profile
viewpoint

gibfahn/dot 24

My dotfiles and setup scripts

a-roberts/dragonball 2

The greatest version of table football

gibfahn/gnt 2

Github Notifications in Tabs - Open a page of GH notifications in new tabs

gibfahn/codelets 1

Misc coding solutions

annaet/node-red-labs 0

Node-RED labs on the use of the Watson Developer Cloud services

gibfahn/bazel 0

a fast, scalable, multi-language and extensible build system

gibfahn/bazel-glossary 0

A Glossary of Bazel Terminology

issue commentrust-lang/rust

Error messages about modules are confusing to beginners

The error message seems particularly confusing as extern crate util should never be needed in 2018 edition crates, so this error will I think be always incorrect.

jmikkola

comment created time in a day

issue commentcli/cli

Use a consistent title for the Pull Request editor file

I'm not sure how I would configure vim to use the correct filetype settings.

I believe you can do it the same way vim does it for other types, see: https://github.com/neovim/neovim/blob/045e86474b75992834f54157b04ea2a941ab03a7/runtime/filetype.vim

With the filenames I proposed above you could do:

au BufNewFile,BufRead gh_*.md	setf gh-markdown
gibfahn

comment created time in a day

issue openedjunegunn/fzf

Maintain cursor position when deleting text

  • [x] I have read through the manual page (man fzf)
  • [x] I have the latest version of fzf
  • [x] I have searched through the existing issues

Info

  • OS
    • [x] Mac OS X
  • Shell
    • [x] zsh

Problem / Steps to reproduce

It would be really nice if when you delete a character, fzf would keep the cursor on the same line (as it will always still be matched).

This would allow better use of fzf for browsing things like git history, before filtering.

For example:

  1. I'm looking for a commit, I know it was near the 1.0.0 commit.
  2. I open my interactive git log to search for it.
  3. I type 1.0.0 to find the 1.0.0 commit.
  4. I press <kbd>Backspace</kbd> to delete the 1.0.0, so I can see the surrounding commits.
  5. Unfortunately fzf preserves the line I was on (i.e. 4th line down) rather than the text I was looking at (i.e. 100th row of text) so I can't achieve this feature.

Not sure if this should be the default or an option, maybe there are good uses for the default (like not wanting your cursor to get stuck all the way at the bottom).

See https://asciinema.org/a/304635 for a visual example:

asciicast

created time in 2 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 69080780a6de36460cdc8775b5003553c58d88d2

brew: remove python2 as it's apparently no longer available by default ``` ❯❯❯ Installing: brew packages. Warning: 'python2' formula is unreadable: No available formula with the name "python2" Error: No available formula with the name "python2" Warning: homebrew/core is shallow clone. To get complete history run: git -C "$(brew --repo homebrew/core)" fetch --unshallow ==> Searching for a previously deleted formula (in the last month)... Error: No previously deleted formula found. Error: No similarly named formulae found. ==> Searching taps on GitHub... ==> Searching for similarly named formulae... ==> Searching taps... Error: No formulae found in taps. Installing python2 has failed! Homebrew Bundle failed! 1 Brewfile dependency failed to install. ```

view details

Gibson Fahnestock

commit sha 1cbaf50092bd2f7c3e8f4796065e68a5bd9fb8de

gibrc: fix `b` alias to point to real `bazelisk` The multiple levels of aliasing didn't seem to be working, and there's no real need for this "cleverness".

view details

Gibson Fahnestock

commit sha b6db960726cb16bed791a6277bf11239f3dc5920

git: make default-branch set the branch if it's not already defined Avoids having to go manually run the command to update.

view details

push time in 2 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha d6523ce76b12bacfb3f108f3f8494edd9fdb88db

Add Apache License

view details

push time in 3 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha f57131aab275a1de5dc0b5a7c6bbc12b5cbedb1a

Remove nightly-only features

view details

Gibson Fahnestock

commit sha 4056be8ade7a57a90d7fa8887c5d8989bca9ee46

Move LICENSE to LICENSE-MIT

view details

push time in 3 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 6563484209c81f3abddd8dd94c3dbc213a19eec7

editorconfig: fix typo in editorconfig

view details

push time in 4 days

delete branch gibfahn/up-rs

delete branch : tmp

delete time in 4 days

issue openedcli/cli

Use a consistent title for the Pull Request editor file

Describe the feature or problem you’d like to solve

Related to https://github.com/cli/cli/issues/390, but right now the opened PR or issue file seems to be of the form $TMPDIR/630826291.md.

hub uses .git/PULLREQ_EDITMSG

There are a couple of differences:

  1. Adding the .md extension -> This is 💯
  2. Not having a consistent file name or prefix -> This is less desirable as you can't configure global settings in your editor.
  3. Not being in the git repo itself -> probably cleaner, but means you can no longer use things like an editorconfig to set things like line lengths and other formatting on a per-repo basis. On the other hand I'm not aware of many things you'd want to configure on a per-repo basis.

Proposed solution

Add a consistent prefix, e.g. $TMPDIR/630826291.md -> $TMPDIR/gh_pr_630826291.md and $TMPDIR/gh_issue_630826291.md.

This will allow users to configure their editors appropriately, e.g. having a custom GFM syntax type.

created time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 64ca46e39911ac9d6d61022d4b6769e9b41942f4

squash!

view details

push time in 4 days

create barnchgibfahn/up-rs

branch : tmp

created branch time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 79b775f27afb19845c5dbf50e4d6440de49f08d6

Switch to clippy-preview and fix suggestions Among other things, clippy recommended splitting up the giant linking function (which is a good idea).

view details

Gibson Fahnestock

commit sha 0673fd49ba57b2bd4e500d5714254d5b38d1332b

editorconfig: fix typo in editorconfig

view details

push time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha cb3ac55e88b9875378ac20476744bdd23bcb4997

Add MIT license

view details

push time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 0f721b1f33003cbeeb255770fd3a26409b9a15e2

CI: try adding --no-fail-fast to cargo test

view details

push time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha e38da46560e667058d01ee80daa5bbfe6326a324

Create rust.yml Just run the basic build and test for now

view details

push time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 1a694b5f1d5884218d4dcd218d863de0bf208c40

Update README.md with new linking command

view details

push time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 1494c84ea81a7488505852318bc5382f3411b061

Fix up tests, get clippy to pass Also remove ripgrep dependency from tests by adding the ignore crate (which is ironically part of ripgrep itself).

view details

Gibson Fahnestock

commit sha 83ea3d2b4f5a391cd72f02a2b1284611c643073f

Add first set of contributing docs Brain-dump of all the useful information I've gleaned so far.

view details

push time in 4 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha b7c1fb94abd1c7e1b99c14a4ba22c79ff2427178

tests: use the integration test setup that ripgrep uses Allows you to avoid dead code warnings for things in tests/common.rs by building everything as a single test binary.

view details

Gibson Fahnestock

commit sha fcd74c04d8983efac9f72edd1b179f315933222f

Remove color-backtrace as it doesn't support std::backtrace::Backtrace yet Refs: https://github.com/athre0z/color-backtrace/issues/24

view details

Gibson Fahnestock

commit sha 32359dfed9c330ecb81ea86c4a14374e77cde4fa

tests: move testutils to a separate crate Looks like this is the best option for shared utils that are used in both src/ and tests/. I tried having a module in src/ that was `#[cfg(test)]`, but it isn't compiled when the module is built for integration tests. Refs: https://stackoverflow.com/questions/37993886/where-should-i-put-test-utility-functions-in-rust Refs: https://stackoverflow.com/questions/44539729/what-is-an-idiomatic-way-to-have-shared-utility-functions-for-integration-tests Refs: https://stackoverflow.com/questions/46867652/cannot-import-a-module-in-an-integration-test

view details

Gibson Fahnestock

commit sha 156b6be6d7e919be9e442983743ba27ea1c2c97b

tests: add more info when unexpected symlink found

view details

Gibson Fahnestock

commit sha 6a741bae9f76b4f8105c685463d6f8dfead2328d

Remove mistakenly added symlink No idea how this got added.

view details

Gibson Fahnestock

commit sha e6fed198b1fb35137bdcf35e3360eaf7a986c5b8

Respect env_logger's RUST_LOG_STYLE env var Also set it in tests so logs are coloured. Update test runner to pass correct args to linker.

view details

Gibson Fahnestock

commit sha 8e4ebff77b15b712fe4d409c502fa2eefadb03ab

tests::z_style: check testutils crate Extract duplicated command running code.

