commits from 2013.
Clearly there was a problem with the clock.
Since git log --date-order has a behavior of
"Show no parents before all of its children are shown", the data still
gets processed ok. The future timestamp just prevented displaying data
after that commit. It seems better, when the clock was wrong, to display
a wrong date, and then return to right dates.
It would be nice to filter out the wrong dates from display entirely,
but that seems it would need to buffer the whole output. This command is
too slow to buffer it all before displaying anything, and anyway this
kind of problem is probably rare.
Sponsored-by: Joshua Antonishen on Patreon
Joey Hess [Mon, 13 Nov 2023 18:46:51 +0000 (14:46 -0400)]
treat dead repos as 0 size
With this, git annex log --totalsizes can be compared with
git-annex info's "combined annex size of all repositories"
to double-check it works correctly.
In my sound repo, the two match.
In my big repo, the two report slightly different sizes,
with the former being 1.3 gb smaller than the latter.
I don't know the reason for this disreprency. Given the 30+tb size of
the repo, it's a small difference.
It seems possible that a bug in an old version of git-annex could
explain it. Eg, if an old git-annex lost a line when updating trust.log
or a location log in a merge, git-annex info would see only what it
replaced it with, while git-annex log will see the previous value as
well.
Sponsored-by: Leon Schuermann on Patreon
Joey Hess [Mon, 13 Nov 2023 18:16:55 +0000 (14:16 -0400)]
newtype MapLog
Noticed that Semigroup instance of Map is not suitable to use
for MapLog. For example, it behaved like this:
ghci> parseTrustLog "foo 1 timestamp=10\nfoo 2 timestamp=11" <> parseTrustLog "foo X timestamp=12"
fromList [(UUID "foo",LogEntry {changed = VectorClock 11s, value = SemiTrusted})]
Which was wrong, it lost the newer DeadTrusted value.
Luckily, nothing used that Semigroup when operating on a MapLog. And this
provides a safe instance.
Sponsored-by: Graham Spencer on Patreon
Joey Hess [Mon, 13 Nov 2023 17:15:07 +0000 (13:15 -0400)]
git-annex log --totalsizes
Note that dead repositories are not yet handled so their sizes show as
nonzero after they are marked dead.
Sponsored-By: unqueued on Patreon
Joey Hess [Mon, 13 Nov 2023 17:09:39 +0000 (13:09 -0400)]
update
Joey Hess [Mon, 13 Nov 2023 17:07:22 +0000 (13:07 -0400)]
git-annex log --sizes
CSV format so it can be fed into a program to graph it.
Note that dead repositories are not yet handled so their sizes show as
nonzero after they are marked dead.
Sponsored-By: k0ld on Patreon
Joey Hess [Sat, 11 Nov 2023 18:43:13 +0000 (14:43 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Lukey [Sat, 11 Nov 2023 16:13:28 +0000 (16:13 +0000)]
Added a comment
Joey Hess [Fri, 10 Nov 2023 21:37:55 +0000 (17:37 -0400)]
the last line is for the current time
Joey Hess [Fri, 10 Nov 2023 20:17:15 +0000 (16:17 -0400)]
git-annex log --sizesof
This can take a lot of memory. I decided to violate the usual rule in
git-annex that it operate in constant memory no matter how many annexed
objects. In this case, it would be hard to be fast without using a big
map of the location logs. The main difficulty here is that there can be
many git-annex branches and it needs to display a consistent view at a
point in time, which means merging information from multiple git-annex
branches.
I have not checked if there are any laziness leaks in this code. It
takes 1 gb to run in my big repo, which is around what I estimated
before writing it.
2 options that are documented are not yet implemented.
Small bug: With eg --when=1h, it will display at 12:00 then 1:10 if the
next change after 12:59 is then. Then it waits until after 2:10 to
display the next change. It ought to wait until after 2:00.
