gitgitgadget / git Goto Github PK

This project forked from git/git

GitGitGadget's Git fork. Open Pull Requests here to submit them to the Git mailing list

Home Page: https://gitgitgadget.github.io/

License: Other

Shell 38.33% Roff 0.12% Makefile 0.75% C 50.00% C++ 0.12% Perl 5.25% M4 0.17% Emacs Lisp 0.01% Python 0.79% Go 0.07% Tcl 3.84% JavaScript 0.22% CSS 0.05% Batchfile 0.04% SmPL 0.07% CMake 0.18%

hacktoberfest

git's Introduction

Contributing git.git patches via GitHub PRs

This project's goal is to make contributions to the Git project (almost) as easy as opening a GitHub PR. The idea is for users to open a Pull Request at

https://github.com/gitgitgadget/git

with a good description of their patch series. Then, the command /submit, issued via a comment on said PR will tell GitGitGadget to send the patches as mail thread to the Git mailing list, with the Pull Request's description as cover letter.

As is common, reviewers on the Git mailing list will probably ask for modifications. These should be folded into the respective commits (or inserted as stand-alone commits at an appropriate place in the patch series) via git rebase -i, followed by a force-push. Once everything is in a good shape, update the description to include information about changes performed relative to the latest patch series iteration, and then another /submit will ask GitGitGadget to send a new iteration of the patch series.

All relevant information, such as the current iteration of the patch series, the Message-ID of the sent mails, etc is stored in the Git notes in refs/notes/gitgitgadget.

Note: GitGitGadget will Cc: the original authors when sending patches on their behalf, and people mentioned in the Cc: footer of the Pull Request description.

Furthermore, for all iterations of a patch series but the first one, GitGitGadget will insert a machine-generated representation of what changed between revisions, and reply to the cover letter of the previous iteration. This patch revision diff can be supressed if the change may be too large or irrelevant by adding a Range-Diff: false footer in the Pull Request description.

For convenience of reviewers, GitGitGadget will generate tags for each iteration it sent, and push those to https://github.com/gitgitgadget/git. Links to those tags will be included in the cover letter.

git's People

Contributors

Stargazers

Watchers

Forkers

webstech fcharlie johnlinp dilijev mark-rushakoff dakotahawkins echuber2 kdnakt aviskarkc10 jpninanjohn normanr utsav2 qusielle 2asoft ryenus qdigital 7960 fmstrat frenk8585 rextge nirmalhk7 dinoallo mildsunrise isengart gsus333 podsvirov sluongng atooq1 dmalan calinou kontesanjana yuuri trivalik underpressure01 adlternative claude743 sura907 eiadshahtout chinmoy12c bukforks tfidfwastaken ali-yekta110 emclain certified-git taok oded-ist abashewsergei dungsaga brendankristiansen mstrap mattiza curieuxdansvie justinrdonnelly vtbassmatt thiago-figueredo rob-3 rwilliams251 tapaswenipathak dylay binsee agreenbhm metolinam ourworld8 iza1977 bnjf wutthichai46 prplr edecosta-mw puchongcnxx29 kboxox jerry-skydio calvin-wan-google ashba1995 cklim79 srz-zumix enochprime mhucka harshil-jani oiuytredc aetotvl6789 ahlinc changren-wcr datacrusade1999 mika-f tydick87 forktonguedpygmy ightevenmckane bronkhonkytonk jcash0464 rushikesh90 zosopentools vahid55555 timosha135 nizhal soulkingx wingt runsascoded hamed516 mystarset icbd

git's Issues

Offer an easy way to fetch/pull the branch matching the name of the current branch

When fetching from a remote, one can specify which ref(s) to fetch.

When pulling, there is even a very convenient convention to fetch and merge the "upstream branch", i.e. the one connected via git branch --set-upstream-to=<remote>/<branch>.

However, there is no convenient way to specify the name of the current branch, other than typing it out. This is a problem e.g. when working on a multi-platform contribution for the Git project itself, where the upstream branch might be origin/master, but the contributor needs work on the same branch on multiple machines. In this instance, the branch might need to be updated with git pull --rebase my-fork <current-branch>.

Let's introduce the --current-branch option for this, automagically appending the name of the current branch to the commands git fetch and git pull, allowing e.g. git pull --rebase --current-branch my-fork.

Weekly Digest (30 June, 2019 - 7 July, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 4 issues were created.
Of these, 0 issues have been closed and 4 issues are still open.

OPEN ISSUES

💚 #283 Follow-up on top of js/mingw-use-utf8, by dscho
💚 #282 contrib/completion: use __git() in bash completor to avoid aliased tools, by zaquestion
💚 #281 git-prompt: improve cherry-pick/revert detection, by phillipwood
💚 #280 Fix test bug with spaces in file owner, by derrickstolee

NOISY ISSUE

🔈 #280 Fix test bug with spaces in file owner, by derrickstolee
It received 8 comments.

PULL REQUESTS

Last week, 26 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 26 pull requests were updated.
💛 #283 Follow-up on top of js/mingw-use-utf8, by dscho
💛 #282 contrib/completion: use __git() in bash completor to avoid aliased tools, by zaquestion
💛 #281 git-prompt: improve cherry-pick/revert detection, by phillipwood
💛 #280 Fix test bug with spaces in file owner, by derrickstolee
💛 #277 Update git-clone doc: refer to long form of the options and list short form of the options first, by kenden
💛 #276 t3420 remove progress from output, by phillipwood
💛 #275 Quieter sequencer status parsing, by phillipwood
💛 #267 t3404: fix a typo, by dscho
💛 #265 Support building with GCC v8.x/v9.x, by dscho
💛 #263 submodule foreach: fix recursion of options, by momoson
💛 #262 add -p: fix checkout -p with pathological context, by phillipwood
💛 #261 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee
💛 #254 [RFC] Create 'core.featureAdoptionRate' setting to update config defaults, by derrickstolee
💛 #253 Make rebase.reschedulefailedexec less overzealous, by dscho
💛 #220 mingw: support spawning programs with paths containing spacesnames, by dscho
💛 #219 Show an error if too-long paths are seen by git clean -dfx, by dscho
💛 #218 Give Git a HOME on Windows, by dscho
💛 #217 gettext(windows): always use UTF-8, by dscho
💛 #216 windows: avoid illegal filenames during a build, by dscho
💛 #215 mingw: enable GCC's stack smashing protector, by dscho
💛 #214 windows: embed a manifest, by dscho
💛 #213 diff: release all handles before running external diff, by dscho
💛 #184 Commit-graph: Write incremental files, by derrickstolee
💛 #151 Fix a test on NTFS (and probably HFS+), by dscho
💛 #149 Fix MSVC support, at long last, by dscho
💛 #147 Let's use the Win32 API more precisely, by dscho

COMMITS

Last week there were no commits.