view details

Gibson Fahnestock

commit sha 93ec0895488dc58cf49a282cf111d9c6a1033431

Update `up link` args, log backup dir contents on exit

view details

Gibson Fahnestock

commit sha eeacd8e23bbdfa1fa9146065eaaecc534a4221ee

Fix tests, move testutils::assert_* functions to testutils::assert::

view details

push time in 4 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 171cca3d767befade4ba2d49d99f0a5a99ad8eed

git: fix Ctrl-o commands, make `g li` open `g shi` If you're interactively viewing a log of commits, you can now Ctrl-o to interactively view the files and diffs for the selected commit (Ctrl-c to exit). e.g. I used this to find when a deleted and renamed file was changed or introduced: ``` g li --follow -- tests/fixtures/link/hidden_and_nested/home_dir/nothing ```

view details

push time in 5 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 2932aef4242a62789fa0bab21050a350b17c253b

zsh: add default fzf shortcuts for wrapping and sorting Needed these when working with git fzf mappings today.

view details

Gibson Fahnestock

commit sha b0b7088f7ac4e29675ec329bda2964846231eae5

mac: set com.apple.swipescrolldirection as a bool not an int Seems to be required to get things to work properly.

view details

Gibson Fahnestock

commit sha b97ef2c3f003ab86e0a48e8fd61e0e893b27dd42

git: update shi/di interactive log/diff to handle complex cases - Set `--stat=1000` to stop wrapping (truncation). Don't want to use `--numstat` because it doesn't look as nice in the fzf window. - Handle rename lines: `{ => }`. e.g. convert: ``` tests/testutils/mod.rs | 3 +- tests/{test_config.rs => config.rs} | 10 +- tests/fixtures/fake_home_dir_with_upconfig/.config/{dot/dot.toml => up/up.toml} | 0 tests/fixtures/{test_link => link}/backup_files/dotfile_dir/already_in_backup | 0 ``` to: ``` tests/common/mod.rs tests/test_config.rs tests/config.rs tests/fixtures/fake_home_dir_with_upconfig/.config/dot/dot.toml tests/fixtures/fake_home_dir_with_upconfig/.config/up/up.toml tests/fixtures/test_link/backup_files/dotfile_dir/already_in_backup tests/fixtures/link/backup_files/dotfile_dir/already_in_backup ``` That way when something is moved we show both the old and the new path so git shows us the renames.

view details

push time in 5 days

push eventgibfahn/up-rs

Gibson Fahnestock

commit sha 411d6ee124c525ce69d9ef7f215c8459b64b9200

Update dependencies, rename tests, remove ~/.up/ path support

view details

push time in 6 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha b5d1df1091f94c55263e21e86becab8e4dffb58a

link: swap FROMPATH and TOPATH These were the wrong way round.

view details

Gibson Fahnestock

commit sha 563ffca93591cc8bf36370d3171b738742c0db29

git: add `g di` and change existing to `g dni` Interactive diff is useful for the same reasons as interactive `log` or `show`.

view details

Gibson Fahnestock

commit sha f522fa0f04fb9710142a4ca342b63ade507a2ad8

cargo: add `cargo watch` aliases

view details

push time in 6 days

push eventgibfahn/dot-rs

Gibson Fahnestock

commit sha f0d6d172b4c1d00880031739cd7f854fe55e64e1

Add an editorconfig

view details

Gibson Fahnestock

commit sha 82699e6fc9feeceb996168b3050d7aaf369bd7f6

Run on nightly for now

view details

Gibson Fahnestock

commit sha 4cbc6a1a1a483e2f8229119c33da347e58a4fdd8

rustfmt: merge imports on `cargo fmt`

view details

Gibson Fahnestock

commit sha 759eac887b0891aeca88f7659d107428e1289d71

Change name dot-rs -> up-rs, migrate to anyhow/thiserror Also bumps dependencies.

view details

push time in 6 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha cc570685f8990f387460565505ad6c6cf73f7486

git: improve ^G^H and `g li`, and add `g shi` for interactive show

view details

push time in 7 days

issue commentpqrs-org/Karabiner-Elements

Does not work on MacOS Catalina

karabiner_grabber and karabiner_observer are in my input monitoring and Karabiner-Elements.app has full disc access

That's all I have as far as I can tell. By re-grant permissions I mean that on the update they added karabiner_grabber and karabiner_observer to the Input Monitoring, and I had to grant those.

Short of uninstalling and reinstalling (and maybe removing and re-adding to System Preferences), I don't have any other ideas.

phoffer

comment created time in 7 days

PR opened ap/vim-readdir

Document mappings

I found these in the plugin source, hopefully the descriptions are accurate.

I've had to look these up a couple of times now, so figured I might as well write them down.

+29 -3

0 comment

1 changed file

pr created time in 7 days

push eventgibfahn/vim-readdir

Gibson Fahnestock

commit sha 1ab679b77e218ec0679d4db0e0749f634b951524

Document mappings I found these in the plugin source, hopefully the descriptions are accurate.

view details

push time in 7 days

create barnchgibfahn/vim-readdir

branch : document_mappings

created branch time in 7 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 03553d497ea5fe2506aa6c78aaa745a0f212b915

gibrc: set BAT_THEME globally This way we don't have to set it in individual delta commands.

view details

Gibson Fahnestock

commit sha 31c6d993a6b77c3e918525426ceebef522c54e95

vim: add case-insensitive search Usually you want an exact match, but not always.

view details

Gibson Fahnestock

commit sha 978ef458ebc17b473b1efdfc017f83e74f772dab

gibrc: remove broken `alias find="gfind --color=auto"` gfind doesn't seem to support this flag.

view details

Gibson Fahnestock

commit sha d58a396cf240f25f5fc03aba86ffb70181942adc

Work around lack of ability to force color in delta Refs: https://github.com/dandavison/delta/issues/101

view details

Gibson Fahnestock

commit sha ec57eca6b1d54f5be9702e5279f93eef455b32b8

copyq: use Clipboard not clipboard as default tab Also adds a command to show changes in a plist file, e.g. ```console $ g diff-plist dotfiles/Library/Preferences/com.copyq.copyq.plist 274c274 < 0 => "&clipboard" --- > 0 => "&Clipboard" 276d275 < 2 => "&Clipboard" ```

view details

push time in 7 days

delete branch gibfahn/vim-readdir

delete branch : master

delete time in 7 days

delete branch gibfahn/vim-readdir

delete branch : assets

delete time in 7 days

create barnchgibfahn/vim-readdir

branch : oldmaster

created branch time in 7 days

startedsolarhess/rules_build_secrets

started time in 7 days

issue commentpqrs-org/Karabiner-Elements

Does not work on MacOS Catalina

Yep, no issues resuming from sleep either. Not sure what issues others are seeing, but probably needs more info than just "broken for me on Catalina".

phoffer

comment created time in 7 days

issue commentpqrs-org/Karabiner-Elements

Does not work on MacOS Catalina

I haven't seen this at all on Catalina, I did re-grant permissions when I updated, but since then everything has been working, and I use a bunch of complex modifications (and put it to sleep all the time) without any issues. Also installed on a new 16" MBP and everything works fine there too.

phoffer

comment created time in 7 days

issue openeddandavison/delta

Support overriding tty detection (forcing interactive mode)

Simplest repro: GIT_PAGER=delta git show | cat gives the same output as GIT_PAGER=cat git show as it detects that the output is not a tty.

In bat you can set --decorations=always --color=always to force color and decorations to be output.

Refs: https://github.com/sharkdp/bat/issues/539 and https://github.com/sharkdp/bat/issues/448

This is useful for scripts, for example I have a more complicated version of this in my git config:

  lp = "!git log --graph --color=always --format=\"%C(auto)%h%d %s %C(black)%C(bold)<%an> %cr%C(auto)\" \"$@\" \
        | fzf --ansi -m --no-sort --reverse --tiebreak=index \
        --preview \"(grep -o '[a-f0-9]\\{7\\}' | head -1 | xargs -I % sh -c 'git -c core.pager=delta show --color=always %') <<< {}\" #"

I would like to be able to force delta to show colour and page in the preview window.