Sponsored-by: Brock Spratlen on Patreon
Joey Hess [Fri, 10 Nov 2023 16:37:01 +0000 (12:37 -0400)]
split out generic git log parser
Sponsored-By: Jack Hill on Patreon
Joey Hess [Wed, 8 Nov 2023 18:14:41 +0000 (14:14 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Wed, 8 Nov 2023 18:14:35 +0000 (14:14 -0400)]
todo
Joey Hess [Wed, 8 Nov 2023 17:15:00 +0000 (13:15 -0400)]
info: Added calculation of combined annex size of all repositories
Factored out overLocationLogs from CmdLine.Seek, which can calculate this
pretty fast even in a large repo. In my big repo, the time to run git-annex
info went up from 1.33s to 8.5s.
Note that the "backend usage" stats are for annexed files in the working
tree only, not all annexed files. This new data source would let that be
changed, but that would be a confusing behavior change. And I cannot
retitle it either, out of fear something uses the current title (eg parsing
the json).
Also note that, while time says "402108maxresident" in my big repo now,
up from "54092maxresident", top shows the RES constant at 64mb, and it
was 48mb before. So I don't think there is a memory leak. I tried using
deepseq to force full evaluation of addKeyCopies and memory use didn't
change, which also says no memory leak. And indeed, not even calling
addKeyCopies resulted in the same memory use. Probably the increased memory
usage is buffering the stream of data from git in overLocationLogs.
Sponsored-by: Brett Eisenberg on Patreon
Joey Hess [Wed, 8 Nov 2023 16:06:03 +0000 (12:06 -0400)]
improve comments
jcjgraf [Tue, 7 Nov 2023 20:42:11 +0000 (20:42 +0000)]
mih [Tue, 7 Nov 2023 15:49:47 +0000 (15:49 +0000)]
Added a comment: What about temporary annex.private declaration?
oadams [Tue, 7 Nov 2023 08:30:21 +0000 (08:30 +0000)]
Added a comment
oadams [Tue, 7 Nov 2023 08:26:44 +0000 (08:26 +0000)]
Joey Hess [Mon, 6 Nov 2023 15:11:50 +0000 (11:11 -0400)]
expand description
unqueued [Sun, 5 Nov 2023 21:32:17 +0000 (21:32 +0000)]
Added a comment
nobodyinperson [Sun, 5 Nov 2023 13:58:25 +0000 (13:58 +0000)]
report
Joey Hess [Thu, 2 Nov 2023 17:35:54 +0000 (13:35 -0400)]
windows hook scripts newlines without CR
Windows: When git-annex init is installing hook scripts, it will
avoid ending lines with CR for portability.
Existing hook scripts that do have CR line endings will not be changed.
While it would be possible to have git-annex init upgrade them, users would
need to know to use that command to do that, and it would add complexity
that does not seem warranted for the portability benefit alone.
Sponsored-by: Luke T. Shumaker on Patreon
Joey Hess [Wed, 1 Nov 2023 18:27:22 +0000 (14:27 -0400)]
convert StorableCipher to ByteString
This allows getting rid of the ugly and error prone handling of
"bag of bytes" String in Remote.Helper.Encryptable.
Avoiding breakage like that dealt with by commit
9862d64bf90de645ef1acfbf862c5b340475aacf
And allows converting Utility.Gpg to use ByteString for IO, which is
a welcome change.
Tested the new git-annex interoperability with old, using all 3
encryption= types.
Sponsored-By: the NIH-funded NICEMAN (ReproNim TR&D3) project
Joey Hess [Wed, 1 Nov 2023 17:14:39 +0000 (13:14 -0400)]
remove unused import
Joey Hess [Wed, 1 Nov 2023 17:09:42 +0000 (13:09 -0400)]
bring back "bag of bytes" handling for ciphers
Fixes test suite failure with LANG=C caused by commit
3742263c99180d1391e4fd51724aae52d6d02137
Sponsored-By: the NIH-funded NICEMAN (ReproNim TR&D3) project
Joey Hess [Wed, 1 Nov 2023 16:19:50 +0000 (12:19 -0400)]
reproduced
Joey Hess [Wed, 1 Nov 2023 16:03:55 +0000 (12:03 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Wed, 1 Nov 2023 12:43:32 +0000 (08:43 -0400)]
update
Joey Hess [Wed, 1 Nov 2023 12:42:15 +0000 (08:42 -0400)]
update
Joey Hess [Tue, 31 Oct 2023 18:06:46 +0000 (14:06 -0400)]
update
yarikoptic [Mon, 30 Oct 2023 22:03:20 +0000 (22:03 +0000)]
initial bug report
jkniiv [Mon, 30 Oct 2023 21:32:07 +0000 (21:32 +0000)]
Added a comment
Joey Hess [Mon, 30 Oct 2023 18:59:09 +0000 (14:59 -0400)]
lookupkey: Sped up --batch
When the file is relative, it does not need to be passed
through git lsfiles to normalize it.