CONTRIBUTORS

Last week there were no contributors.

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Weekly Digest (9 June, 2019 - 16 June, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 7 issues were created.
Of these, 0 issues have been closed and 7 issues are still open.

OPEN ISSUES

💚 #267 t3404: fix a typo, by dscho
💚 #266 ci: split linux-gcc into linux-gcc and linux-gcc-extra, by dscho
💚 #265 Support building with GCC v8.x/v9.x, by dscho
💚 #264 Adding git-ignore command, tests, and documentation., by tstone2077
💚 #263 submodule foreach: fix recursion of options, by momoson
💚 #262 add -p: fix checkout -p with pathological context, by phillipwood
💚 #261 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee

NOISY ISSUE

🔈 #262 add -p: fix checkout -p with pathological context, by phillipwood
It received 11 comments.

PULL REQUESTS

Last week, 18 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 18 pull requests were updated.
💛 #267 t3404: fix a typo, by dscho
💛 #266 ci: split linux-gcc into linux-gcc and linux-gcc-extra, by dscho
💛 #265 Support building with GCC v8.x/v9.x, by dscho
💛 #264 Adding git-ignore command, tests, and documentation., by tstone2077
💛 #263 submodule foreach: fix recursion of options, by momoson
💛 #262 add -p: fix checkout -p with pathological context, by phillipwood
💛 #261 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee
💛 #228 sha1-file: split OBJECT_INFO_FOR_PREFETCH, by derrickstolee
💛 #223 Fix racy fsmonitor, by dscho
💛 #221 Fix two issues pointed out by Coverity, by dscho
💛 #196 status: remove the empty line after hints, by johnlinp
💛 #189 trace2: Add variable description to git.txt, by derrickstolee
💛 #184 Commit-graph: Write incremental files, by derrickstolee
💛 #183 p4: fix "Not a valid object name HEAD0" when unshelving, by mdymike
💛 #151 Fix a test on NTFS (and probably HFS+), by dscho
💛 #145 Avoid spawning gzip in git archive, by dscho
💛 #142 Optimize run_diff_files()' rename detection, by dscho
💛 #112 Commit-graph write refactor (was: Create commit-graph file format v2), by derrickstolee

COMMITS

CONTRIBUTORS

Last week there was 1 contributor.
👤 gitster

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Make OPT_RERERE_AUTOUPDATE's "unspecified" value consistent with other unspecified values

In Git's source code, the (unwritten) convention is to use -1 for unspecified values of command-line options of config settings.

Currently, the rerere.autoUpdate setting disagrees, and we should rectify this.

See https://public-inbox.org/git/[email protected]/

Don't run Cocci on compat/ files that are pulled externally

Since some files in compat/ are pulled externally, ideally we should not update them unless absolutely necessary. As a result, we should find what these files are and blacklist Cocci from being run on them.

One thing to consider, some files are Windows-only and are never Coccinelled whenever we compile on Windows. Do we want to make Cocci run on all files, regardless of whether or not they're compilable on the current platform? Upside is that Szeder found some files that could be patched that we weren't catching. Downside is that there is no way to test the code to ensure that it works properly.

Another stretch goal is to do the extern removal on files that we do own in compat/, since those were explicitly excluded during the big extern cleanup since those were intentionally excluded in b199d71 (.[ch]: remove extern from function declarations using sed, 2019-04-29) and 5545442 (.[ch]: remove extern from function declarations using spatch, 2019-04-29).

some PRs need effective rebasing between gfs and upstream git.

It is tricky playing both sides against each other for commit diffs that should transferable between the two code bases (gfw <-> git), as seen in #237

Not sure if GitHub could support this?

Implement command-line options to change `branch.<name>.remote/pushRemote` configs

Perhaps something like --set-remote-to, --set-push-remote-to, --set-remote-save-to-push?

For the last, this would behave similarly to f39a9c6 (remote: add --save-to-push option to git remote set-url, 2018-12-10).

That option is useful for a situation such as mirroring. Consider if I wanted to mirror an upstream branch on my own remote. I would have something such as

[branch "master"]
    remote = "upstream"
    pushRemote = "origin"

in my config.

We could make this efficient to setup by doing:

git checkout master # is currently tracking origin's master
git branch --set-remote-save-to-push upstream

This is already implemented here but I think this idea might be a little too stupid to contribute. Maybe I'll send it in one day...

Make Git subcommand documentation more consistent

In remote's documentation, in the Synopsis, the subcommands (add, remove, etc...) are underlined and Commands section, they are also underlined.

In submodule, the subcommands are not underlined in both they Synopsis and the Commands sections.

In worktree, the subcommands are underlined in the Synopsis but not in the Commands section.

I'm sure there are other Git commands that have subcommands and their documentation is probably all over the place as well. Anyway, we should look over this, find one style to stick with and standardise it for all the documentation.

Fix flaky "t5516.81 deny fetch unreachable SHA1, allowtipsha1inwant=false"

This test seems to fail infrequently in our Azure Pipeline, apparently most often in the msvc job. See https://dev.azure.com/Git-for-Windows/git/_build/results?buildId=36820&view=ms.vss-test-web.build-test-results-tab for an example.

Invent a way to retain reflogs for a while after the ref was deleted

It is a bit of a sad story that reflogs are deleted the moment the ref is deleted. So there is no easy "undo" there, even when the user has an "Ooops, I needed that branch and its reflog!" moment.

This will require a sane design around ref-less reflogs that can be garbage-collected after a configurable time.

Weekly Digest (26 May, 2019 - 2 June, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 26 issues were created.
Of these, 2 issues have been closed and 24 issues are still open.

OPEN ISSUES

CLOSED ISSUES

❤️ #232 Having grep support the -o option like GNU grep et al, by dscho
❤️ #227 Deprecate git rebase -p even more, by dscho

NOISY ISSUE

🔈 #227 Deprecate git rebase -p even more, by dscho
It received 12 comments.