Workaround

This seems to work (although I'm not entirely sure why):

  lp = "!git log --graph --color=always --format=\"%C(auto)%h%d %s %C(black)%C(bold)<%an> %cr%C(auto)\" \"$@\" \
        | fzf --ansi -m --no-sort --reverse --tiebreak=index \
        --preview \"(grep -o '[a-f0-9]\\{7\\}' | head -1 | xargs -I % sh -c 'git -c core.pager=cat show --color=never % | delta') <<< {}\" #"

it shows color, shows the decorations, and pages when you scroll in the preview window.

Basically it changes:

-GIT_PAGER=delta git show | cat
+GIT_PAGER=cat git show | delta | cat

when what I really want is:

GIT_PAGER="delta --decorations=always --color=always --paging=always" git show | cat

created time in 9 days

issue openeddandavison/delta

Support `git stash show`

Looks like git stash show doesn't get formatted by delta:

image

Script to reproduce

export GIT_PAGER='delta --theme=TwoDark'
mkdir tst
cd tst
git init
printf "bar\nbaz\n" >foo_diff
printf "bar\nbaz\n" >foo_stash

git add foo*
git commit -m "Initial Commit"

printf "bat\n" >> foo_stash
git stash
printf "bat\n" >> foo_diff

git diff # Works with delta
git stash show # Doesn't.

created time in 9 days

delete branch gibfahn/ripgrep

delete branch : update_brew_instructions_readme

delete time in 11 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 2a44060953146936ba1453c150c975eeb06311cf

setup: remove rust-analyzer cargo install Binaries are now available, and the VS Code and coc.nvim (neovim) extensions seem to download it themselves.

view details

Gibson Fahnestock

commit sha b666400fa566c96400cbd6057fa6e84297c9d0af

mac: note that we're updating brew Currently just always run these commands.

view details

Gibson Fahnestock

commit sha a5ab7d6825745d0b047f8cfec8c9a2fc52df2d4f

brew: add buildozer CLI for Bazel

view details

push time in 11 days

delete branch gibfahn/bazel

delete branch : note_macos_output_root

delete time in 12 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha cc8157e4b76536789dc7fd41e5d89c230146b3cd

mac: run brew update as well as brew upgrade Seems to be needed to update taps to their latest remote versions.

view details

push time in 12 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha ec50c988ebef55f342abb556013777e820535735

git: fix repeated `g bdd` and use delta for `g lp` and ^G^H + ^G^R - `g bdd` would fail when the previous branch no longer exists (e.g. when you use it twice in a row). Now it falls back to changing to the default branch. - `g lp`, ^G^H, and ^G^R were using less rather than delta, which is much nicer.

view details

push time in 13 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha ddd95c0a31e6e25e9ad4d2de9186102ae2fdbf0f

brew: add new gh CLI

view details

Gibson Fahnestock

commit sha d213f9eae9d5ccd516d85940036cf0129c8f142a

vim: add support for shell code blocks in markdown

view details

push time in 14 days

startedcli/cli

started time in 14 days

delete branch gibfahn/homebrew-cask

delete branch : fix_copyq_on_case_sensitive_fs

delete time in 15 days

issue commentdandavison/delta

Show binary files changes in delta output

Checked and it works, thanks for fixing it so quickly!

gibfahn

comment created time in 15 days

startedSycnex/Windows10Debloater

started time in 15 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha f0294def9b8a1def05c6d0bcfe0741981a77a466

git: add a `g lpr` option to create a log for pasting into PRs

view details

Gibson Fahnestock

commit sha 52cbb91329de21f63bdc893ae8af4fddb06ab7dc

copyq: add copyq preferences to dotfiles Run `copyq info config` to get the path to the config file. Usually `~/.config/copyq/` on non-Windows. Refs: https://github.com/hluk/CopyQ/issues/428#issuecomment-538730229

view details

Gibson Fahnestock

commit sha 7f00388631bcb0bfa6a5f1743e845059072bf3b8

Use Hasklig font with Nerd Font Extensions Font is called Hasklug to avoid something copyright-related.

view details

Gibson Fahnestock

commit sha 39f3290e4e874cd405fd7286d551242f4d38ca3f

git: add lpr and mfp commands for working with GitHub PRs - `g lpr`: nicest log format I can find to paste into a Pull Request. - `g mfp`: merge in changes that were committed in the GitHub UI. The previous `lpr` format didn't show links in commit messages (because they were \<pre\> formatted.

view details

Gibson Fahnestock

commit sha 6a81fa47aac059c9071514a7addc994fd9715989

zsh: update to new p10k version that supports @{push} deltas Refs: https://github.com/romkatv/powerlevel10k/issues/395

view details

Gibson Fahnestock

commit sha 675f508c38f0ca5e529243fc4b62030e1e39abbd

brew: add vlc for everyone

view details

push time in 15 days

PR opened Homebrew/homebrew-cask

copyq: fix shimscript for case-sensitive filesystems

Otherwise it fails for me with:

$ copyq --version
/usr/local/bin/copyq: line 2: /Applications/CopyQ.app/Contents/MacOS/copyq: No such file or directory
/usr/local/bin/copyq: line 2: exec: /Applications/CopyQ.app/Contents/MacOS/copyq: cannot execute: No such file or directory

$ ls /Applications/CopyQ.app/Contents/MacOS/
/Applications/CopyQ.app/Contents/MacOS/CopyQ

$ brew cask reinstall Casks/copyq.rb
...

$ copyq --version
CopyQ Clipboard Manager v3.10.0

After making all changes to the cask:

  • [x] brew cask audit --download {{cask_file}} is error-free.
  • [x] brew cask style --fix {{cask_file}} reports no offenses.
  • [x] The commit message includes the cask’s name and version.
  • [x] The submission is for a stable version or documented exception.
+1 -1

0 comment

1 changed file

pr created time in 15 days

push eventgibfahn/homebrew-cask

Gibson Fahnestock

commit sha 4ecb2f84d65d8af17079d9f1607d7b256922afe4

copyq: fix shimscript for case-sensitive filesystems Otherwise it fails for me with: ```console $ copyq --version /usr/local/bin/copyq: line 2: /Applications/CopyQ.app/Contents/MacOS/copyq: No such file or directory /usr/local/bin/copyq: line 2: exec: /Applications/CopyQ.app/Contents/MacOS/copyq: cannot execute: No such file or directory $ ls /Applications/CopyQ.app/Contents/MacOS/ /Applications/CopyQ.app/Contents/MacOS/CopyQ $ brew cask reinstall Casks/copyq.rb ... $ copyq --version CopyQ Clipboard Manager v3.10.0 ```

view details

push time in 15 days

create barnchgibfahn/homebrew-cask

branch : fix_copyq_on_case_sensitive_fs

created branch time in 15 days

issue openeddandavison/delta

Show binary files changes in delta output

When I use GIT_PAGER=less git show I see the following section for each binary file:

diff --git a/foo b/foo
new file mode 100644
index 0000000..4088911
Binary files /dev/null and b/foo differ

When I use delta with GIT_PAGER="delta --theme=TwoDark" git show, I don't see any output for foo, which is a binary file.

It would be great to see something that shows that there was a binary file and that it has changed, what less shows would be fine.

Quick repro:

mkdir -p tst && cd tst
git init
echo -n -e \\x48\\x00\\x49\\x00 > foo
git add foo
GIT_PAGER="delta --theme=TwoDark" git diff --staged
GIT_PAGER=less git diff --staged

Output:

