git-lfs support #10153

b-camacho · 2024-03-04T22:39:38Z

Motivation

nix fetches git repos using libgit2, which does not run filters by default. This means LFS-enabled repos can be fetched, but LFS pointer files are not smudged.

This change adds a lfs attribute to fetcher URLs. With lfs=1, when fetching LFS-enabled repos, nix will smudge all the files.

Context

See #10079.
Git Large File Storage lets you track large files directly in git, using git filters. A clean filter runs on your LFS-enrolled files before push, replacing large files with small "pointer files". Upon checkout, a "smudge" filter replaces pointer files with full file contents. When this works correctly, it is not visible to users, which is nice.

Changes

builtins.fetchGit has new bool lfs attr
when lfs=true, GitSourceAccessor will smudge any pointer files with the lfs filter attribute
as verified by new test in tests/nixos/fetchgit (this is why lfs is now enabled on the test gitea instance)

Priorities and Process

Add 👍 to pull requests you find important.

The Nix maintainer team uses a GitHub project board to schedule and track reviews.

b-camacho · 2024-03-05T22:18:27Z

Small complication:
it seems that nix flake lock calls fetch which in turn calls Input::fetch -> InputScheme::fetch -> fetchToStore.

In other words, I don't currently see a way to:

materialize LFS files when fetching -source store paths, but
don't materialize LFS files during nix flake lock.

Using git-lfs, when the flake copies the repo to the store (for purity) the 'virtual file' stored in git is copied (with oid/size info of the object in LFS) instead of the actual (large) file :/ ref: NixOS/nix#10153 I think it was working before because the file was in git temporarily at some point, then I moved it to LFS, but after the system was built.. (or something like that 🤷)

L-as · 2024-03-09T11:42:43Z

What use case do you have in mind? Isn't LFS typically for large files, that wouldn't usually affect evaluation anyway?

b-camacho · 2024-03-11T16:47:38Z

What use case do you have in mind? Isn't LFS typically for large files, that wouldn't usually affect evaluation anyway?

builtins.fetchGit populates the nix store with a <hash>-store path. This path is used as the source when building a derivation. Currently, the builder will see the unsmudged LFS pointer files, but I'd like the builder to optionally see smudged files. I agree that smudged files are usually not needed at eval time, but I don't see a good alternative of making them available at build time, besides a fixed-output derivation (but fixed-output derivations have their own problems)

L-as · 2024-03-14T14:04:16Z

A FOD seems optimal here, in general you shouldn't use builtins.fetchGit if you're only going to use it at build time.

b-camacho · 2024-03-14T23:18:10Z

In general I agree, but (afaik) other fetchers can't use git credentials.

nixos-discourse · 2024-04-08T15:27:27Z

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/2024-03-11-nix-team-meeting-132/42960/1

khoitd1997 · 2024-06-22T02:07:05Z

@roberth have you had a chance to take a look at this issue? We have been staying at older versions of Nix as a workaround but newer versions now have fixes for critical issues so sticking to old ones would no longer be optimal.

roberth · 2024-06-22T10:03:40Z

Hi @b-camacho, thanks for the ping and sorry for the delay. This PR was assigned to me, but I hadn't prioritized it because it was a draft. Wrong assumption on my end, because I do think this is valuable, and I have some things to say :)

when fetching LFS-enabled repos, nix will smudge all the files.

That's a good start, but we need to make sure that the smudging happens in a controlled manner; otherwise we risk adding impurities.

Specifically, we should parse the attribute to check that they're supposed to be unsmudged by lfs; if not, ignore the smudge rule. It seems you were already investigating how this could be implemented.

Furthermore, we should validate the sha256 so that we don't increase the potential for silent errors by a whole external program. The hash should be easy to parse from the pointer file, and while reading other programs' inputs is a little ad hoc, I don't expect any serious issues from this, as we won't cause users to accidentally rely on a bug this way.

the LFS-tracked files are materialized during nix flake lock - this is bad

This won't happen unnecessarily either of these are implemented

If we need to backtrack on the removal of narHashes (#6530), we can also avoid re-locking transitive inputs whose lock has already been computed by the dependency's lock.

Reuse input lock files #7730

So yes, this isn't efficient yet, but it will be.

A FOD seems optimal here, in general you shouldn't use builtins.fetchGit if you're only going to use it at build time.