PULL REQUESTS

Last week, 16 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 16 pull requests were updated.
💛 #238 Document that Unspecified options should initialize to -1, by PhilipOakley
💛 #229 git-completion: sanitize the command names, by MrMebelMan
💛 #228 sha1-file: split OBJECT_INFO_FOR_PREFETCH, by derrickstolee
💛 #226 bundle verify: improve the user experience when called without a .git/ directory, by dscho
💛 #223 Fix racy fsmonitor, by dscho
💛 #221 Fix two issues pointed out by Coverity, by dscho
💛 #219 Show an error if too-long paths are seen by git clean -dfx, by dscho
💛 #208 Close commit-graph before calling 'gc', by derrickstolee
💛 #196 status: remove the empty line after hints, by johnlinp
💛 #195 Drop support for git rebase --preserve-merges, by dscho
💛 #194 Clean up after the removal of the scripted rebase, by dscho
💛 #189 trace2: Add variable description to git.txt, by derrickstolee
💛 #184 [RFC] Commit-graph: Write incremental files, by derrickstolee
💛 #183 p4: fix "Not a valid object name HEAD0" when unshelving, by mdymike
💛 #177 documentation: add lab for first contribution, by nasamuffin
💛 #112 Commit-graph write refactor (was: Create commit-graph file format v2), by derrickstolee

COMMITS

CONTRIBUTORS

Last week there were 7 contributors.
👤 gitster
👤 pclouds
👤 dscho
👤 jonathantanmy
👤 jeffhostetler
👤 szeder
👤 nasamuffin

STARGAZERS

Last week there was 1 stargazer.
⭐ rafasc
You are the star! 🌟

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Add more config IncludeIf conditions

Recently there was a mention on git-users to extend the includeIf statement to read config.

Now that seems like a nasty circular dependency but there's other low-hanging fruit there, like make it match a given env name to a value (or glob?).

Copied from: https://public-inbox.org/git/[email protected]/

Weekly Digest (2 June, 2019 - 9 June, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 7 issues were created.
Of these, 2 issues have been closed and 5 issues are still open.

OPEN ISSUES

💚 #259 Add support for drop!/reword! commits to git rebase -i, by dscho
💚 #258 Fix flaky t5504 (in test case 9 - push with transfer.fsckobjects), by dscho
💚 #256 Avoid using Perl in the test suite, by dscho
💚 #254 [RFC] Create 'core.size=large' setting to update config defaults, by derrickstolee
💚 #253 Make rebase.reschedulefailedexec less overzealous, by dscho

CLOSED ISSUES

❤️ #257 Fix flaky t5516-fetch-push (test case 79 - deny fetch unreachable SHA1, allowtipsha1inwant=true), by dscho
❤️ #255 Improve instructions around how to set git-prompt preference variables., by ghedsouza

LIKED ISSUE

👍 #253 Make rebase.reschedulefailedexec less overzealous, by dscho
It received 👍 x1, 😄 x0, 🎉 x0 and ❤️ x0.

NOISY ISSUE

🔈 #255 Improve instructions around how to set git-prompt preference variables., by ghedsouza
It received 16 comments.

PULL REQUESTS

Last week, 14 pull requests were created, updated or merged.

OPEN PULL REQUEST

Last week, 1 pull request was opened.
💚 #256 Avoid using Perl in the test suite, by dscho

UPDATED PULL REQUEST

Last week, 13 pull requests were updated.
💛 #254 [RFC] Create 'core.size=large' setting to update config defaults, by derrickstolee
💛 #253 Make rebase.reschedulefailedexec less overzealous, by dscho
💛 #228 sha1-file: split OBJECT_INFO_FOR_PREFETCH, by derrickstolee
💛 #223 Fix racy fsmonitor, by dscho
💛 #221 Fix two issues pointed out by Coverity, by dscho
💛 #196 status: remove the empty line after hints, by johnlinp
💛 #194 Clean up after the removal of the scripted rebase, by dscho
💛 #184 Commit-graph: Write incremental files, by derrickstolee
💛 #183 p4: fix "Not a valid object name HEAD0" when unshelving, by mdymike
💛 #151 Fix a test on NTFS (and probably HFS+), by dscho
💛 #149 Fix MSVC support, at long last, by dscho
💛 #142 Optimize run_diff_files()' rename detection, by dscho
💛 #112 Commit-graph write refactor (was: Create commit-graph file format v2), by derrickstolee

COMMITS

CONTRIBUTORS

Last week there were 13 contributors.
👤 gitster
👤 jiangxin
👤 cmalard
👤 ruester
👤 ralfth
👤 newren
👤 alshopov
👤 fangyi-zhou
👤 ChrisADR
👤 AlessandroMenti
👤 jnavila
👤 vnwildman
👤 tmzullinger

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Add another set of GIT_{AUTHOR,COMMITTER}_{NAME,EMAIL} with lower priorities

It would be nice if git supported a way to do override the fallback when neither the config nor the current environment variables specify the author/committer name.

Copy-munged from: https://public-inbox.org/git/[email protected]/

Fix inconsistent regexp usage

If you supply --perl-regexp to git log it only applies to --grep. The documentation says "Consider the limiting patterns to be Perl-compatible regular expressions". Which might lead you to think that e.g. -G uses it too. It doesn't, only grep.c does PCRE, but -G is handled by diffcore-pickaxe.c.

Looking at git grep -l regexec -- '*.c' this whole thing is a mess. Ideally you should be able to say you want to use PCRE for everything, except maybe things that end up in your .gitconfig or e.g. the diff driver. But we could really use a more generic regexp interface.

In general Git's regexp use is a huge mess, e.g. there's --regexp-ignore-case but no way to supply various other regexp options like REG_NEWLINE or PCRE options consistently.

Copy-munged from: https://public-inbox.org/git/CACBZZX5wdnA-96e11edE7xRnAHo19RFTrZmqFQj-0ogLOJTncQ@mail.gmail.com/

Weekly Digest (23 June, 2019 - 30 June, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 4 issues were created.
Of these, 1 issues have been closed and 3 issues are still open.

OPEN ISSUES

💚 #277 Update git-clone doc: refer to long option, by kenden
💚 #276 t3420 remove progress from output, by phillipwood
💚 #275 Quieter sequencer status parsing, by phillipwood

CLOSED ISSUES

❤️ #278 l10n: de.po: Fix German translation, by derphilipp

NOISY ISSUE

🔈 #275 Quieter sequencer status parsing, by phillipwood
It received 13 comments.