$ mkdir -p tst && cd tst
$ git init
Initialized empty Git repository in ~/tmp/tst/.git/
$ echo -n -e \\x48\\x00\\x49\\x00 > foo
$ git add foo
$ GIT_PAGER="delta --theme=TwoDark" git diff --staged
$ GIT_PAGER=less git diff --staged
  diff --git a/foo b/foo
  new file mode 100644
  index 0000000..7a42ff1
  Binary files /dev/null and b/foo differ

created time in 15 days

push eventgibfahn/homebrew-cask

Randall

commit sha 64f8fcb43e4f3b6279088b608076db3ce7047890

Update nuclear from 0.6.0,8595ed to 0.6.1,6fa42c (#75839)

view details

Randall

commit sha 21d7590a1087ea2c853274375e7c1a4c3a6bef9e

Update dcp-o-matic-kdm-creator from 2.14.21 to 2.14.23 (#75857)

view details

Randall

commit sha b5893afe9d4a09393ec1ca910248781944b072f8

Update dcp-o-matic-batch-converter from 2.14.21 to 2.14.23 (#75856)

view details

Randall

commit sha 3934f52932d25b59b97e97353bdb23b26c51c716

Update dcp-o-matic-player from 2.14.21 to 2.14.23 (#75855)

view details

Randall

commit sha e0380a034648f02c8dbacd9eaa7647fe3f29c785

Update dcp-o-matic from 2.14.21 to 2.14.23 (#75854)

view details

Randall

commit sha 53ff81fab7ed75b88f307ca016ec862158852767

Update dcp-o-matic-encode-server from 2.14.21 to 2.14.23 (#75853)

view details

Randall

commit sha 62b2d7ca713598ffa38d8b03b7293cf982ef0a11

Update zoomus from 4.6.14768.0101 to 4.6.17383.0119 (#75852)

view details

Garth Kidd

commit sha 7d91c8759260f85ff7e25ac1c5957e0c32363b21

Correct freecad 0.18.4 build 16146 SHA256... again (#75851) More than a dozen new bundles since the version or its build number was last bumped. Not sure what's happening upstream. SHA matches that downloadable from their [release page]. [release page]: https://github.com/FreeCAD/FreeCAD/releases/tag/0.18.4

view details

Randall

commit sha e8d1469009721ab98aefaf0768ff2dd8f05ca12b

Update retroarch from 1.8.2 to 1.8.4 (#75850)

view details

Cap.雪ノ下八幡

commit sha a39e8abe659c961e37b967f516af9b97dbb99ccc

Add IINA+ 0.4.12 (#75822)

view details

Ralph Martin

commit sha 4dc0f31ad2930067d9dfa5f4fa76d8041cf253ef

Update amadeus-pro (#75858)

view details

Erik Minekus

commit sha d14969bfc61c03d0068de15e60ea17c4b7cf32bf

Update toggl from 7.5.3 to 7.5.16 (#75861)

view details

Bogdan Popescu

commit sha 39bf19ba9a336e7013e45ade49545c2bacf61560

Update dash to 5.1.1 (#75863)

view details

Bogdan Popescu

commit sha 79615f8b5bc67c91e6a39d1b96d64549adfb8a6f

Update dash to 5.1.1 (#75865)

view details

suschizu

commit sha 81684580e5b4c9c47afd4ab8092cdaf30cce1f85

Update popo from 3.2.0 to 3.2.1 (#75868)

view details

suschizu

commit sha c32dd540db6f95cdc710595d89c4f449d7d3bceb

Update jamf-migrator from 5.0.1 to 5.0.3 (#75867)

view details

suschizu

commit sha 325f42866e4d316f9cbac0cc8e4603c520be1bdb

Update bluestacks from 4.140.12.2902,4804ca37dfe9b788ed01b47f9d19bdd9 to 4.140.13.2803,7b51e6488db874e4f7fb6a253786c8f1 (#75872)

view details

lwouis

commit sha d46b6a70d44aa2673a8de5f6f9a550e9ebffc291

Update alt-tab 2.3.2 with correct macOS minimum version (#75859)

view details

Kargins Rem

commit sha bf01fcf41a649f27300ac034a2a77fd2b8aa37d3

Add qView.app v3.0 (#75860)

view details

suschizu

commit sha da1229f04b68dbce5dda00a7b12c353f518f871b

Revert "Change version from 4.6 to 4.6.18 progressive-downloader.rb (#75829)" (#75862) This reverts commit 274ce08853f3a102bd68e6b4d91615529959d48b.

view details

push time in 15 days

delete branch gibfahn/homebrew-cask

delete branch : chrome-binary

delete time in 15 days

issue commentromkatv/powerlevel10k

Allow showing delta from @{push} as well as @{upstream}

Updated and tried this and it works perfectly, thanks!

gibfahn

comment created time in 18 days

issue commentromkatv/gitstatus

Show delta from @{push} as well as @{upstream}

https://github.com/romkatv/gitstatus/commit/11198dfe3b07388df0c85afc2c09bed5adb7f3f8 works perfectly, thanks!

gibfahn

comment created time in 18 days

issue commentzdharma/zinit

Parallelize `zplugin update --all`

Could maybe a donation be granted for this? Like $5 or $10?

Sure! Done.

gibfahn

comment created time in 18 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha ab84a9b9c19376b857637b8c3f30e8d5cfee1e6e

brew: add watch command-line tool

view details

push time in 21 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha a12cb3b006bfaad361bc8a03caa481a22c642f2e

git: more UX improvements following push remote changes - `g pr` should raise a PR against `@{upstream}` by default, as that's what the "upstream tracking branch" means for PR branches. - Add `g ppr` as the most common flow is "push to fork, then raise PR" - Change "head-remote" to "up-remote" as it's really just the remote part of `@{upstream}`. - Add "up-branch" to return the other part of `@{upstream}`, note the `2-` to return all the rest of the branch, so `origin/gib/pr/foo` would return `gib/pr/foo` not just `gib` (branches can have `/` in them).

view details

push time in 23 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha f4187df8e90b06f647e8b675b806b800875eb0cb

git: more UX improvements following push remote changes - `g pr` should raise a PR against `@{upstream}` by default, as that's what the "upstream tracking branch" means for PR branches. - Add `g ppr` as the most common flow is "push to fork, then raise PR" - Change "head-remote" to "up-remote" as it's really just the remote part of `@{upstream}`. - Add "up-branch" to return the other part of `@{upstream}`, note the `2-` to return all the rest of the branch, so `origin/gib/pr/foo` would return `gib/pr/foo` not just `gib` (branches can have `/` in them).

view details

push time in 23 days

push eventgibfahn/bazel

John Cater

commit sha 1db2faa17150d19da4dd411a5f9504925014eaed

Re-enable //src/test/res:res_test on Windows. We are far past Bazel release 0.29.0. Fixes #9148. Change-Id: I6cbd907f9a3e63c2337b367626441d8698f56d60 Closes #10666. Change-Id: I6cbd907f9a3e63c2337b367626441d8698f56d60 PiperOrigin-RevId: 291923311

view details

aehlig

commit sha 6a3147f26d29f126f403d0c444ce67e4d883ca8d

Update documentation with respect to exports_files All source files used outside their package should be exported explicitly. PiperOrigin-RevId: 291923415

view details

elenairina

commit sha 5eade37562bd84ea6354feb124a00779ed96e3c0

Add native dependencies artifacts to the test runner action's tools. This change is useful for the persistent test runner to include the native dependencies in the worker's key and re-build a worker if the native dependencies have changed. PiperOrigin-RevId: 291927096

view details

Yun Peng

commit sha 63f0b36e61864cc805f50618ac72f76440ba4b84

Add build scripts for release Closes #10671. PiperOrigin-RevId: 291927177

view details

ulfjack

commit sha c428ee515380b90c9105df4b88bd397af9a84fd5

Use a ForkJoinPool for async include scanning Previously, switching to a FJP was rolled back due to introducing a performance regression in some cases. While I was unable to reproduce the regression due to other changes in Bazel, this ties the switch to the async include scanner flag. I have carefully benchmarked that, and I think that switching to a FJP is required to roll out the async include scanner. PiperOrigin-RevId: 291931477

view details

Nikolay Shelukhin

commit sha a9b82978acccb7249bcef09637860ec95ce3e48a

Windows environment variable name can contain brackets Bazel cuts off environment variables with brackets in name. It leads to a problem for me. I have Visual Studio 2017 with workloads "Desktop C++ development", "TypeScript 3.1 SDK", "Python Web Support" installed. I use Visual Studio toolchain configuration script similar to `tools/cpp/windows_cc_configure.bzl`. My script in contrast to original finds more tools and checks a result of `VCVARSALL.BAT` also. I noticed that `VCVARSALL.BAT` fails with error: ``` [ERROR:typescript.bat] TypeScript was not added to PATH since a valid installation was not found [ERROR:VsDevCmd.bat] *** VsDevCmd.bat encountered errors. Environment may be incomplete and/or incorrect. *** [ERROR:VsDevCmd.bat] In an uninitialized command prompt, please 'set VSCMD_DEBUG=[value]' and then re-run [ERROR:VsDevCmd.bat] vsdevcmd.bat [args] for additional details. [ERROR:VsDevCmd.bat] Where [value] is: [ERROR:VsDevCmd.bat] 1 : basic debug logging [ERROR:VsDevCmd.bat] 2 : detailed debug logging [ERROR:VsDevCmd.bat] 3 : trace level logging. Redirection of output to a file when using this level is recommended. [ERROR:VsDevCmd.bat] Example: set VSCMD_DEBUG=3 [ERROR:VsDevCmd.bat] vsdevcmd.bat > vsdevcmd.trace.txt 2>&1 ``` It happens because `typescript.bat` can't access `ProgramFiles(x86)` environment variable: **typescript.bat**: ``` ... if exist "%ProgramFiles(x86)%\Microsoft SDKs\TypeScript\3.1" ( set "PATH=%ProgramFiles(x86)%\Microsoft SDKs\TypeScript\3.1;%PATH%" set _TypeScript_Found=1 ) ... ``` Exit code of `VCVARSALL.BAT` is ignored in original `tools/cpp/windows_cc_configure.bzl` so it is not critical bug but it can lead to unobvious errors for users who find external tools (TypeScript compiler, for example) with own scripts. Closes #10605. PiperOrigin-RevId: 291932741

