[pull] master from git:master#73
Merged
pull[bot] merged 28 commits intoturkdevops:masterfrom Jul 2, 2025
Merged
Conversation
Remotes can be configured either via a repository's config or by using
the ".git/branches/" or ".git/remotes/" directories. Back when the new
config-based mechanism has been introduced we also introduced a helper
script that migrates from the old-style remote configuration to the new
config-based mechanism.
With the recent removal announcement for the two directories we also
started to instruct users to migrate repositories that still use these
mechanism to use config-based remotes. Notably though, the migration
path doesn't even use the migration script. Instead, git-remote(1)
itself knows how to migrate any such remote via `git remote rename`.
In fact, a full migration _cannot_ use the script as it only knows to
migrate remotes from ".git/remotes/", but not ".git/branches/". As such,
the migration path via `git remote rename` is the only feasible way to
fully migrate repositories over to the new format.
Last but not least, the script doesn't even work as-is as it sources
"git-sh-setup". For this to work it would need to be invoked either via
Git so that this script is in our PATH, users would have to manually
call it with an adjusted PATH, or distributions need to install the
script into "$prefix/libexec/git-core" with a "git-" prefix. All of
these steps are unlikely enough to underpin the claim that this script
is not used at all.
So given that:
- The script cannot perform a full migration of all deprecated remote
types.
- We don't advertise it anywhere.
- It has been basically untouched since 2007.
- It doesn't even work unless users do manual steps.
It should be safe enough to just remove it. Do so.
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "examples" directory used to contain scripted versions of some of our builtins. These have all been removed in 49eb8d3 (Remove contrib/examples/*, 2018-03-25), but we left a note in the directory to make it discoverable that there used to be examples. It is unlikely that anybody still looks at these examples more than 7 years after they have been removed. Remove the note and its directory. Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "remote-helpers" directory contains two remote helper scripts for Mercurial and Bazaar. These scripts have since been converted into stubs in b2c851a (Revert "Merge branch 'jc/graduate-remote-hg-bzr' (early part)", 2014-05-20) as the helpers have been moved into their own upstream projects [1][2]. Given that these stubs have been created more than a decade ago it is very unlikely that anybody still tries to use them. Remove them. [1]: https://github.com/felipec/git-remote-bzr [1]: https://github.com/felipec/git-remote-hg Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "thunderbird-patch-inline" directory in "contrib/" contains a script to send patch files via Thunderbird. This script depends on the ExternalEditor extension [1], which seems to be effectively unmaintained with the last update being in 2008. While the extension has eventually been maintained in [2], that fork hasn't received any updates since 2020, either. As such, the ExternalEditor extension does not work with modern versions of Thunderbird anymore, and as the "thunderbird-patch-inline" script depends on the ExternalEditor extension it likely doesn't work anymore, either. The fact that this script hasn't been touched for the last 10 years outside of some global cleanup supports the idea that it is not useful anymore. Remove it. [1]: https://globs.org/articles.php?lng=en&pg=2 [2]: https://github.com/exteditor/exteditor/releases Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "hooks" directory contains a handful of example hooks. Most of these hooks are highly specific and haven't really received any updates over the last couple of years, except for some global cleanups. The multimail hook has also been removed in f74d114 (multimail: stop shipping a copy, 2021-06-10) in favor of its upstream project [1]. Remove those hooks. If we want to provide examples for how to use Git hooks we should do that as part of our documentation, for example in githooks(5). [1]: https://github.com/git-multimail/git-multimail Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "mw-to-git" directory contains tools for accessing MediaWiki via Git. The scripts are essentially unmaintained in Git: despite a couple of global cleanups, the last changes were a couple of security-related issues part of 9a86064 (remote-mediawiki: use "sh" to eliminate unquoted commands, 2020-09-21) and its parents. We don't ever run any of the tests so it is more likely than not that many of the tests have been bitrotting, like e.g. documented in f8ab018 (remote-mediawiki tests: annotate failing tests, 2020-09-21). According to Matthieu Moy [1], one of the original developers of this tool, it didn't receive any attention recently and there is no motivation to keep maintaining it anymore in the community. The project has been spun out of Git [2] and thus has a new official