PULL REQUESTS

Last week, 22 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 22 pull requests were updated.
💛 #277 Update git-clone doc: refer to long option, by kenden
💛 #275 Quieter sequencer status parsing, by phillipwood
💛 #273 fetch: add --[no-]show-forced-updates, by derrickstolee
💛 #272 git-status: create config for ahead/behind calculation, by derrickstolee
💛 #269 Wip/fix continue/ skip in synopsis, by phillipwood
💛 #267 t3404: fix a typo, by dscho
💛 #266 ci: split linux-gcc into linux-gcc and linux-gcc-extra, by dscho
💛 #265 Support building with GCC v8.x/v9.x, by dscho
💛 #263 submodule foreach: fix recursion of options, by momoson
💛 #262 add -p: fix checkout -p with pathological context, by phillipwood
💛 #261 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee
💛 #254 [RFC] Create 'core.featureAdoptionRate' setting to update config defaults, by derrickstolee
💛 #253 Make rebase.reschedulefailedexec less overzealous, by dscho
💛 #217 gettext(windows): always use UTF-8, by dscho
💛 #215 mingw: enable GCC's stack smashing protector, by dscho
💛 #214 windows: embed a manifest, by dscho
💛 #184 Commit-graph: Write incremental files, by derrickstolee
💛 #151 Fix a test on NTFS (and probably HFS+), by dscho
💛 #149 Fix MSVC support, at long last, by dscho
💛 #147 Let's use the Win32 API more precisely, by dscho
💛 #123 fixup! trace2: collect Windows-specific process information, by jeffhostetler
💛 #112 Commit-graph write refactor (was: Create commit-graph file format v2), by derrickstolee

COMMITS

Last week there were no commits.

CONTRIBUTORS

Last week there were no contributors.

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Fix flakey t7519.15

The fsmonitor feature is apparently not quite robust, or at least the mentioned test case isn't (see e.g. https://dev.azure.com/gitgitgadget/git/_build/results?buildId=9064&view=ms.vss-test-web.build-test-results-tab&runId=19542&resultId=101590&paneView=debug).

Weekly Digest (16 June, 2019 - 23 June, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 5 issues were created.
Of these, 1 issues have been closed and 4 issues are still open.

OPEN ISSUES

💚 #273 fetch: add --[no-]show-forced-updates, by derrickstolee
💚 #272 git-status: create config for ahead/behind calculation, by derrickstolee
💚 #271 refactor: rename is_directory() to dir_exists() and use it in clone.c, by johnlinp
💚 #269 Wip/fix continue/ skip in synopsis, by phillipwood

CLOSED ISSUES

❤️ #270 blame: read diff config, by derrickstolee

NOISY ISSUE

🔈 #273 fetch: add --[no-]show-forced-updates, by derrickstolee
It received 3 comments.

PULL REQUESTS

Last week, 16 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 16 pull requests were updated.
💛 #273 fetch: add --[no-]show-forced-updates, by derrickstolee
💛 #272 git-status: create config for ahead/behind calculation, by derrickstolee
💛 #271 refactor: rename is_directory() to dir_exists() and use it in clone.c, by johnlinp
💛 #269 Wip/fix continue/ skip in synopsis, by phillipwood
💛 #267 t3404: fix a typo, by dscho
💛 #265 Support building with GCC v8.x/v9.x, by dscho
💛 #264 Adding git-ignore command, tests, and documentation., by tstone2077
💛 #263 submodule foreach: fix recursion of options, by momoson
💛 #262 add -p: fix checkout -p with pathological context, by phillipwood
💛 #261 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee
💛 #254 [RFC] Create 'core.featureAdoptionRate' setting to update config defaults, by derrickstolee
💛 #196 status: remove the empty line after hints, by johnlinp
💛 #184 Commit-graph: Write incremental files, by derrickstolee
💛 #151 Fix a test on NTFS (and probably HFS+), by dscho
💛 #149 Fix MSVC support, at long last, by dscho
💛 #112 Commit-graph write refactor (was: Create commit-graph file format v2), by derrickstolee

COMMITS

CONTRIBUTORS

Last week there was 1 contributor.
👤 gitster

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Weekly Digest (12 May, 2019 - 19 May, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 21 issues were created.
Of these, 2 issues have been closed and 19 issues are still open.

OPEN ISSUES

💚 #211 rebase -i: implement --reschedule, by dscho
💚 #210 rebase -i: error on fixup/squash when the corresponding pick was skipped, by dscho
💚 #209 rebase -r: implement an equivalent to skip_unnecessary_picks(), by dscho
💚 #208 Close commit-graph before calling 'gc', by derrickstolee
💚 #207 range-diff: allow comparing branches to mboxes, by dscho
💚 #206 range-diff: add support for --left-only and --right-only, by dscho
💚 #205 Offer an easy way to fetch/pull the branch matching the name of the current branch, by dscho
💚 #204 rebase -ir: optionally auto-update local branches when executing label commands, by dscho
💚 #203 git cherry-pick --pick-merges , by dscho
💚 #202 rebase -ir should warn (or error out) when commits were produced but not included in the final commit history, by dscho
💚 #201 Make OPT_RERERE_AUTOUPDATE's "unspecified" value consistent with other unspecified values, by dscho
💚 #200 Mention core.excludesFile in git clean's documentation, by dscho
💚 #199 Allow reset <tag> in git rebase -ir, by dscho
💚 #198 Avoid duplicate entries in git ls-files for unmerged entries, by dscho
💚 #197 Fix flakey t7519.15, by dscho
💚 #196 status: remove the empty line after hints, by johnlinp
💚 #195 Drop support for git rebase --preserve-merges, by dscho
💚 #194 Clean up after the removal of the scripted rebase, by dscho
💚 #191 Remove fast export munging, by newren

CLOSED ISSUES

❤️ #193 Document the stash.useBuiltin escape hatch, by dscho
❤️ #192 pkt-line: fix incorrect function declaration, by dscho

NOISY ISSUE

🔈 #192 pkt-line: fix incorrect function declaration, by dscho
It received 10 comments.