view details

adonovan

commit sha e8d482496f705304d21c938ec44dc578997ae942

bazel syntax: move Selector{List,Value} into lib.packages The Bazel 'select' function is no longer in Starlark.UNIVERSE. It is pre-declared in all Bazel file environments (BUILD, .bzl, WORKSPACE). This is a new category (though depset belongs in it too), exposed as StarlarkLibrary.COMMON. Also: - make various methods private - improve the annotation doc comment - remove legacyNamed for select(x={...}). There are no uses in Google's depot. A later change will merge Selector{List,Value} into a single data type that represents a node in a binary tree whose branch nodes are deferred '+' operations and whose leaves are dictionaries. While removing what looked like unnecessarily qualified class names, I discovered that BuildType.SelectorList is yet a third (!) Starlark value class representing the same concept. PiperOrigin-RevId: 291938455

view details

ulfjack

commit sha e411fa74176cdc54f84d62c189d042e9fb6d5c79

Add a flag to declare coverage dir as tree artifact This adds an experimental flag that changes the behavior of tests when run with coverage such that the _coverage directory is declared as a tree artifact. This makes it significantly easier to debug coverage runs when sandboxing (the default) or remote execution are enabled. Without this, the files are implicitly ignored, and only the coverage.dat file is brought back into the output tree. If for some reason the additional steps to generate the coverage.dat file fail, then that the file is empty or non-existent. This also provides a workaround for languages that cannot currently generate lcov format coverage data. By bringing back the individual files, the user can use a post-process to merge the data. It's not perfect, but it's simple and effective. I used it myself to debug a coverage issue due to incorrect file paths w/ gcc on MacOS. Unfortunately, the coverage collector script generates a number of temporary files in the _coverage directory which we generally do not want (or need) to bring back into the output tree. This will have to be addressed separately. PiperOrigin-RevId: 291940271

view details

jcater

commit sha c864f876dc7e6d7758103047f721a605d1d5899b

Automated rollback of commit 37aeabcd39fe326d1c4e55693d8d207f9f7ac6c4. *** Reason for rollback *** Causes all actions to use the "sandbox" strategy, effectively disabling remote builds in all cases. *** Original change description *** Split SpawnStrategy from ActionContext. This change removes SpawnStrategy from the ActionContext hierarchy. Spawn strategies thus are officially no longer queryable from any place in the execution phase (they already hadn't been but it was not obvious at all). Instead all access to them must be mediated by the SpawnStrategyResolver (formerly ProxySpawnStrategy) which uses the SpawnStrategyRegistry (formerly SpawnActionContextMaps) to do its work. Other kinds of action contexts are accessible... *** PiperOrigin-RevId: 291943951

view details

adonovan

commit sha ed7d241f10c5ebbb7243a86479d2157b5d012b19

Automated rollback of commit e8d482496f705304d21c938ec44dc578997ae942. *** Reason for rollback *** Skydoc's immensely fragile simulation of the .bzl environment no longer includes 'select', causing tests to fail, and there is no simple fix because Skydoc is not allowed to depend on lib.packages. *** Original change description *** bazel syntax: move Selector{List,Value} into lib.packages The Bazel 'select' function is no longer in Starlark.UNIVERSE. It is pre-declared in all Bazel file environments (BUILD, .bzl, WORKSPACE). This is a new category (though depset belongs in it too), exposed as StarlarkLibrary.COMMON. Also: - make various methods private - improve the annotation doc comment - remove legacyNamed for select(x={...}). There are no uses in Google's depot. A later change will merge Selector{List,Value} into... *** ROLLBACK_OF=291938455 PiperOrigin-RevId: 291966278

view details

philwo

commit sha ddfc7d5c83cbde0dc4ec92bd0f0b8e24239a4fb2

Make bazel_rules_cc_test and bazel_rules_java_test compatible with repository overriding. RELNOTES: None. PiperOrigin-RevId: 291972443

view details

adonovan

commit sha d8ee69ab679594cb47fa8c5fa24b309a2a23b54c

bazel syntax: move Selector{List,Value} into lib.packages (Second attempt at commit e8d482496f705304d21c938ec44dc578997ae942, rolled back in commit ed7d241f10c5ebbb7243a86479d2157b5d012b19, because it broke Skydoc tests. The only new code is in SkydocMain.) The Bazel 'select' function is no longer in Starlark.UNIVERSE. It is pre-declared in all Bazel file environments (BUILD, .bzl, WORKSPACE). This is a new category (though depset belongs in it too), exposed as StarlarkLibrary.COMMON. Also: - make various methods private - improve the annotation doc comment - remove legacyNamed for select(x={...}). There are no uses in Google's depot. A later change will merge Selector{List,Value} into a single data type that represents a node in a binary tree whose branch nodes are deferred '+' operations and whose leaves are dictionaries. While removing what looked like unnecessarily qualified class names, I discovered that BuildType.SelectorList is yet a third (!) Starlark value class representing the same concept. PiperOrigin-RevId: 291982062

view details

felly

commit sha cda604b1bf8a97184de7fa2a0b57285079f7d40d

Allow overriding of bes internal retries. Not intended as an end-user interface. PiperOrigin-RevId: 291987090

view details

plf

commit sha 3545cb1b5509ad185071c900c621dbfa08767fcb

C++: Fix linker_input not being noneable in create_linking_context RELNOTES:none PiperOrigin-RevId: 291989441

view details

twerth

commit sha 9fd125535ba7ed3f3667a19aa42c340b1c9849af

Use allInputs.size() as size hint for the ActionInputMap. Without NestedSet-on-Skyframe, getInputDeps returned a map whose size was equal to the number of all inputs of that specific action. With NestedSet-on-Skyframe, getInputDeps returned a singleton map which was growing over the course of action executions. The resulting size hint was growing with the number of actions being executed (or the total number of inputs), resulting in a ~30% GC tax for big builds. After this CL, the size hint corresponds again to the expected size of the inputs, also for NestedSet-on-Skyframe. RELNOTES: None PiperOrigin-RevId: 291998196

view details

adonovan

commit sha df05243768049192c10e21316c4ee2c56bc28a10

bazel syntax: eliminate calls to StarlarkThread.getGlobals during evaluation Calls during evaluation want the innermost enclosing Starlark source file on the call stack, which is now provided by the ugly helper function Module.ofInnermostEnclosingStarlarkFunction. All remaining calls to getGlobals are called before or after execution, so StarlarkThread call/return no longer needs to save/restore this.globalFrame (now called this.module). These calls will be eliminated once the StarlarkThread and Module construction API is simplified. This is a breaking API change for Copybara. References to StarlarkThread have been replaced by Module where the one appeared to be a proxy for the other. No behavior change. PiperOrigin-RevId: 292000083

view details

cparsons

commit sha da5ee491b7af41d18d035e42228f518077164f65

Rollback "Partially fix MacOS llvm..." due to rules_cc breakage RELNOTES: None. PiperOrigin-RevId: 292005414

view details

Googler

commit sha c7703a9c63a8f8d365b5935acc8daf391ece77d6

Automatic code cleanup. PiperOrigin-RevId: 292107739

view details

jcater

commit sha aea8dbe942a2489d48e1bbc171413b3e0391cded

Enable extended regexps in expect_log and expect_not_log. PiperOrigin-RevId: 292142060

view details

jcater

commit sha b3250e50a24fa9175e2541758690b9eac1392cce

Automated rollback of commit aea8dbe942a2489d48e1bbc171413b3e0391cded. *** Reason for rollback *** Change breaks test detection leading to sh_test targets passing with the message: 0 / 0 tests passed. *** Original change description *** Enable extended regexps in expect_log and expect_not_log. PiperOrigin-RevId: 292159521

view details

push time in 24 days

pull request commentbazelbuild/bazel

Note default cache directory root on macOS

Gentle nudge @aiuto , I believe the existing feedback has been addressed.

