You can do this but adding another remote i.e. ¹:

git remote set-url --add --push origin git@gitlab.com:hmajid2301/go-routinely.git

Make sure to add both repositories with this command

Then when you go git push it will push to all repositories.

You can double check with:

goroutinely on  fix/ci [$] via 🐹 v1.25.5 via ❄  impure (nix-shell-env)
❯ git remote show origin
Welcome to this knot!
* remote origin
  Fetch URL: git@git.haseebmajid.dev:majiy00.tngl.sh/go-routinely
  Push  URL: git@git.haseebmajid.dev:majiy00.tngl.sh/go-routinely
  Push  URL: git@gitlab.com:hmajid2301/go-routinely.git
  HEAD branch: main

Well turns out it seems to be the way NixOS works with the firewall, you can read all the details here ¹. Where the poster explains it really well.

My understanding is the following:

This changes RPF (reverse path filtering) from “strict” to “loose” mode:

• Strict: Reply must go back out the exact same interface it came in on
• Loose: Reply can go out any interface, as long as there’s a valid route back to the source

Solution

But the fix is enabling this option: networking.firewall.checkReversePath = "loose";. Please make sure you understand exactly what you are doing when enabling this.

Sometimes the app store “App Lounge” gets stuck updating apps, my old solution was to delete the app and reinstall it but that can be a pain. Then I discovered a slightly better solution.

• Go to settings
• Apps & Notifications
• Find App Lounge
• Then go to Storage & cache
• Clear storage and “Delete app data”

You will need to log in again to the app, but the update process shouldn’t get stuck anymore hopefully.

[^1] https://community.e.foundation/t/applounge-updates-stuck/47296/11

The fix ended up being pretty simple ¹:

Nautilus > Preference > Show thumbnails.

Set it to "On this computer only" for only seeing thumbnails on your system, .

If u need it to do same for a remote server, select "All files"

Or since we are managing this via NixOS and want to do things declaratively we can do do the following in our Nix config ².

{
dconf.settings = {
  "org/gnome/nautilus/preferences" = {
    show-image-thumbnails = "always";
  };
};
}

That’s it, in your SMB share you should now be able to view thumbnails for all of your videos.

Recently I have been trying to update the emojis when I run nh. Particularly the nom output.

The old output was like so:

To fix this update your nh derivation like so:

environment.systemPackages = with pkgs; [
  (pkgs.nh.override {
    nix-output-monitor = pkgs.nix-output-monitor.overrideAttrs (old: {
      postPatch = old.postPatch or "" + ''
        substituteInPlace lib/NOM/Print.hs \
          --replace 'down = "↓"' 'down = "\xf072e"' \
          --replace 'up = "↑"' 'up = "\xf0737"' \
          --replace 'clock = "⏱"' 'clock = "\xf520"' \
          --replace 'running = "⏵"' 'running = "\xf04b"' \
          --replace 'done = "✔"' 'done = "\xf00c"' \
          --replace 'todo = "⏸"' 'todo = "\xf04d"' \
          --replace 'warning = "⚠"' 'warning = "\xf071"' \
          --replace 'average = "∅"' 'average = "\xf1da"' \
          --replace 'bigsum = "∑"' 'bigsum = "\xf04a0"'
      '';
    });
  })
];

Note you will no longer be able to use the default nix cache to build this. So everytime nh updates you will need to manually build it.

That’s it!

Background

In my app, I show the user 25 feedbacks per page, and they can delete them which opens a modal to confirm the action. If the user presses the delete button in the modal I want to send a HTTP DELETE request to an endpoint like /feedback/{id}, but the ID is dynamic, as we are using one modal for all the feedbacks vs having one modal per one feedback.

I am using HTMX, generated using templ and using AlpineJS for some frontend interactivity, where it doesn’t make sense to send a request back to the server.

Solution

The solution is relatively easily in the end

We have a button like this which triggers the modal to open, this is rendered in templ via Go. Hence the fmt.Sprintf function. Here we are setting a variable which will be used by AlpineJS.

<button
    class="btn btn-ghost btn-xs text-neutral"
    @click={ fmt.Sprintf("currentItemId = '%s'; document.getElementById('delete-modal').showModal()", item.ID) }
    aria-label="Delete Feedback"
>
    <i class="text-lg hgi hgi-solid hgi-delete-02"></i>
</button>

In one of the parent div make sure we set the x-data attribute, which matches the same variable we set on the @click.

<div class="" x-data="{ currentItemId: '' }">

Finally here is a simplified version of the button in the modal when pressed will send the delete request. The key line is x-effect="currentItemId;htmx.process($el)", we need HTMX to reevaluate the hx-delete attribute. Else the request will go to /feedback/ without the ID.

<button
    x-bind:hx-delete="'/feedback/' + currentItemId"
    x-effect="currentItemId;htmx.process($el)"
    hx-target="#feedback_list"
    hx-swap="outerHTML"
>
    Delete
</button>

That’s it, now you can set the endpoint dynamically, generally speaking I don’t need to use this pattern much. As most of the time this can be set when we render the template as they are usually static.