PULL REQUESTS

Last week, 8 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 8 pull requests were updated.
💛 #196 status: remove the empty line after hints, by johnlinp
💛 #195 Drop support for git rebase --preserve-merges, by dscho
💛 #191 Remove fast export munging, by newren
💛 #189 trace2: Add variable description to git.txt, by derrickstolee
💛 #183 p4: fix "Not a valid object name HEAD0" when unshelving, by mdymike
💛 #177 documentation: add lab for first contribution, by nasamuffin
💛 #170 git add -i: add a rudimentary version in C (supporting only status and help so far), by dscho
💛 #92 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee

COMMITS

CONTRIBUTORS

Last week there were 10 contributors.
👤 gitster
👤 tmzullinger
👤 peff
👤 carenas
👤 dscho
👤 avar
👤 Denton-L
👤 derrickstolee
👤 sunshineco
👤 ismail

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

git cherry-pick --pick-merges <range>

This feature would work just like git rebase --rebase-merges, i.e. try to recreate the merge commits contained in the commit range.

Document --stdin (and other options in setup_revisions)

Currently, Git accepts --stdin for the log family of commands. Well really, anything that calls setup_revisions from revisions.c and passes in its argv will implicitly have the --stdin option.

We should explicitly document this (and other) options.

Weekly Digest (7 July, 2019 - 14 July, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 1 issue was created.
It is still open.

OPEN ISSUES

💚 #285 Update gpg.txt to correct gpg --verify syntax, by rtmorgan

NOISY ISSUE

🔈 #285 Update gpg.txt to correct gpg --verify syntax, by rtmorgan
It received 10 comments.

PULL REQUESTS

Last week, 15 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 15 pull requests were updated.
💛 #285 Update gpg.txt to correct gpg --verify syntax, by rtmorgan
💛 #282 contrib/completion: use __git() in bash completor to avoid aliased tools, by zaquestion
💛 #281 git-prompt: improve cherry-pick/revert detection, by phillipwood
💛 #280 Fix test bug with spaces in file owner, by derrickstolee
💛 #276 t3420 remove progress from output, by phillipwood
💛 #275 Quieter sequencer status parsing, by phillipwood
💛 #264 Adding git-ignore command, tests, and documentation., by tstone2077
💛 #261 Create 'expire' and 'repack' verbs for git-multi-pack-index, by derrickstolee
💛 #254 [RFC] Create 'core.featureAdoptionRate' setting to update config defaults, by derrickstolee
💛 #218 Give Git a HOME on Windows, by dscho
💛 #217 gettext(windows): always use UTF-8, by dscho
💛 #216 windows: avoid illegal filenames during a build, by dscho
💛 #213 diff: release all handles before running external diff, by dscho
💛 #184 Commit-graph: Write incremental files, by derrickstolee
💛 #145 Avoid spawning gzip in git archive, by dscho

COMMITS

CONTRIBUTORS

Last week there were 3 contributors.
👤 gitster
👤 dscho
👤 szeder

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

CodingGuidelines: recommend to use -1 for unspecified values of command-line options of config settings

following from #201, document same in Coding Guidelines.

Fix flaky t5504 (in test case 9 - push with transfer.fsckobjects)

This seems to hit the linux-gcc job the most, maybe in the part where it runs the test suite with all kinds of GIT_TEST_* variables set to non-default settings.

See e.g. Attempt #1 of this linux-gcc run:

expecting success: 
	rm -rf dst &&
	git init dst &&
	(
		cd dst &&
		git config transfer.fsckobjects true
	) &&
	test_must_fail git push --porcelain dst master:refs/heads/test >act &&
	test_cmp exp act

+ rm -rf dst
+ git init dst
Initialized empty Git repository in /home/vsts/work/1/s/t/trash directory.t5504-fetch-receive-strict/dst/.git/
+ cd dst
+ git config transfer.fsckobjects true
+ test_must_fail git push --porcelain dst master:refs/heads/test
+ _test_ok=
+ git push --porcelain dst master:refs/heads/test
remote: fatal: object of unexpected type        
error: remote unpack failed: unpack-objects abnormal exit
error: failed to push some refs to 'dst'
+ exit_code=1
+ test 1 -eq 0
+ test_match_signal 13 1
+ test 1 = 141
+ test 1 = 269
+ return 1
+ test 1 -gt 129
+ test 1 -eq 127
+ test 1 -eq 126
+ return 0
+ test_cmp exp act
+ diff -u exp act
--- exp	2019-06-03 20:37:40.928834316 +0000
+++ act	2019-06-03 20:37:41.064834272 +0000
@@ -1,2 +0,0 @@
-To dst
-!	refs/heads/master:refs/heads/test	[remote rejected] (unpacker error)
error: last command exited with $?=1
not ok 9 - push with transfer.fsckobjects
#	
#		rm -rf dst &&
#		git init dst &&
#		(
#			cd dst &&
#			git config transfer.fsckobjects true
#		) &&
#		test_must_fail git push --porcelain dst master:refs/heads/test >act &&
#		test_cmp exp act
#

Make `git diff ..` less confusing for newbies

-phiresky link

The semantics of .. vs ... are essentially swapped between diff and the log family of commands. This is because diff uses setup_revisions but abuses the way that the returned endpoints are used.

There are several options to deal with this:

Print a warning into the output if a user uses git diff A..B instead of git diff A B (see here)
Update the documentation for git-diff so that it's obvious why this is happening (so we tell users about how range-notation is implemented). Upside is users will understand why git diff foo@^! works as well. Downside is that I'm not sure it's productive to expose implementation details to end-uers.
Swap the meaning of .. and .... One can dream...
Something else? I've been mulling over it and not wanting to contribute a patch because there might be some cleaner way to resolve this. Not really sure how.

See:

Teach `git stash` and `git commit` the `-A` option

The git add -A command can be used to conveniently add all files, whether they are tracked or not yet tracked (excluding the ones ignored via .gitignore/.git/info/excludes).

This convenience would come in quite nicely in the stash/commit commands, too (in the latter case, many test cases could be simplified using this new option).

The git stash command already has the --include-untracked option that does this, but not the short-and-sweet short option -A. The git commit command does not have any equivalent (it has the --untracked-files option, but that is really just modifying the output of git status).

Sadly, we cannot add --all, too, as that is already used for something different in git stash: it includes the ignored files, too. Likewise, git commit --all commits all tracked files, and does not add the untracked files (unlike git add --all).