A fixed output derivation works best when all you're using it for is as an input to another derivation (and it's publicly available, as mentioned).
However, if the "fetched" source is a flake (e.g. you have a flake.nix in a repo with LFS files), then you also need to evaluate files from the fetched source, which would constitute import from derivation, which is not optimal. Furthermore you'd need to produce fixed-output hashes for your local repo files, which is such horrible UX we don't need to consider it as a solution.

To summarize, this is worth implementing, I see no blocking issues, design or otherwise, and the following needs to be done:

lfs attribute with default false, LGTM
figure out which files are LFS
invoke the Git LFS filter from $PATH; no need for a rigid dependency or makeWrapper
check the sha256
add a test, perhaps extending tests/nixos/fetch-git
documentation for the lfs attribute (currently under fetchGit's entry in doc/manual/src/language/builtins.md); mention the runtime dependency on the LFS package.

kip93 · 2024-07-23T12:33:19Z

What's the state on this PR? Seems to unfortunately be a bit stale given the delayed review. This issue has been plaguing us for a while, so I'm willing to pick up the torch here and try to get this out the door (was actually starting to see how to fix this myself back in March when I saw this PR and decided to see what came out of this).

roberth · 2024-07-23T12:59:14Z

@kip93 I think your question was directed towards @b-camacho, but I'd like to add that we would welcome and support anyone who'd like to work on this.

Feel free to ask questions here or in the meetings if you can make them. We generally have some agenda, but we also like to make time for contributors during or after, when we often hang out while we get some things done. Link to the video conference is in the scratchpad linked there. We also have a matrix room, although personally I'm guilty of neglecting that one sometimes.

b-camacho · 2024-07-24T07:15:06Z

Thanks for the thorough writeup @roberth !
I owe you all an update. To avoid shelling out and implement some features we need to merge, I need a subset of a git-lft C/C++ client. I reimplemented one from Python into C++ here https://github.com/b-camacho/git-lfs-fetch-cpp.

Once I add some tests and integrate git-lfs-fetch-cpp here, we should be ready for another review!

I'm still on vacation with not-great internet, but back in 6 days and will update you all on 7/31 regardless.

Thanks for the feedback and sorry for the wait!

roberth · 2024-07-24T12:07:02Z

Oh, I don't think shelling out was such a big deal because we can verify the correctness of the result, kind of like how fixed output derivations are allowed to do "grossly impure" things because we can verify the output.

I guess a library implementation of it is still nice for a consistent UX with a small closure size though.

kip93 · 2024-10-10T15:11:10Z

Hey! It's me again! I just want to ask if there's anything I can help with here. Maybe I can try and doing some testing, or do a smaller version of this that uses the git-lfs CLI tools while the full implementation gets done?

We have a lot of repos with LFS files that would greatly benefit from this, so I'm willing to do whatever work is needed, but also don't want to add extra work for others where it's not wanted.

b-camacho · 2024-11-28T20:16:35Z

I'm happy to hop back in! Can you tell me more about the non-zero padding errors? Just a dump of your terminal output would be a lot to go off of, I don't think I ran into that one before.

kip93 · 2024-11-29T12:35:56Z

There's a test I've already added which currently fails with this error and should explain how to reproduce, and here's the output:

vm-test-run-fetch-git> client # error:
vm-test-run-fetch-git> client #        … while updating the lock file of flake 'path:/build/tmpvffq_lct?lastModified=1732571537&narHash=sha256-/lQnNQ7IGCkZTsV32qCzPYSWyD1x44pl1DyaxnxR%2Blg%3D'
vm-test-run-fetch-git> client # 
vm-test-run-fetch-git> client #        … while updating the flake input 'foo'
vm-test-run-fetch-git> client # 
vm-test-run-fetch-git> client #        … while fetching the input 'git+http://gitea:3000/test/lfs?lfs=1&ref=main&rev=56afd59511b955d8e589513b58dde01cf287a364'
vm-test-run-fetch-git> client # 
vm-test-run-fetch-git> client #        error: non-zero padding

b-camacho · 2024-12-02T08:14:30Z

Thanks for the repro! Pulled on a couple different threads, almost positive this sizeCallback invocation is the culprit, it seems to underreport size for files that use lfs. I can confirm and fix tomorrow.

kip93 · 2024-12-02T12:19:36Z

Oh that sounds very likely! If needed, I expect I'll have some more time come Wednesday.