home, but did not receive much attention over there, either. As such, it seems like the MediaWiki transport helper is slowly fading away. But given that there is a new home, it doesn't make sense to have it as part of Git anymore only to let it rot. Remove the directory. [1]: <108f297a-b415-4742-80e4-51ea02af18e9@matthieu-moy.fr> [2]: https://github.com/Git-Mediawiki/Git-Mediawiki Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "persistent-https" remote helper supposedly speeds up SSL operations by running a daemon that keeps a connection open to a remote server. It is effectively unmaintained nowadays: the last time it received an update was in accb613 (contrib/persistent-https: use Git version for build label, 2016-07-20) and its parent commits to make it compile with Go 1.7+. This Go toolchain is somewhat dated by now though and unsupported. The oldest still-supported toolchain is Go 1.23, which was released in August 2024. It is not possible to compile the remote helper with that Go version anymore: $ go version go version go1.23.8 linux/amd64 $ make case $(go version) in \ "go version go"1.[0-5].*) EQ=" " ;; *) EQ="=" ;; esac && \ go build -o git-remote-persistent-https \ -ldflags "-X main._BUILD_EMBED_LABEL${EQ}GIT_VERSION=2.49.0.943.g965a70ebf62" go: cannot find main module, but found .git/config in /home/pks/Development/git to create a module there, run: cd ../.. && go mod init make: *** [Makefile:31: git-remote-persistent-https] Error 1 The problem is that modern Go toolchains require a "go.mod" file, but we don't have any such files. This requirement exists since quite a while already, so it's clear that nobody has tried to use this remote helper anytime recent. Remove the remote helper. Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "git-resurrect.sh" script can be used to find traces of a branch tip in the reflog and resurrect that branch. Despite a couple of global cleanups, the script hasn't seen any activity since it was introduced in e1ff064 (contrib git-resurrect: find traces of a branch name and resurrect it, 2009-02-04). Furthermore, the tool does not work with the "reftable" backend at all as it directly reads ".git/logs/HEAD". As reflogs are stored as part of the individual tables though that file wouldn't exist in a "reftable"- enabled repository. Last but not least, the tool doesn't even work unless it is explicitly invoked via `git resurrect` as it sources "git-sh-setup". As none of our build systems know to install this script, users thus have to go out of their way to really make it work, which is highly unlikely. Another source that indicates that this tool can be removed is a question for how to restore deleted branches on StackOverflow [1]. The top-voted answer uses git-reflog(1) directly and has received more than 3000 votes to date. While "git-resurrect.sh" is also mentioned, it only got 16 upvotes, and comments mention the above caveat that users have to do some manual setup to make it work. It's thus rather clear that the tool doesn't have a lot or even any users. Remove it. [1]: https://stackoverflow.com/questions/3640764/can-i-recover-a-branch-after-its-deletion-in-git Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
While the "emacs/" directory still exists, all of its code has been replaced with stubs in 6d5ed48 (git{,-blame}.el: remove old bitrotting Emacs code, 2018-04-11). Instead, the recommendation is to use Emacs' own vc-annotate mode. Remove the code altogether. Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "git-new-workdir" command has been introduced to make it possible to have a separate working directory in a different place. The command thus predates git-worktree(1), which is what people use nowadays to create any such working directory. As such, the script doesn't really have much of a reason to exist nowadays anymore. It also doesn't seem like the script is still in use: the last time it has received an update was in e32afab (git-new-workdir: don't fail if the target directory is empty, 2014-11-26), more than a decade ago. Remove it as well as the tests that depend on it. Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "stats" directory contains a couple of scripts to do some statistics
on a repository:
- "git-common-hash" shows the longest common hash prefixes and can be
used to determine the minimum prefix length to use for object names
to be unique. The script has last been touched in 53474eb
(contrib: update stats/mailmap script, 2012-12-12) and searching for
it on the internet doesn't really surface any potential use cases or
even mentions of it.
Modern Git also shouldn't really need this tool as it knows to
automatically scale printed prefixes via some heuristics.
- "mailmap.pl" performs some statistics on the number of mailmapped
commits in a repository. It has last been modified in 53474eb
(contrib: update stats/mailmap script, 2012-12-12) and has since
been bitrotting. It doesn't even compile nowadays anymore:
$ perl contrib/stats/mailmap.pl
Experimental keys on scalar is now forbidden at contrib/stats/mailmap.pl line 57.