Mention `core.excludesFile` in `git clean`'s documentation

"man git-clean" explicitly mentions two locations related to cleaning:

  -x
        Don't use the standard ignore rules read from .gitignore
        (per directory) and $GIT_DIR/info/exclude, ...

without additionally explicitly mentioning core.excludesFile. if the
man page simply said something like, "using the standard ignore
processing" and left it at that, it would be fine, but to list two of
the locations without the third is potentially confusing.

See https://public-inbox.org/git/[email protected]/

Weekly Digest (19 May, 2019 - 26 May, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 12 issues were created.
Of these, 0 issues have been closed and 12 issues are still open.

OPEN ISSUES

💚 #224 Use user supplied origin name for extension.partialcone instead of, by delphij
💚 #223 Fix racy fsmonitor, by dscho
💚 #222 trace2: fix tracing when NO_PTHREADS is defined, by jeffhostetler
💚 #221 Fix two issues pointed out by Coverity, by dscho
💚 #220 mingw: support spawning programs with paths containing spacesnames, by dscho
💚 #219 Show an error if too-long paths are seen by git clean -dfx, by dscho
💚 #218 Give Git a HOME on Windows, by dscho
💚 #217 gettext(windows): always use UTF-8, by dscho
💚 #216 windows: avoid illegal filenames during a build, by dscho
💚 #215 mingw: enable GCC's stack smashing protector, by dscho
💚 #214 windows: embed a manifest, by dscho
💚 #213 diff: release all handles before running external diff, by dscho

NOISY ISSUE

🔈 #222 trace2: fix tracing when NO_PTHREADS is defined, by jeffhostetler
It received 7 comments.

PULL REQUESTS

Last week, 21 pull requests were created, updated or merged.

OPEN PULL REQUEST

Last week, 3 pull requests were opened.
💚 #220 mingw: support spawning programs with paths containing spacesnames, by dscho
💚 #219 Show an error if too-long paths are seen by git clean -dfx, by dscho
💚 #213 diff: release all handles before running external diff, by dscho

UPDATED PULL REQUEST

Last week, 18 pull requests were updated.
💛 #224 Use user supplied origin name for extension.partialcone instead of, by delphij
💛 #223 Fix racy fsmonitor, by dscho
💛 #222 trace2: fix tracing when NO_PTHREADS is defined, by jeffhostetler
💛 #221 Fix two issues pointed out by Coverity, by dscho
💛 #218 Give Git a HOME on Windows, by dscho
💛 #217 gettext(windows): always use UTF-8, by dscho
💛 #216 windows: avoid illegal filenames during a build, by dscho
💛 #215 mingw: enable GCC's stack smashing protector, by dscho
💛 #214 windows: embed a manifest, by dscho
💛 #208 Close commit-graph before calling 'gc', by derrickstolee
💛 #195 Drop support for git rebase --preserve-merges, by dscho
💛 #194 Clean up after the removal of the scripted rebase, by dscho
💛 #189 trace2: Add variable description to git.txt, by derrickstolee
💛 #184 [RFC] Commit-graph: Write incremental files, by derrickstolee
💛 #149 Fix MSVC support, at long last, by dscho
💛 #148 Windows: add a configuration location to be shared among different Git implementations, by dscho
💛 #147 Let's use the Win32 API more precisely, by dscho
💛 #112 Commit-graph write refactor (was: Create commit-graph file format v2), by derrickstolee

COMMITS

CONTRIBUTORS

Last week there was 1 contributor.
👤 gitster

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Add support for drop!/reword! commits to `git rebase -i`

This was already discussed briefly on the Git mailing list: https://public-inbox.org/git/alpine.DEB.2.21.1.1710151754070.40514@virtualbox/

In short, the reword! semantics are relatively easy in my mind: it would be called as

git commit --reword <commit>

would take staged changes (if any), otherwise implicitly --allow-empty,
then create a commit message to be edited in this style:

reword! <original-oneline>

<original-oneline>

<original-body>

This would be presented to the user in an editor (similar to git commit --amend).

Upon rebase --autosquash, it would work like a squash! but comment out all
previous commit messages, and also comment out the reword! line and the
empty line after that.

In case of multiple reword!, the last one would win, and it would
naturally integrate in any fixup!/squash! workflow.

What is more difficult is something else I frequently need: drop!
. That is, I want to explicitly mark a commit to be excluded in
subsequent rebase --autosquash. I guess the best way would be to
implement a

git revert --drop <commit>

that would work as if you called git revert -n <commit> && git commit -m 'drop! '"$(git show -s --oneline <commit>)", and upon rebase --autosquash, it would reorder like fixup!/squash!/reword!, replace the pickof the previous line (if it was apick) by dropand comment out the currentpick drop! ` line.

The reason why the semantics are more difficult in that case is that drop!
does not mix well with fixup!/squash!/reword!.

Fix flaky t5516-fetch-push (test case 79 - deny fetch unreachable SHA1, allowtipsha1inwant=true)

This seems to fail every once in a while, especially in Git for Windows' experimental MSVC job of our Azure Pipeline.

The issue seems to be that multiple processes report fatal errors at the same time, and our grep fails to verify that the expected error message was printed because they were interleaved. Example:

++ grep 'remote error:.*not our ref.*64ea4c133d59fa98e86a771eda009872d6ab2886$' err
++ echo 'error: '\''grep remote error:.*not our ref.*64ea4c133d59fa98e86a771eda009872d6ab2886$' 'err'\'' didn'\''t find a match in:'
error: 'grep remote error:.*not our ref.*64ea4c133d59fa98e86a771eda009872d6ab2886$ err' didn't find a match in:
++ test -s err
++ cat err
fatal: git uploadfatal: remote error: upload-pack: not our ref 6-pack: not our ref 64ea4c1334ea4c133d59fa98e86a771eda009872d6ab2886
d59fa98e86a771eda009872d6ab2886

Split the `linux-gcc` in our CI builds into two separate jobs

The linux-gcc job is a bit overloaded: it actually runs the test suite twice. Once with regular options, and once after setting all kinds of GIT_TEST_* options.

Since the running time is dominated by the test suite, not by the build, it would make more sense for those two runs to be run in parallel.