Sponsored-by: Kevin Mueller on Patreon
Joey Hess [Mon, 30 Oct 2023 18:45:58 +0000 (14:45 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Mon, 30 Oct 2023 18:45:52 +0000 (14:45 -0400)]
comment
Joey Hess [Mon, 30 Oct 2023 18:43:43 +0000 (14:43 -0400)]
Windows: Consistently avoid ending output lines with CR
This matches the behavior of git on Windows, which does not end lines with
CR either.
Previously, git-annex used to always write lines with putStrLn, so would
output CR on Windows. Then parts of it changed to use ByteString.putStrLn,
which does not output CR. That left its output inconsistent, sometimes
within the same command.
The point of this commit is to get back to consistency. Having the same
behavior as git is a nice bonus. It would be much harder to make it
consistently output CR, because every place it uses ByteString.putStrLn or
similar would need to be changed.
Sponsored-by: Nicholas Golder-Manning on Patreon
Joey Hess [Mon, 30 Oct 2023 18:23:23 +0000 (14:23 -0400)]
Windows: Fix CRLF handling in some log files
In particular, the mergedrefs file was written with CR added to each line,
but read without CRLF handling. This resulted in each update of the file
adding CR to each line in it, growing the number of lines, while also
preventing the optimisation from working, so it remerged unncessarily.
writeFile and readFile do NewlineMode translation on Windows. But the
ByteString conversion prevented that from happening any longer.
I've audited for other cases of this, and found three more
(.git/annex/index.lck, .git/annex/ignoredrefs, and .git/annex/import/). All
of those also only prevent optimisations from working. Some other files are
currently both read and written with ByteString, but old git-annex may have
written them with NewlineMode translation. Other files are at risk for
breakage later if the reader gets converted to ByteString.
This is a minimal fix, but should be enough, as long as I remember to use
fileLines when splitting a ByteString into lines. This leaves files written
using ByteString without CR added, but that's ok because old git-annex has
no difficulty reading such files.
When the mergedrefs file has gotten lines that end with "\r\r\r\n", this
will eventually clean it up. Each update will remove a single trailing CR.
Note that S8.lines is still used in eg Command.Unused, where it is parsing
git show-ref, and similar in Git/*. git commands don't include CR in their
output so that's ok.
Sponsored-by: Joshua Antonishen on Patreon
Joey Hess [Mon, 30 Oct 2023 18:10:46 +0000 (14:10 -0400)]
add PackageImports
This makes loading it in ghci work when both crypton and cryptonite are
installed.
mih [Mon, 30 Oct 2023 07:45:05 +0000 (07:45 +0000)]
Report on a possibly unknown slowness of `lookupkey`
Joey Hess [Fri, 27 Oct 2023 21:47:50 +0000 (17:47 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Fri, 27 Oct 2023 21:47:23 +0000 (17:47 -0400)]
remove wrongly nested comment
This happened to prevent checkout on windows, due to filename too long.
jkniiv [Fri, 27 Oct 2023 18:45:03 +0000 (18:45 +0000)]
Added a comment
Joey Hess [Thu, 26 Oct 2023 19:22:22 +0000 (15:22 -0400)]
comment
Joey Hess [Thu, 26 Oct 2023 18:23:08 +0000 (14:23 -0400)]
Merge branch 'master' of ssh://git-annex.branchable.com
Joey Hess [Thu, 26 Oct 2023 18:19:41 +0000 (14:19 -0400)]
display summary after concurrent output is complete
Otherwise, it could have buffered output that was displayed after the
summary. I started seeing that on my 12 core laptop, probably due to the
increased amount of concurrency.
Also seems possible, that, since the summary did exit, it might have
prevented displaying buffered output sometimes, or had other bad
results.