gibfahn

comment created time in 24 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 1b52d931e015d05b5186b1cc7131280b22eab21e

manual: note things that shouldn't auto-start

view details

push time in 24 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 8bad023b57b0a3c00def8608d1fa0acf303e078e

manual: note things that shouldn't auto-start

view details

push time in 24 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha adeb1a91b7e1295a58ea8dd91feb900973a4ff23

mac: add manual NTFS setup instructions

view details

Gibson Fahnestock

commit sha 37aa749b6add41be70e02bd01c83e112e218a98e

mac: debug log the actual `defaults write` actual command run Otherwise with the function indirection it's a little painful to work out the command being run.

view details

Gibson Fahnestock

commit sha 07439e5b450087fdc8bf5d32eff8da1cde85aa85

mac: set battery percentage, fix killall checks and add debug logging Was previously always killing the Dock due to a copypasta error. Make it so we now have logging to see why we kill things.

view details

Gibson Fahnestock

commit sha 51095193a873d32f1599f285097a801a72c96d56

brew: note the `brew bundle cleanup` command to run This way you can see what would be removed, and if it's a lot can run it manually.

view details

Gibson Fahnestock

commit sha a111f0c25c18313f2ef526d24fcf2e1dc3c086ab

brew: add work and keyboard formulae

view details

push time in 24 days

push eventgibfahn/qmk_firmware

William Chang

commit sha 156fd4e9693e05fb12933bddb49e1a387fa1fab3

Initial commit

view details

William Chang

commit sha cd379c69a03d339e1eb87646824c2b49cb4db982

Modified keymap and updated readme.

view details

William Chang

commit sha a5337b349552da25009f05fff127cd2228329912

fixed numbering in readme

view details

William Chang

commit sha 5a86db22592cc23d42803e08fc83e0e22f100cc8

Update README.md

view details

William Chang

commit sha fa73d43818e3e2a7dcd4b0982ca1e7a33b25c499

Update README.md

view details

William Chang

commit sha 7f8b0906c60e3d659089bfe7ab4ea7dfa65eb9a4

Update README.md

view details

William Chang

commit sha 5b80e10b8239b038719355441989ef392572cdd6

Update README.md

view details

William Chang

commit sha 472060d3338b509f82fc4c1cfaa6c0483109dfed

Update README.md

view details

William Chang

commit sha dc3a8ddb6be228d5385b1d10982715c6072d740b

Update README.md

view details

William Chang

commit sha 42cb78f98eb3ab0e4de638ec8e0db99fcab8d4a4

Made requested changes in PR.

view details

William Chang

commit sha b42ca9bc5fd13b2d5d3cefa8f329d78611bea53f

Merge branch 'master' of https://github.com/billiams/qmk_firmware

view details

William Chang

commit sha da1dc28d31f1d48d558ad88bd2b5b5cdd4fa15cb

Rename README.md to readme.md

view details

William Chang

commit sha a41e6804fcf97ac85a1b21c923df12a2fcaf9f7a

swapped media control keys

view details

William Chang

commit sha d7c5cf6e5bfe0be17173f42aabdd981036c4c940

Merge branch 'master' of https://github.com/billiams/qmk_firmware

view details

William Chang

commit sha 89e8e0d2774064362fa5e73d19df3b1e760d7016

updated right alt key to apply shift key for emacs

view details

William Chang

commit sha 57a6ea11df685d84a1ea07953e88f224ce2b24f7

updated R_Alt to remove shift

view details

William Chang

commit sha 7b0200660ee23b4bec278747ec010d16575bce42

Updated readme

view details

William Chang

commit sha 4b2d3288d013b1a71ea25402224c4a8225a099e9

resolved conflicts

view details

William Chang

commit sha 9ce186860e78d5afe350636cb7445e18c2577f33

Merge branch 'qmk-master'

view details

William Chang

commit sha 8d8f2d09cc228a459be8348c5ce61327e9862be8

Merge pull request #2 from billiams/feature/billiams Updated readme

view details

push time in 24 days

push eventgibfahn/dot

Gibson Fahnestock

commit sha 771e9943aa51c9295197c7a40c4b326cc962584e

brew: move Brewfiles to $XDG_CONFIG_HOME/brew/ and fix update script Allows installing multiple Brewfiles at once, and shows what things would be cleaned up if only Brewfile things were installed.

view details

Gibson Fahnestock

commit sha 48b05b7d30d9f0b9643686cefa24e8bf167d0bba

setup: improve logging functions

view details

push time in a month

delete branch gibfahn/rbenv

delete branch : re-add_zsh_completions_and_remove_fish

delete time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 58b56cc72d55de0bf2f114f576a1c17a86b0e81d

bazel: use a persistent disk cache

view details

Gibson Fahnestock

commit sha 5a5f0b94336c35fde75359e69cb2366f3094fa0d

brew: install steam by default

view details

Gibson Fahnestock

commit sha ea65d38fd94fbbd88ddcc3f117ec633da92e294e

git: add a `git fork` command to set up forks - Create a new fork with remote name "fork". - Create a new branch "oldmaster" mirroring "up/master". - Make "oldmaster" the default branch. - Clean all other branches and tags from the fork.

view details

Gibson Fahnestock

commit sha d51e4d8e036a2c378a108e31fd42f2b8e792e86c

brew: add wechat, fzf, buildifier

view details

Gibson Fahnestock

commit sha 5f98ccf2c6ff4a1a266662c627ce508b46d39bfa

vim: add indentation file for Bazel Otherwise indentation is broken for `.bzl` files. Refs: https://github.com/sitaktif/dotfiles/commit/4fa3c71a81c4d64ccb8b59e4f27226252967f4b8

view details

push time in a month

PR opened rbenv/rbenv

Re-add zsh completion script and remove fish completion script

Looks like the zsh completion script was removed by mistake.

Fixes: https://github.com/rbenv/rbenv/issues/1215 Refs: https://github.com/rbenv/rbenv/commit/569d464d361ba02ed77f20cff3399adf70953a58

+18 -23

0 comment

2 changed files

pr created time in a month

create barnchgibfahn/rbenv

branch : re-add_zsh_completions_and_remove_fish

created branch time in a month

delete branch gibfahn/rbenv

delete branch : version-aliases

delete time in a month

delete branch gibfahn/rbenv

delete branch : user-gems

delete time in a month

delete branch gibfahn/rbenv

delete branch : master

delete time in a month

delete branch gibfahn/rbenv

delete branch : gh-pages

delete time in a month

create barnchgibfahn/rbenv

branch : oldmaster

created branch time in a month

fork gibfahn/rbenv

Groom your app’s Ruby environment

fork in a month

issue openedrbenv/rbenv

zsh completion script removed

It looks like completions/rbenv.zsh was removed by mistake in https://github.com/rbenv/rbenv/commit/569d464d361ba02ed77f20cff3399adf70953a58 , whose commit message says it was attempting to remove completions/rbenv.fish to fix https://github.com/rbenv/rbenv/issues/1212.

cc/ @mislav

created time in a month

push eventgibfahn/bazel

ichern

commit sha 585647e7c0564b7be708bb6985570dd2a3b00ea9

Add --experimental_ninja_actions flag for Ninja execution Closes #10571. PiperOrigin-RevId: 289808802

view details

plf

commit sha cc8809ba2ac0ab13ec6c47b4f5422798dcfc5d60

C++: Change type of attribute linked_statically_by The attribute linked_statically_by part of --experimental_cc_shared_library has a NODEP_LABEL_LIST type to avoid cyclic dependencies. By changing the type from string we get Label syntax validation for free as well as relative paths working. RELNOTES:none PiperOrigin-RevId: 289812404

view details

ulfjack

commit sha 5e1847c23edb6c5860498af7a246ade7d6b21ad5

Update tests to work with NestedSet not Iterable (part 3) PiperOrigin-RevId: 289813755

view details

leba

commit sha d2343d90f19dc9a3f3a227e1c0c113fde60a7747

Fix JSON format for aquery analysis_v2.proto. The JSON format doesn't work with the streamed output since by nature it needs to be grouped by message type. This CL adds a monolithic output handler for analysis_v2.proto that can be used for JSON format. RELNOTES: None PiperOrigin-RevId: 289815268