Therefore, let's split linux-gcc into two: a regular one, and on with all the extra options.

rebase -r: implement an equivalent to skip_unnecessary_picks()

Take this todo list, for example:

label onto

# Branch dscho
reset onto
pick a123 first
label dscho

# Branch avar
reset onto
pick b789 second
label avar

reset onto
merge -C c124 dscho
merge -C d314 avar

In other words: two branches, both one patch deep, both merged, one after the other. Now,
if you insert pick abc zeroth before the first pick, obviously the first branch would no longer be skippable, but the second one totally would be!

Since we are not performing a linear rebase, the skip_unnecessary_picks() function has no prayer of helping us here.

To remedy this, we would probably need code in pick_commits(), right where TODO_RESET is handled, that would:

parse the argument (this is currently done in do_reset() and would have to be refactored out) and pretend that it is HEAD,
look at the following command: if it is
- a pick, and if its parent agrees with HEAD, pretend that the pick was actually a reset, update the pretended HEAD and keep looking at the next command,
- a merge, and if its option was -C <orig-merge> (not lower-case -c!), and if its parent agrees with HEAD, and if its merge head(s) agree with the original merge commit's (if any), pretend that it was actually a reset <orig-merge>, update the pretended HEAD and keep looking at the next command,
- a label, perform it, but with the pretended HEAD, and keep looking for the next command,
- a reset, update the done and git-rebase-todo files and start the entire spiel from the top,
- otherwise perform the reset.
all while skipping, this code would need to take care of updating the done and git-rebase-todo files,
if a reset is necessary, and if it fails, the done and git-rebase-todo files should not be updated, but the original reset should be re-scheduled, and
since this adds quite a bit of code, it should probably be done in a separate function.

Instead of tapping into the TODO_RESET code block, it might make more sense, though, to start at the top of the loop, with whatever HEAD is at the moment.

Fix formatting of config values in documentation

Currently, in our config documentation, there are many inconsistencies with the way values are being marked up. This patch is a very small start but we should try to go through all of the config/*.txt files and mark up config values with surrounding backticks. This way, it'll be obvious that these are special values that a user should pay attention to.

In particular, true and false are major offenders for not being surrounded by backticks.

Investigate -Wformat-signedness

According to Peff, compiling with -Wformat-signedness might reveal some places where there output might be incorrectly formatted. (https://public-inbox.org/git/[email protected]/)

Compile Git with -Wformat-signedness and fix some (or all) of these warnings? Maybe consider turning on the warning in config.mak.dev after these are fixed?

rebase -ir: optionally auto-update local branches when executing `label` commands

The --rebase-merges mode allows users to rebase entire branch thickets, rather than only a linear commit history.

This feature can also be used to structure otherwise too-long patch series. Essentially the history looks like this:

o - A - B - C -D -E -F - G - H - I ...
  \          \             \
   ------------ first-batch - second-batch - ...

This makes sure that the whole feature can be developed from beginning to end, still keeping the "first-batch" in a shape that can be contributed individually to the Git mailing list. And then fixes (e.g. as a reaction to code review comments) can be integrated via git commit --fixup and then git rebase -ir.

However, the local branches first-batch and second-batch are not updated that way; They still have to be updated manually, which is tedious and error-prone.

How about an option to --rebase-merges that teaches label commands to see whether there is a local branch of the same name whose tip commit corresponds to the pre-rebase state of this label, and if such a branch was found, it is updated automatically (even updating a worktree if there is one, much like the receive.denyCurrentBranch=updateInstead feature does on push).

Weekly Digest (21 July, 2019 - 28 July, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 5 issues were created.
Of these, 0 issues have been closed and 5 issues are still open.

OPEN ISSUES

💚 #296 Add the REBASE_P prerequisite to two more test cases, by dscho
💚 #295 git svn: use rebase --rebase-merges instead of preserve-merges, by dscho
💚 #294 rebase -r: support merge strategies other than recursive, by dscho
💚 #293 Clean up more of the fall-out from dropping the scripted git rebase, by dscho
💚 #292 Create 'feature.*' config area and some centralized config parsing, by derrickstolee

NOISY ISSUE

🔈 #294 rebase -r: support merge strategies other than recursive, by dscho
It received 8 comments.

PULL REQUESTS

Last week, 10 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 10 pull requests were updated.
💛 #296 Add the REBASE_P prerequisite to two more test cases, by dscho
💛 #295 git svn: use rebase --rebase-merges instead of preserve-merges, by dscho
💛 #294 rebase -r: support merge strategies other than recursive, by dscho
💛 #293 Clean up more of the fall-out from dropping the scripted git rebase, by dscho
💛 #292 Create 'feature.*' config area and some centralized config parsing, by derrickstolee
💛 #290 Travis CI: Lint for Python syntax errors and undefined names, by cclauss
💛 #287 Reinstate support for Visual Studio, by dscho
💛 #219 Show an error if too-long paths are seen by git clean -dfx, by dscho
💛 #195 Drop support for git rebase --preserve-merges, by dscho
💛 #170 git add -i: add a rudimentary version in C (supporting only status and help so far), by dscho

COMMITS

CONTRIBUTORS

Last week there were 2 contributors.
👤 gitster
👤 peff

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

rebase -ir should warn (or error out) when commits were produced but not included in the final commit history

It is easy to do this:

pick something

reset somewhere else
pick something different
label this-is-my-branch

reset onto
merge this-is-my-branch

Obviously, the first pick was not included, by mistake.

So every time reset is executed, it should record the HEAD's hash in a growing list, and verify in the end that they are reachable (possibly using the refs/rewritten/* refs with name-rev to display nicer names).

This feature is similar in spirit to the warning when calling git checkout <branch> on a detached HEAD, where the user is told about the commits that would now be lost.

Migrated from https://bugs.chromium.org/p/git/issues/detail?id=32

range-diff: allow comparing branches to mboxes

A neat trick that could come in real handy at times would be to teach git range-diff to compare a local branch to a patch series that is provided as an mbox, e.g. to use git range-diff to see
whether a given patch series on the mailing list has made it into git.git's pu branch yet.

We could introduce range-diff mbox:<file> A..B for that, for example.

This would still require some scripting to e.g. pull down the mbox files of a patch series from https://public-inbox.org/git.

But it would solve the problem where current range-diff requires local branches, and therefore the user would have to find a commit where the patch series actually applies first.

format.outputDirectory should be based on the top of the repository

Currently, if one has format.outputDirectory set, the output directory will be relative to wherever they've run the command, not the top of the repository. We should make it relative to the Git repository instead so that users don't end up creating patches in unintended locations.