b-camacho · 2024-12-02T20:08:47Z

I was right about the size callback, the nixos test passes again!
Remerged master, would love to get this in before we fall too far behind again.
Last thing I can think of adding is fixes for fetching over ssh, but I agree with you it may be better to handle those in a followup.

kip93 · 2024-12-02T20:26:02Z

Nice! I would do some testing on wednesday and if http works fine then I'll call this PR mergeable from my side (still needs a review from someone else of course). I do want to get the ssh part working but won't block this PR just for that.

kip93 · 2024-12-04T15:44:55Z

I did some quick tests, validated that I can pull via HTTPS from both github and gitlab, and since our tests also run on gitea, that's 3 different services working as expected. SSH is not working, but now that the rest looks like it works, I'll give up the tiny hope that this would get resolved in this PR.

b-camacho · 2024-12-05T21:33:29Z

Hm, just re-checked I can pull from GitHub and GitLab over SSH from my machine, this gist has the commands and outputs I ran. Current implementation intentionally uses the system ssh binary.

Do you have any more details on the ssh failures? You probably took this into account, but asking for completeness: even publicly available repos do not allow ssh access without auth, so you have to try this on a repo where you have owner/collaborator access on a privkey in your keychain.

Alas, while I don't exactly enjoy chasing subtle git{ea,hub,lab} impl differences, I'll try to set up an automated test for ssh fetching while we wait for a review.

kip93 · 2024-12-06T04:01:29Z

Not sure what happened before (I was a bit burnt out so it might just have been a silly mistake) but now it's working fine, both public and private repos on github and gitlab work ok 🎉

Sorry for the false alarm 🙈

jsqu4re · 2024-12-09T14:01:28Z

Thank you so much for your efforts! I tried this PR and it works well for gitlab.com. Is there a chance to also allow for local remote repos using url = "file:///path/to/repo/.git"? Currently there is this error:

ssh: Could not resolve hostname file: Name or service not known
error: git-lfs-authenticate: no output (cmd: ssh file git-lfs-authenticate ///path/to/repo/.git download)

kip93 · 2024-12-09T14:08:10Z

I'd say this is a nice to have but probably not worth delaying this PR

A follow up PR would be more appropriate.

roberth

Reviewed a bit. It seems that the gitattributes support is incomplete, but could perhaps be reused from libgit2.

Git consults [...] the same directory as the path in question, and its parent directories up to the toplevel of the work tree

man gitattributes

see comments

roberth · 2024-12-09T13:25:28Z

src/libfetchers/git-utils.cc

+        if (lfsFetch && !lookup(CanonPath(".gitattributes"))) {
+            warn("Requested to fetch lfs files, but no .gitattributes file was found, ignoring");
+        }


It's possible for these to be in subdirectories.
I'd be ok with not checking this, because checking the whole tree may be slow.

I think cloning with LFS is quite slow as it is, adding a few extra seconds at most to recurse the entire tree is probably not very noticeable

src/libfetchers/git-lfs-fetch.hh

roberth · 2024-12-09T13:38:08Z

src/libfetchers/git-lfs-fetch.hh

+    std::istringstream iss(content_str);
+    std::string line;
+
+    while (std::getline(iss, line)) {


Shouldn't libgit2 handle this for us?

Don't know if libgit2 has lfs support, but the syntax was simple enough that we just wrote a quick parser. I'll take another look at libgit2's provided tools next week to see if I missed anything

git attributes exist at the git layer (as opposed to lfs), you can read them with libgit2's git_attr_get (the ExportIgnore fetcher does this), but I could never get git_attr_get to work, but I'll give it another shot tonight.

src/libfetchers/git-utils.cc

roberth · 2024-12-09T13:48:40Z

tests/nixos/fetch-git/test-cases/lfs/default.nix

+
+    # check that file was smudged
+    file_size_lfs = client.succeed(f"stat -c %s {fetched_lfs}/beeg").strip()
+    assert int(file_size_lfs) >= expected_max_size_lfs_pointer, \


We should know the exact size based on the dd command; would be better to check with ==.

This was a conceptual test, since a pointer cannot be >= 1024 by definition, can definitely also check for the full 1MiB size. Will rewrite this with the other test changes when I have the time

roberth · 2024-12-09T13:55:49Z

tests/nixos/fetch-git/test-cases/lfs/default.nix

+    # Request lfs fetch without any .gitattributes file
+    ############################################################################


This seems to be the first test case; a base case that doesn't actually use the lfs feature.
The subtest feature would be nice for this. It doesn't do much, but other NixOS tests use it for this purpose, and I think it's nice to mark structure with indentation as well.