view details

laurentlb

commit sha 8cb336ac3dc4d95005ab5b5243b51bf88cd6fdfe

Delete the flag --incompatible_remap_main_repo https://github.com/bazelbuild/bazel/issues/7130 RELNOTES: The flag --incompatible_remap_main_repo is removed. PiperOrigin-RevId: 289822459

view details

ulfjack

commit sha a928a5f42ba8e4e646ca7106d015db833afba09c

Update code for NestedSet not implementing Iterable This is the last significant code change before making NestedSet not implement Iterable. PiperOrigin-RevId: 289824349

view details

pcloudy

commit sha 9db93bc96b898b9614926d6347d785b6047e581d

Automated rollback of commit 4162cc54ba0b128b616c0bd05b65bf9ad5e66f2e. *** Reason for rollback *** Broken Bazel CI in Downstream pipeline https://github.com/bazelbuild/bazel/issues/10588 *** Original change description *** proto_lang_toolchain: Add original sources of ProtoInfo to blacklistedProtos Fixes #10484 Closes #10493. PiperOrigin-RevId: 289828390

view details

philwo

commit sha 817fa5ec2cc6a8d56282f7536e53235707da18a8

Improve WORKSPACE file generation in shell tests. By passing in the workspace name to create_workspace_with_default_repos, we can remove custom code for this from a few tests. This also ensures that the tests get a fresh WORKSPACE file with known contents when they call that method. The old code simply appended a few repositories to any existing WORKSPACE file which caused problems when overriding repositories. RELNOTES: None. PiperOrigin-RevId: 289830653

view details

laszlocsomor

commit sha 50ebca165d1d3d2a919d2e7af155e299283c0be5

Docs: discourage using MSYS2 to run Bazel Update the "Using Bazel on Windows" page to discourage running Bazel from MSYS2 Bash. See https://github.com/bazelbuild/bazel/issues/10573 RELNOTES: Windows: we now discourage running Bazel from MSYS2 because of a newly found bug (#10573) PiperOrigin-RevId: 289833002

view details

laszlocsomor

commit sha be0e36f6a379cc84d2758fd455ae231327419010

incompatible_windows_bashless_run_command: remove Remove this flag and every occurrence of it. It was flipped in Bazel 1.0, so nobody should be using it in production. Fixes https://github.com/bazelbuild/bazel/issues/8240 RELNOTES[INC]: The --[no]incompatible_windows_bashless_run_command flag is no longer supported. It was flipped in Bazel 1.0 PiperOrigin-RevId: 289835848

view details

ulfjack

commit sha b74cf3020f4ae711f4ebd7f8cf3932760a1df91b

NestedSet no longer implements Iterable This makes the Java version match the Starlark depset to make it more explicit when / where nested sets are flattened. In many, but not all, of these cases, the flattening implies quadratic cpu/memory consumption. PiperOrigin-RevId: 289836851

view details

laszlocsomor

commit sha ba87ba9a9b9dd30156ad6dcbd71c1d86b474f683

incompatible_windows_native_test_wrapper: remove Remove this flag and every occurrence of it. It was flipped in Bazel 1.0, so nobody should be using it in production. Fixes https://github.com/bazelbuild/bazel/issues/6622 RELNOTES[INC]: The --[no]incompatible_windows_native_test_wrapper flag is no longer supported. It was flipped in Bazel 1.0 PiperOrigin-RevId: 289840656

view details

plf

commit sha d9ca5f9e4c92d932a71b692a6a45de763644ab93

C++: Add linked_statically_by_all to cc_library RELNOTES:none PiperOrigin-RevId: 289844278

view details

Yun Peng

commit sha 7e4c9d5d21f051f096d65af637f53470cc94d116

Make --verbose_failures work again --verbose_failures is broken since Bazel 0.28.0, Bazel won't print the failing command for local execution even --verbose_failures is enabled. This change fixes the problem and adds a test for it. Fixes https://github.com/bazelbuild/bazel/issues/10572 Closes #10581. PiperOrigin-RevId: 289854746

view details

Yannic Bonenberger

commit sha 1821bd5e7af0649f5e29861ad97a3c89c34fa511

Rollforward of commit 4162cc54ba0b128b616c0bd05b65bf9ad5e66f2e. proto_lang_toolchain: Add original sources of ProtoInfo to blacklistedProtos This fixes the problem reported in #10588 Fixes #10484 Closes #10592. PiperOrigin-RevId: 289857885

view details

Benjamin Peterson

commit sha e0d325b44b9941931aa3c0d94fb2e446d60ba98c

"Stream" action cache digest computations. Now that ActionAnalysisMetadata.getInputs() returns a NestedSet, we don't need allocate an intermediate (path -> metadata) map during action digest computation. Encapsulate the logic for this in a OrderIndependentHasher class. This class also caches fingerprints and digest arrays to avoid allocations. Update the size constant in CompactPersistentActionCacheTest.testToStringIsntToBig because the digests are all 32 byte arrays instead of 1 byte arrays. Additionally, fix ActionCache.Entry so it's stateless and immutable by making a builder class for it. Delete CompactPersistentActionCacheTest.testEntryToStringIsIdempotent because the property it wants to proven by the type system. Closes #10537. PiperOrigin-RevId: 289858351

view details

ulfjack

commit sha 7c20af3ee4888437ad08d43fab7d4919874b1dec

Refactor test case summarization This improves accuracy in cases where no or limited test case information is available. I also simplified the API and made the implementation of TestSummary only do a single transitive traversal over the TestCase tree. Fixes #8188. PiperOrigin-RevId: 289859356

view details

laszlocsomor

commit sha 0f2c626038358688e7c5013fdb807ce8a3f2ae69

Evaluator: delete this unused interface Nothing implements the nested Factory interface so that can be deleted. Only ParallelEvaluator implements Evaluator so the interface can be deleted and the few references to Evaluator updated. I seem to remember that there used to be various Evaluator implementations, but evidently these are no more. We could rename ParallelEvaluator to just Evaluator but that'd touch more than 50 files, which would be too much effort for negligible benefit. RELNOTES: none PiperOrigin-RevId: 289872590

view details

schmitt

commit sha f8f59bc6804e3609132818286c2797bc4d0fd7c3

Replace @ExecutionStrategy with module-based registration. Strategies and other action contexts used an annotation to indicate their identifying type (for example: "I'm something that can write a file for an action") and commandline identifiers (for example: "local", "remote"). This approach has a number of drawbacks: * Reflection is required to obtain the values, making it harder to understand control flow/dependencies. * Classes have to be created just so an annotation can be put upon them (everything else is handled as constructor parameters passed to a common base class). * Identifying information is not visible in places usually associated with configuration (Bazel modules) which merely register the class. This change removes the annotation-based propagation of identifying information and instead adds registration methods (called in Bazel modules that would previously simply register an annotation context instance) that take the data explicitly. RELNOTES: None PiperOrigin-RevId: 289881876

view details

schmitt

commit sha 31d494ac02db0b11ed79ca20d9e1f560274ef117

Delete @ExecutionStrategy. This annotation was made superfluous in a previous CL but is only removed in this separate CL to keep changes small(er). RELNOTES: None. PiperOrigin-RevId: 289882670

view details

push time in a month

push eventgibfahn/qmk_firmware

Gibson Fahnestock

commit sha 2a1214fc1979b6aba3521098b306acdd0c382158

Copy the bracket []{} layout from Keyboardio layout

view details

push time in a month

issue commentromkatv/powerlevel10k

POWERLEVEL9K_SHOW_RULER not showing ruler since a9715367d4547dabaa724f9907bc971078453528

Yep, fixed, that was super-fast, thanks!

Thanks for the high quality bug report!

Thanks for making such an awesome prompt. I've tried a lot over the years (including rolling my own), and this is somehow the fastest and at the same time the most customisable! I especially love Instant Prompt, such a simple idea, but gamechanging.

gibfahn

comment created time in a month

issue closedromkatv/powerlevel10k

POWERLEVEL9K_SHOW_RULER not showing ruler since a9715367d4547dabaa724f9907bc971078453528

On the latest master commit I am no longer seeing the ruler between prompts.

I have the following ruler config:

  # Ruler, a.k.a. the horizontal line before each prompt. If you set it to true, you'll
  # probably want to set POWERLEVEL9K_PROMPT_ADD_NEWLINE=false above and
  # POWERLEVEL9K_MULTILINE_FIRST_PROMPT_GAP_CHAR=' ' below.
  typeset -g POWERLEVEL9K_SHOW_RULER=true
  typeset -g POWERLEVEL9K_RULER_CHAR='▁'        # reasonable alternative: '·'
  typeset -g POWERLEVEL9K_RULER_FOREGROUND=240

A git bisect suggests that the ruler stopped showing in a9715367d4547dabaa724f9907bc971078453528 .