Or maybe the current behaviour is desired and we should just force users to use an absolute path if they don't want this to happen. This requires discussion.

Allow `reset <tag>` in `git rebase -ir`

The interactive rebase allows resetting HEAD to a specified revision, but it currently cannot handle tags. Let's fix that.

See https://public-inbox.org/git/[email protected]/

range-diff: add support for `--left-only` and `--right-only`

When contributing patches to Git, developers often need to compare in what form those patches made it upstream (e.g. to pick up fixes made by the Git maintainer in preparation for a new iteration). In such a case, a --left-only (or --right-only) option would come in real handy, to allow git range-diff --left-only <branch>..upstream/pu to skip commits that are upstream only (as there will be a metric ton of them).

rebase -i: implement --nested

The problem: during an interactive rebase, the user might realize that they want to redo (part of) the rebase, e.g. to resolve merge conflicts in a different way, say, by re-picking the latest three commits. Usually one would do git rebase -i HEAD~3, but that does not work because we are already in a rebase.

What I typically do is to create a throw-away worktree with git worktree add --detach, call git rebase -i HEAD~3 in that worktree, then copy the output of git rev-parse HEAD, then remove the worktree, and call git reset --hard <commit> in the original worktree.

That is fraught with danger, though, and very inconvenient.

The proposed solution: a new git rebase -i --nested [...] option. The --nested option would allow

This would generate the todo list as usual, but then prepend it to the current todo list. At the end, it would insert a new command that indicates that the nested rebase is now complete, and at the same time provides enough information to allow git rebase --abort to abort the nested rebase.

Some care will have to be taken in case of combining --nested with --rebase-merges: that will potentially override the onto label. In that instance, at the end of the nested rebase we would have to re-set the onto label to the original value.

It is not yet possible to call label <name> <commit>, only label <name>. We would have to introduce the former form to make it possible to re-set the onto label. Or use reset <commit> and then label <name>, but that is somewhat ugly.

Add documentation for "git commit --date=now"

There are some special strings that we can pass to --date in git commit, e.g. now, yesterday, noon, tea...

Personally I think --date=now is the most useful because I will use it with git commit --amend. I guess we don't have to make an exhaustive list of special strings in the documentation, but I think we can at least mention now.

Avoid duplicate entries in `git ls-files` for unmerged entries

This is a left-over bit: https://public-inbox.org/git/[email protected]/

Fix flaky "t0021.15 required process filter should filter data"

This test case seems to fail infrequently, most often in the linux-gcc job. See e.g. https://dev.azure.com/gitgitgadget/git/_build/results?buildId=9226&view=ms.vss-test-web.build-test-results-tab&runId=19716&resultId=100893&paneView=debug

Weekly Digest (14 July, 2019 - 21 July, 2019)

Here's the Weekly Digest for gitgitgadget/git:

ISSUES

Last week 4 issues were created.
Of these, 0 issues have been closed and 4 issues are still open.

OPEN ISSUES

💚 #290 Travis CI: Lint for Python syntax errors and undefined names, by cclauss
💚 #289 Fix rendering of text in angle brackets in README.md, by dilijev
💚 #288 ci: automatically build Git using MSVC in our Azure Pipeline, by dscho
💚 #287 Reinstate support for Visual Studio, by dscho

NOISY ISSUE

🔈 #287 Reinstate support for Visual Studio, by dscho
It received 11 comments.

PULL REQUESTS

Last week, 12 pull requests were created, updated or merged.

UPDATED PULL REQUEST

Last week, 12 pull requests were updated.
💛 #290 Travis CI: Lint for Python syntax errors and undefined names, by cclauss
💛 #289 Fix rendering of text in angle brackets in README.md, by dilijev
💛 #288 ci: automatically build Git using MSVC in our Azure Pipeline, by dscho
💛 #287 Reinstate support for Visual Studio, by dscho
💛 #285 Update gpg.txt to correct gpg --verify syntax, by rtmorgan
💛 #264 Adding git-ignore command, tests, and documentation., by tstone2077
💛 #254 [RFC] Create 'core.featureAdoptionRate' setting to update config defaults, by derrickstolee
💛 #220 mingw: support spawning programs with paths containing spacesnames, by dscho
💛 #219 Show an error if too-long paths are seen by git clean -dfx, by dscho
💛 #213 diff: release all handles before running external diff, by dscho
💛 #170 git add -i: add a rudimentary version in C (supporting only status and help so far), by dscho
💛 #145 Avoid spawning gzip in git archive, by dscho

COMMITS

CONTRIBUTORS

Last week there was 1 contributor.
👤 gitster

STARGAZERS

Last week there were no stargazers.

RELEASES

Last week there were no releases.

That's all for last week, please 👀 Watch and ⭐ Star the repository gitgitgadget/git to receive next weekly updates. 😃

You can also view all Weekly Digests by clicking here.

Reconcile `dir_exists()` and `is_directory()`

In https://github.com/git/git/blob/v2.22.0-rc1/builtin/clone.c#L885-L889, there is a function with the short and sweet name dir_exists() (which is the obvious companion to file_exists()). However, it is marked as static, so nobody can use it outside builtin/clone.c.

There is also is_directory() which obviously tries to do the very same, but it uses a name that few developers will think of when they see file_exists() and look for the equivalent function to see whether a given directory exists.

Let's reconcile these, probably by renaming is_directory() to dir_exists() and using it also in builtin/clone.c.

Having grep support the -o option like GNU grep et al

We have most of the code for this already in the form of our color hi-lighting, it would mostly just be a matter of "just print out the stuff you'd have colored", with the small exception that if you have more than one match on a line they should be printed out on their own lines.

Copied from: https://public-inbox.org/git/[email protected]/

Turn the `fetch_if_missing` global into a field of `struct repository`

It is high time that we reduced our globals, especially in light of finally having struct repository and working towards using it everywhere.

rebase -i: error on fixup/squash when the corresponding pick was skipped

Imagine this todo list:

pick a123 this will conflict
fixup b124 want to fix up

Now, as the oneline suggests, let's assume that the pick results in merge conflicts, and let's further assume that the user decided to skip it via git rebase --skip.

Obviously, the fixup should not blindly amend HEAD at this point, so we will probably want to error out with a helpful message.