Suggested change

# Request lfs fetch without any .gitattributes file

############################################################################

with subtest("Request lfs fetch without any .gitattributes file"):

< ... indented code ... >

Didn't know there was a subtest feature, will rewrite these as soon as I have time (probably next week :/ )

tests/nixos/fetch-git/test-cases/lfs/default.nix

roberth · 2024-12-09T14:05:17Z

tests/nixos/fetch-git/test-cases/lfs/default.nix

+    # Use as flake input
+    ############################################################################
+    with TemporaryDirectory() as tempdir:
+      client.succeed(f"mkdir -p {tempdir}")
+      client.succeed(f"""
+        printf '{{
+          inputs = {{
+            foo = {{
+              url = "git+{repo.remote}?ref=main&rev={lfs_file_rev}&lfs=1";
+              flake = false;
+            }};
+          }};
+          outputs = {{ foo, self }}: {{ inherit (foo) outPath; }};
+        }}' >{tempdir}/flake.nix
+      """)
+      fetched_flake = client.succeed(f"""
+        nix eval --raw {tempdir}#.outPath
+      """)
+
+    assert fetched_lfs == fetched_flake, \
+      f"fetching as flake input (store path {fetched_flake}) yielded a different result than using fetchGit (store path {fetched_lfs})"


This test case seems redundant to me.

This was added because the flake fetch does some things on top of what fetchGit does, and it was failing at some point. So while it may look redundant, it apparently does have a subtle difference.

roberth · 2024-12-09T14:14:09Z

src/libfetchers/git-lfs-fetch.hh

+}
+
+bool Fetch::hasAttribute(const std::string & path, const std::string & attrName, const std::string & attrValue) const
+{


Do we have a compelling reason not to use git_attr_get_ext?

Didn't implement this part, @b-camacho may have more context, but I could try my hand at using this instead to see if it still works.

Not a satisfying explanation but I could not get it to work without materializing the files on disk, but I'll try again

Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>

b-camacho · 2024-12-09T20:07:10Z

Thank you so much for your efforts! I tried this PR and it works well for gitlab.com. Is there a chance to also allow for local remote repos using url = "file:///path/to/repo/.git"? Currently there is this error:
ssh: Could not resolve hostname file: Name or service not known
error: git-lfs-authenticate: no output (cmd: ssh file git-lfs-authenticate ///path/to/repo/.git download)

I'd be happy to add this, but I agree this PR is already pretty big so we should save it for a followup.

jsqu4re · 2024-12-09T20:15:06Z

I'd be happy to add this, but I agree this PR is already pretty big so we should save it for a followup.

Makes total sense, and thanks for tackling this task!

nixos-discourse · 2024-12-11T21:21:36Z

This pull request has been mentioned on NixOS Discourse. There might be relevant details there:

https://discourse.nixos.org/t/2024-12-09-nix-team-meeting-minutes-201/57280/1

CHN-beta · 2024-12-12T01:09:33Z

Thank you so much for your efforts! I tried this PR and it works well for gitlab.com. Is there a chance to also allow for local remote repos using url = "file:///path/to/repo/.git"? Currently there is this error:
ssh: Could not resolve hostname file: Name or service not known
error: git-lfs-authenticate: no output (cmd: ssh file git-lfs-authenticate ///path/to/repo/.git download)
I'd be happy to add this, but I agree this PR is already pretty big so we should save it for a followup.

Seems git LFS is only supported over https now. Enabling other transfers (e.g. ssh) needing NixOS/nixpkgs#350130. Please let me know if I am wrong.

kip93 · 2024-12-16T15:03:03Z

Today I'll try to rewrite all tests, but wanted to let you know that the guys from libgit2 actually replied to my issue, and in there I've found out that this function we're using (git_pathspec_matches_path) is not the right one (and there's currently no right function for this job) since a pathspec is not the same as a path pattern. So I'll try to use the git_attr_get_ext again and see if that works for our purposes. This should also fix issues with negated patterns that I've noticed today.

b-camacho · 2024-12-16T15:04:28Z

Oh I had no idea they were different, that explains a lot!

kip93 · 2024-12-18T17:30:59Z

Ok I've reintroduced the git_attr_get_ext and have it working. I'm currently testing that it does not import files into the store too early like it used to be the case the first time around.