Type of arg 1 to keys must be hash or array (not hash element) at contrib/stats/mailmap.pl line 57, near "}) "
Experimental keys on scalar is now forbidden at contrib/stats/mailmap.pl line 57.
Type of arg 1 to keys must be hash or array (not private variable) at contrib/stats/mailmap.pl line 57, near "$h)"
Experimental keys on scalar is now forbidden at contrib/stats/mailmap.pl line 64.
Type of arg 1 to keys must be hash or array (not private variable) at contrib/stats/mailmap.pl line 64, near "$h)"
Execution of contrib/stats/mailmap.pl aborted due to compilation errors.
This should be good-enough signal to indicate that nobody is using
this script at all anymore.
- "packinfo.pl" takes the output from git-verify-pack(1) and performs
some pretty printing thereof. On the one hand it reformats the
output to be easier to read and provide some summaries. On the other
hand it may also print filenames of blobs.
We don't have any replacement for this tool. Ideally, we should move
its functionality into git-verify-pack(1) itself.
Remove the first two scripts, but retain "packinfo.pl".
Signed-off-by: Patrick Steinhardt <ps@pks.im>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6d1f198 (imap-send: fix leaking memory in `imap_server_conf`, 2024-06-07) resulted a change in static int git_imap_config which resulted in cfg->folder being incorrectly set to NULL in case imap.user, imap.pass, imap.tunnel and imap.authmethod were defined. Because of this, since Git 2.46.0, git-imap-send is not usable at all. The bug seems to have been unnoticed for a long time, likely due to better options like git-send-email. Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
This patch fixes a memory leak by running free(response) in case auth_cram_md5 fails. Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
…thout OpenSSL Unlike PLAIN, XOAUTH2 and OAUTHBEARER, CRAM-MD5 authentication is not supported by libcurl and requires OpenSSL. If the user tries to use CRAM-MD5 authentication without OpenSSL, the previous behaviour was to attempt to authenticate and fail with a die(error). Handle this in a better way by first checking if OpenSSL is available and then attempting to authenticate. If OpenSSL is not available, print an error message and exit gracefully. Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
OAuth2.0 is a new way of authentication supported by various email providers these days. OAUTHBEARER and XOAUTH2 are the two most common mechanisms used for OAuth2.0. OAUTHBEARER is described in RFC5801[1] and RFC7628[2], whereas XOAUTH2 is Google's proprietary mechanism (See [3]). [1]: https://datatracker.ietf.org/doc/html/rfc5801 [2]: https://datatracker.ietf.org/doc/html/rfc7628 [3]: https://developers.google.com/workspace/gmail/imap/xoauth2-protocol#initial_client_response Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
The current implementation for PLAIN in imap-send works just fine if using curl, but if attempted to use for OpenSSL, it is treated as an invalid mechanism. The default implementation for OpenSSL is IMAP LOGIN command rather than AUTH PLAIN. Since AUTH PLAIN is still used today by many email providers in form of app passwords, lets add an implementation that can use AUTH PLAIN if specified. Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some users may very often want to imap-send messages to a folder other than the default set in the config. Add a command line argument for the same. While at it, fix minor mark-up inconsistencies in the existing documentation text. Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Various IMAP servers have different ways to name common folders.
For example, the folder where all deleted messages are stored is often
named "[Gmail]/Trash" on Gmail servers, and "Deleted" on Outlook.
Similarly, the Drafts folder is simply named "Drafts" on Outlook, but
on Gmail it is named "[Gmail]/Drafts".
This commit adds a `--list` command to the `imap-send` tool that lists
the available folders on the IMAP server, allowing users to see
which folders are available and how they are named. A sample output
looks like this when run against a Gmail server:
Fetching the list of available folders...