If this isn't reproducible let me know and I'll try to pull out a minimal config to repro.

image

closed time in a month

gibfahn

issue closedromkatv/powerlevel10k

"closing brace expected" output

Occasionally my prompt gets into this state:

Usually a zplugin update --all fixes things, which suggests it's something to do with some commits of p10k. I've seen it twice though, which is odd.

_p9k_reset_prompt:5: closing brace expected
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       2020-01-27 09:19:33 ⇡3 add_docs ~/w/r/ci
_p9k_reset_prompt:5: closing brace expected
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       2020-01-27 09:19:33 ⇡3 add_docs ~/w/r/ci
_p9k_reset_prompt:5: closing brace expected
-- console output produced during zsh initialization follows --

_p9k_maybe_dump:16: closing brace expected
_p9k_maybe_dump:16: closing brace expected
_p9k_maybe_dump:16: closing brace expected
_p9k_maybe_dump:16: closing brace expected
Updating plugin romkatv/powerlevel10k
 ... 
Updating 9586e24..3bca7e8

Not the end of the world, as it has been fixable so far, just raising this in case it's useful to anyone.

closed time in a month

gibfahn

issue commentromkatv/powerlevel10k

"closing brace expected" output

Which zplugin command do you use to load powerlevel10k?

https://github.com/gibfahn/dot/blob/435cb5ed97055452784c5934970b6ac811444606/dotfiles/.config/gibrc#L303-L304

zplugin ice depth=1
zplugin light romkatv/powerlevel10k

On which comment of p10k does it happen?

Updating 9586e24..3bca7e8

Does this mean that you've experienced errors quoted above while synced to 9586e24?

That seems to be what zplugin is telling me, yeah.

What do you need to do to trigger this error? Can you reproduce it if you start zsh without sourcing regular configuration files (zsh -df) and then manually source p10k from there?

I'm not sure, it seems to happen randomly, without me having done anything.

The problem here is that I haven't been able to reproduce it on purpose. If I go back to that commit again I don't see this issue. Once I've fixed it I can't reproduce it again. The two times it has happened were weeks apart.

I'll close this for now as I can't reproduce. I can try to reproduce with the manual sourcing of p10k if it happens again. It seems quite possible that this is an issue elsewhere in the stack, that just manifests itself in p10k.

gibfahn

comment created time in a month

issue commentromkatv/powerlevel10k

POWERLEVEL9K_SHOW_RULER not showing ruler since a9715367d4547dabaa724f9907bc971078453528

FWIW I saw your comment on https://github.com/romkatv/powerlevel10k/issues/383#issuecomment-567549523 :

By the way, few people use ruler nowadays because "prompt gap" seems to work better.

And while that's nice, I don't think it gives the appearance I'm looking for of a minimal prompt with a blank line and a ruled line separating prompts, all in 2 lines. The ruler is invaluable when scrolling back through terminal history.

So while the prompt gap seems like a great feature, I hope the ruler stays in as an option.

image

gibfahn

comment created time in a month

issue openedromkatv/powerlevel10k

POWERLEVEL9K_SHOW_RULER not showing ruler since a9715367d4547dabaa724f9907bc971078453528

On the latest master commit I am no longer seeing the ruler between prompts.

I have the following ruler config:

  # Ruler, a.k.a. the horizontal line before each prompt. If you set it to true, you'll
  # probably want to set POWERLEVEL9K_PROMPT_ADD_NEWLINE=false above and
  # POWERLEVEL9K_MULTILINE_FIRST_PROMPT_GAP_CHAR=' ' below.
  typeset -g POWERLEVEL9K_SHOW_RULER=true
  typeset -g POWERLEVEL9K_RULER_CHAR='▁'        # reasonable alternative: '·'
  typeset -g POWERLEVEL9K_RULER_FOREGROUND=240

A git bisect suggests that the ruler stopped showing in a9715367d4547dabaa724f9907bc971078453528 .

If this isn't reproducible let me know and I'll try to pull out a minimal config to repro.

image

created time in a month

issue openedromkatv/powerlevel10k

"closing brace expected" output

Occasionally my prompt gets into this state:

Usually a zplugin update --all fixes things, which suggests it's something to do with some commits of p10k. I've seen it twice though, which is odd.

_p9k_reset_prompt:5: closing brace expected
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       2020-01-27 09:19:33 ⇡3 add_docs ~/w/r/ci
_p9k_reset_prompt:5: closing brace expected
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       2020-01-27 09:19:33 ⇡3 add_docs ~/w/r/ci
_p9k_reset_prompt:5: closing brace expected
-- console output produced during zsh initialization follows --

_p9k_maybe_dump:16: closing brace expected
_p9k_maybe_dump:16: closing brace expected
_p9k_maybe_dump:16: closing brace expected
_p9k_maybe_dump:16: closing brace expected
Updating plugin romkatv/powerlevel10k
 ... 
Updating 9586e24..3bca7e8

Not the end of the world, as it has been fixable so far, just raising this in case it's useful to anyone.

created time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 435cb5ed97055452784c5934970b6ac811444606

mac: re-add non-HARDCORE things Eve uses - coreutils is needed to provide `dircolors`, which allows the LS_COLORS env var to be set and used. - Eve likes the extra trackpad sensitivity.

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 46740ecf3621433f4ac50f23ad75e2623fd5de44

brew: use the bazelbuild tap for bazelisk Upstream bazel is broken (builds with Java 8), and we need the `bazel` -> `bazelisk` symlink because scripts often directly shell out to the `bazel` in the path.

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 46ce3d780a945351784d470ee3f729ca5d78b4e1

zsh: remove corporate-specific Node.js cert setting Moving this to my internal dotfiles.

view details

push time in a month

issue commentnodejs/build

remove gibfahn from build and release wg and teams

Thanks for cleaning up after me, sorry I didn't do it myself! Feel free to emeritize me from any team you think reasonable.

sam-github

comment created time in a month

issue openedbazelbuild/bazelisk

Bazelisk subprocesses should prepend the correct bazel to the $PATH

Many bazel rules end up running shell scripts that invoke bazel, which resolves to the first bazel binary in the path.

For example:

https://github.com/bazelbuild/rules_jvm_external/blob/0a7cc6a0b6764232a0ddd31ad87b489e1d47b166/private/pin.sh#L5

This should resolve to the bazel that bazelisk downloaded, otherwise you can end up using a different bazel binary in these subcommands. This causes confusing errors.

I believe the most natural fix would be for bazelisk to prepend a directory containing the relevant bazel binary to the $PATH.

created time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha c3b4dd24096de47696a222966f16ab73bc4f11d0

karabiner: fix karabiner on Eve's laptop

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha dbbb641a22080292804a3ff9882dc5dde677a5f3

git: fix git name and email handling Avoids having to commit changes to the dotfiles repo just to change a name.

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 9fac1a3f4f99b7ba910f18e9dd3f64faa731b5e7

update: don't prompt for HARDCORE, set defaults Unless overridden, only run HARDCORE if running as user "gib".

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 734991ac155d3c454c9159733ecf12de448f3cff

brew: install Node.js by default Used in internal scripts.

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha e3bc05a31f26319e0395e95005432f2b2ffeab09

brew: move delta to required section

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha f031aa10dd5dca86053f5d67220afe478c3018e8

karabiner: add profile for Eve Just changes Caps Lock -> Left Shift for now.

view details

push time in a month

push eventgibfahn/dot

Gibson Fahnestock

commit sha 9f75df04d57b8d704bcb5c1153df9caf3aeab23c

mac: resort brew packages If not running in HARDCORE mode, the average user (i.e. my family) need a different set of packages available.

view details

Gibson Fahnestock

commit sha 02f78803b1076bed3e5a076137eadcf179ecff60

mac: pull out non-HARDCORE defaults to always be set

view details

Gibson Fahnestock

commit sha a475d09bf3d70470a97af22f4e75f9b070227aaa

mac: remove manual instructions that are now set with `defaults`

view details

Gibson Fahnestock

commit sha 483db75a6f08b1dde683ca993480644387292816

unix: remove tooling I no longer use

view details

push time in a month

more