Also rewrote the tests to use subtests as suggested by the code review.

kip93 · 2024-12-19T14:33:17Z

With my quick-ish tests (with several large lfs repos which take a long time to clone) it all looks ok? I think it's ready for another review.

Given that holidays are around the corner, I won't have any more time to work on this 'till after new years (of course, I'm also not expecting you to skip xmas to review this, I'm fine if the review is delayed as well). If anything needs fixing I can take this back up again probably 6-ish of january.

edolstra · 2024-12-19T16:01:32Z

src/libfetchers/git-utils.cc

+                try {
+                    _lfsFetch.fetch(blob.get(), pathStr, s, [&s](uint64_t size){ s.s.reserve(size); });
+                } catch (Error &e) {
+                    e.addTrace({}, "while smudging git-lfs file '%s' (std::string interface)", pathStr);


Suggested change

e.addTrace({}, "while smudging git-lfs file '%s' (std::string interface)", pathStr);

e.addTrace({}, "while smudging git-lfs file '%s'", path);

edolstra · 2024-12-19T16:02:54Z

src/libfetchers/git-utils.cc

+        const CanonPath & path,
+        Sink & sink,
+        std::function<void(uint64_t)> sizeCallback = [](uint64_t size){}) override {
+        auto blob = getBlob(path, false);


Do we need nearly identical readBlob() and readFile() functions? SourceAccessor has a default implementation for the string version of readFile() that wraps the streaming version.

edolstra · 2024-12-19T16:03:02Z

src/libfetchers/git-utils.cc

+    void readFile(
+        const CanonPath & path,
+        Sink & sink,
+        std::function<void(uint64_t)> sizeCallback = [](uint64_t size){}) override {


Suggested change

std::function<void(uint64_t)> sizeCallback = [](uint64_t size){}) override {

std::function<void(uint64_t)> sizeCallback = [](uint64_t size){}) override

{

edolstra · 2024-12-19T16:04:27Z

src/libfetchers/git-utils.cc

+            if (_lfsFetch.shouldFetch(pathStr)) {
+                StringSink s;
+                try {
+                    _lfsFetch.fetch(blob.get(), pathStr, s, [&s](uint64_t size){ s.s.reserve(size); });


Not sure why _lfsFetch is needed? Can't you do

Suggested change

if (_lfsFetch.shouldFetch(pathStr)) {

StringSink s;

try {

_lfsFetch.fetch(blob.get(), pathStr, s, [&s](uint64_t size){ s.s.reserve(size); });

if (lfsFetch->shouldFetch(pathStr)) {

StringSink s;

try {

lfsFetch->fetch(blob.get(), pathStr, s, [&s](uint64_t size){ s.s.reserve(size); });

I was mimicking if let Some(v) = v {}, but in C++ pointer-like syntax is more idiomatic for optionals

edolstra · 2024-12-19T17:36:24Z

src/libfetchers/git-lfs-fetch.hh

+
+// if authHeader is "", downloadToSink assumes no auth is expected
+void downloadToSink(
+    const std::string & url, const std::string & authHeader, Sink & sink, std::string_view sha256Expected)


This should use FileTransfer::download() (the sink variant). That way libfetcher doesn't need to reimplement curl support, and we get stuff like connection reuse.

edolstra · 2024-12-19T17:41:37Z

src/libfetchers/git-lfs-fetch.hh

+    size_t size;      // in bytes
+};
+
+struct GitUrl


This class looks very similar to ParsedUrl, maybe you can use that instead (in conjuction with fixGitURL() to handle the scp syntax)?

edolstra · 2024-12-19T17:44:13Z

src/libfetchers/git-lfs-fetch.hh

+    return jArray;
+}
+
+std::vector<nlohmann::json> Fetch::fetchUrls(const std::vector<Md> & metadatas) const


Idem, this should use FileTransfer::download().

edolstra · 2024-12-19T17:50:38Z

src/libfetchers/git-lfs-fetch.hh

+}
+
+void Fetch::fetch(
+    const git_blob * pointerBlob,


This function could be simplified by having the caller read the blob and pass it to this function, i.e.