* LIST (\HasNoChildren) "/" "INBOX"
* LIST (\HasChildren \Noselect) "/" "[Gmail]"
* LIST (\All \HasNoChildren) "/" "[Gmail]/All Mail"
* LIST (\Drafts \HasNoChildren) "/" "[Gmail]/Drafts"
* LIST (\HasNoChildren \Important) "/" "[Gmail]/Important"
* LIST (\HasNoChildren \Sent) "/" "[Gmail]/Sent Mail"
* LIST (\HasNoChildren \Junk) "/" "[Gmail]/Spam"
* LIST (\Flagged \HasNoChildren) "/" "[Gmail]/Starred"
* LIST (\HasNoChildren \Trash) "/" "[Gmail]/Trash"
For OpenSSL, this is achived by running the 'IMAP LIST' command and
parsing the response. This command is specified in RFC6154:
https://datatracker.ietf.org/doc/html/rfc6154#section-5.1
For libcurl, the example code published in the libcurl documentation
is used to implement this functionality:
https://curl.se/libcurl/c/imap-list.html
Signed-off-by: Aditya Garg <gargaditya08@live.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When requesting for passsword, git credential helper used to display
only the host name. For example:
Password for 'imaps://gargaditya08%40live.com@outlook.office365.com':
Now, it will display the port along with the host name:
Password for 'imaps://gargaditya08%40live.com@outlook.office365.com:993':
This has been done to make credential helpers more specific for ports.
Also, this behaviour will also mimic git send-email, which displays
the port along with the host name when requesting for a password.
Signed-off-by: Aditya Garg <gargaditya08@live.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Whenever we sent a message using the `imap-send` command, it would
display a log showing the number of messages which are to be sent.
For example:
sending 1 message
100% (1/1) done
This had been made more informative by adding the name of the destination
folder as well:
Sending 1 message to Drafts folder...
100% (1/1) done
Signed-off-by: Aditya Garg <gargaditya08@live.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some minor mistakes have been found in the logs. Most of them include error messages starting with a capital letter, and ending with a period. Abbreviations like "IMAP" and "OK" should also be in uppercase. Another mistake was that the error message showing unknown authentication mechanism used was displaying the host rather than the mechanism in the logs. Fix them. Signed-off-by: Aditya Garg <gargaditya08@live.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ag/imap-send-resurrection: imap-send: fix minor mistakes in the logs imap-send: display the destination mailbox when sending a message imap-send: display port alongwith host when git credential is invoked imap-send: add ability to list the available folders imap-send: enable specifying the folder using the command line imap-send: add PLAIN authentication method to OpenSSL imap-send: add support for OAuth2.0 authentication imap-send: gracefully fail if CRAM-MD5 authentication is requested without OpenSSL imap-send: fix memory leak in case auth_cram_md5 fails imap-send: fix bug causing cfg->folder being set to NULL
The error message 'no imap store specified' is misleading because it refers to 'store' when the actual missing configuration is 'imap.folder'. Update the message to use the correct terminology that matches the configuration variable name. This reduces confusion for users who might otherwise look for non-existent 'imap.store' configuration when they see this error. Signed-off-by: Jörg Thalheim <joerg@thalheim.io> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Replace basic error messages with more helpful ones that guide users on how to resolve configuration issues. When imap.host or imap.folder are missing, provide the exact git config commands needed to fix the problem, along with examples of typical values. Use the advise() API to display hints in a multi-line format with proper "hint:" prefixes for each line. Signed-off-by: Jörg Thalheim <joerg@thalheim.io> Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git imap-send" has been broken for a long time, which has been resurrected and then taught to talk OAuth2.0 etc. * ag/imap-send-resurrection: imap-send: fix minor mistakes in the logs imap-send: display the destination mailbox when sending a message imap-send: display port alongwith host when git credential is invoked imap-send: add ability to list the available folders imap-send: enable specifying the folder using the command line imap-send: add PLAIN authentication method to OpenSSL imap-send: add support for OAuth2.0 authentication imap-send: gracefully fail if CRAM-MD5 authentication is requested without OpenSSL imap-send: fix memory leak in case auth_cram_md5 fails imap-send: fix bug causing cfg->folder being set to NULL
Remove bunch of stuff from contrib/ hierarchy. * ps/contrib-sweep: contrib: remove some scripts in "stats" directory contrib: remove "git-new-workdir" contrib: remove "emacs" directory contrib: remove "git-resurrect.sh" contrib: remove "persistent-https" remote helper contrib: remove "mw-to-git" contrib: remove "hooks" directory contrib: remove "thunderbird-patch-inline" contrib: remove remote-helper stubs contrib: remove "examples" directory contrib: remove "remotes2config.sh"
Update some error messages from "git imap-send". * jt/imap-send-message-fix: imap-send: improve error messages with configuration hints imap-send: fix confusing 'store' terminology in error message
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.2)
Can you help keep this open source service alive? 💖 Please sponsor : )