Suggested change

const git_blob * pointerBlob,

std::string contents,

since the caller needs to be able to do that anyway if LFS is not enabled. So the caller becomes something like:

auto contents = std::string((const char *) git_blob_rawcontent(blob.get()), git_blob_rawsize(blob.get())); lfsFetch->fetch(contents, ..., sink, ...);

That way all the blob-reading code here could be removed.

edolstra · 2024-12-19T17:56:43Z

src/libfetchers/git-lfs-fetch.hh

+        }
+        const uint64_t size = obj.at("size");
+        sizeCallback(size);
+        downloadToSink(ourl, authHeader, sink, oid); // oid is also the sha256


Should these downloads be cached? If I understand this correctly, Nix will re-download the same LFS objects on every invocation unless the entire tree has been copied into the store (which generally won't be the case with lazy trees).

We have downloadFile() in tarball.cc that can cache files in the store to avoid repeated downloads.

naive lfs support

8fb36a9

github-actions bot added the fetching Networking with the outside (non-Nix) world, input locking label Mar 4, 2024

bcamacho2 and others added 3 commits March 11, 2024 10:13

parse out lfs url attr correctly

9ef1c28

todo: actual smudge

f8bc96a

todo: fix git_attr_foreach

8c6641e

thufschmitt assigned roberth Apr 8, 2024

This comment was marked as off-topic.

Sign in to view

b-camacho added 3 commits August 6, 2024 17:30

Merge branch 'master' into lfs

cec370e

Merge branch 'master' into lfs

787cc04

wip

ba417a2

roberth mentioned this pull request Oct 1, 2024

Support for Git LFS in private repositories #4623

Open

add libcurl to deps, builds now

87e0bc9

b-camacho added 3 commits December 2, 2024 14:52

trim storepath in test

85d6efb

fix sizeCallback

79a6438

Merge branch 'master' into lfs

169d62a

kip93 added 2 commits December 4, 2024 16:39

Fix flake path in test

9a7b14c

Fix format

65f78f7

roberth reviewed Dec 9, 2024

View reviewed changes

Apply suggestions from code review

140b34b

Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>

kip93 added 3 commits December 18, 2024 18:00

Re-introduce git_attr_get_ext

b1663fa

Rework tests

726f8fd

Merge remote-tracking branch 'upstream/master' into lfs

7756b22

edolstra requested changes Dec 19, 2024

View reviewed changes

		# Request lfs fetch without any .gitattributes file
		############################################################################

	e.addTrace({}, "while smudging git-lfs file '%s' (std::string interface)", pathStr);
	e.addTrace({}, "while smudging git-lfs file '%s'", path);

	std::function<void(uint64_t)> sizeCallback = [](uint64_t size){}) override {
	std::function<void(uint64_t)> sizeCallback = [](uint64_t size){}) override
	{

git-lfs support #10153

Are you sure you want to change the base?

git-lfs support #10153

Conversation

b-camacho commented Mar 4, 2024 • edited Loading

Motivation

Context

Changes

Priorities and Process

b-camacho commented Mar 5, 2024

L-as commented Mar 9, 2024 • edited Loading

b-camacho commented Mar 11, 2024

L-as commented Mar 14, 2024

b-camacho commented Mar 14, 2024

nixos-discourse commented Apr 8, 2024

khoitd1997 commented Jun 22, 2024

roberth commented Jun 22, 2024

kip93 commented Jul 23, 2024

roberth commented Jul 23, 2024

b-camacho commented Jul 24, 2024

roberth commented Jul 24, 2024

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

kip93 commented Oct 10, 2024

b-camacho commented Nov 28, 2024

kip93 commented Nov 29, 2024

b-camacho commented Dec 2, 2024

kip93 commented Dec 2, 2024

b-camacho commented Dec 2, 2024

kip93 commented Dec 2, 2024

kip93 commented Dec 4, 2024

b-camacho commented Dec 5, 2024

kip93 commented Dec 6, 2024

jsqu4re commented Dec 9, 2024

kip93 commented Dec 9, 2024 • edited Loading

roberth left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

b-camacho commented Dec 9, 2024 • edited Loading

jsqu4re commented Dec 9, 2024

nixos-discourse commented Dec 11, 2024

CHN-beta commented Dec 12, 2024

kip93 commented Dec 16, 2024

b-camacho commented Dec 16, 2024

kip93 commented Dec 18, 2024

kip93 commented Dec 19, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

b-camacho commented Mar 4, 2024 •

edited

Loading

L-as commented Mar 9, 2024 •

edited

Loading

kip93 commented Dec 9, 2024 •

edited

Loading

b-camacho commented Dec 9, 2024 •

edited

Loading