Appendix

• Github: https://github.com/mvolkmann/htmx-examples/blob/main/dynamic-endpoint/public/index.html
• Reddit Post: https://old.reddit.com/r/htmx/comments/1fyaw8r/htmx_and_alpinejs_dynamic_data/

Where my keybinding looked like this:

vim.keymap.set("n", "<leader>ff", function()
    builtin.find_files({ hidden = true, follow = true })
end, { desc = "Find all files" })

This was because I was ignoring the .env* paths in my gitignore, I could tell telescope to not follow the gitignore. But I would get a lot more junk when searching in Telescope. I just wanted to see certain patterns but still mostly respect the gitignore file. You can do the following:

#  create a ~/.ignore file
nvim ~/.ignore

# Mine looks like this
!.env.local
!*/.env.local
!.env.test
!*/.env.test

Now in Telescope the .env.local and .env.test files are searchable. But still ignored by git I don’t want to commit them as they contain secrets.

This of course assumes you are using ripgrep to do the search, in my config I have the following:

vimgrep_arguments = {
    "rg",
    "-L",
    "--color=never",
    "--no-heading",
    "--with-filename",
    "--line-number",
    "--column",
    "--smart-case",
    "--fixed-strings",
    },

GitHub Issue: https://github.com/LunarVim/LunarVim/discussions/3770#discussioncomment-11523524

In my post last month I thought I had fixed this issue but turns out I did not. The actual fix to stop getting this error:

5108: Error executing lua [string "v:lua"]:1: attempt to call global 'dropbar' (a nil value)
stack traceback:
        [string "v:lua"]:1: in main chunk

bat /home/haseeb/.local/share/nixCats-nvim/sessions/%2Fhome%2Fhaseeb%2Fprojects%2Fvoxicle.vim | rg dropbar
setlocal winbar=%{%v:lua.dropbar()%}
setlocal winbar=%{%v:lua.dropbar()%}
setlocal winbar=%{%v:lua.dropbar()%}

We want the winbar to nil. The fix that worked for me was this:

vim.o.sessionoptions = "blank,buffers,curdir,folds,help,tabpages,winsize,winpos,terminal,localoptions"

require("auto-session").setup({
	pre_save_cmds = {
		function()
			vim.cmd([[
                noautocmd windo set winbar=
                noautocmd windo setlocal winbar=
            ]])
		end,
	},
})

I recently moved to dropbar from barbecue as it has been archived and kept getting a weird error for buffer open when I would reopen Neovim. I use the auto-session plugin, which loads back the previous state of NixVim when I last exited in that folder, i.e. open buffers.

E5108: Error executing lua [string "v:lua"]:1: attempt to call global 'dropbar' (a nil value)
stack traceback:
        [string "v:lua"]:1: in main chunk

I noticed though I would only get this error the second time I would open Neovim, which then made think maybe it was an issue with auto-session. So I looked at the serialised file it generates so it knows what to load when we open Neovim again. Which I found at:

nvim /home/haseeb/.local/share/nvim/sessions/%2Fhome%2Fhaseeb%2Fprojects%2Fvoxicle.vim

One line in there looked like this:

setlocal winbar=%{%v:lua.dropbar()%}

So I removed this line by updating my config:

require("auto-session").setup({
	pre_save_cmds = {
		function()
			vim.opt.winbar = nil -- Clear winbar before saving session
		end,
	},
})

This stopped the error from occurring, as there is no winbar line now. I am not a 100% what was happening, but it seems it was running this line before dropbar had loaded. I am using nixCats. Anyway, that’s it! That should resolve the issues.

Today I learnt that we can cancel old pipelines in MRs. For example, you push a new commit, and you don’t care about the old pipeline running any more. You want to cancel them to save CI minutes etc.

Here is an example, you can see the 2nd pipeline is cancelled:

We need to add this to our .gitlab-ci.yml file. Where the default.interruptible marks every job as it can be cancelled ¹.

workflow:
  auto_cancel:
    on_new_commit: interruptible

default:
  interruptible: true

Then on new commits, we cancel all interruptible jobs. Just make sure to be careful where jobs actually deploy stuff. You don’t want to partially deploy things. But for jobs like running tests, linting etc this should be all good.

• Example gitlab pipeline

When shell completions don’t work with direnv, you may need to use nix develop to load the shell manually.

Background

I am using nix-direnv, with nix devshell to autoload into my development environments. I changed the directory and the devshell is automagically loaded without me doing anything, which is great. Provides me with a bunch of tools specific for that project. With all the benefits of Nix, mainly reproducibility.

❯ z banterbus/
direnv: loading ~/projects/banterbus/.envrc
direnv: using flake
direnv: nix-direnv: Using cached dev shell
direnv: export +AR +AS +CC +CONFIG_SHELL +CXX +GOTOOLDIR +HOST_PATH +IN_NIX_SHELL +LD +NIX_BINTOOLS +NIX_BINTOOLS_WRAPPER_TARGET_HOST_x86_64_unknown_linux_gnu +NIX_BUILD_CORES +NIX_CC +NIX_CC_WRAPPER_TARGET_HOST_x86_64_unknown_linux_gnu +NIX_CFLAGS_COMPILE +NIX_ENFORCE_NO_NATIVE +NIX_HARDENING_ENABLE +NIX_LDFLAGS +NIX_STORE +NM +OBJCOPY +OBJDUMP +PLAYWRIGHT_BROWSERS_PATH +PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD +RANLIB +READELF +SIZE +SOURCE_DATE_EPOCH +STRINGS +STRIP +__structuredAttrs +buildInputs +buildPhase +builder +cmakeFlags +configureFlags +depsBuildBuild +depsBuildBuildPropagated +depsBuildTarget +depsBuildTargetPropagated +depsHostHost +depsHostHostPropagated +depsTargetTarget +depsTargetTargetPropagated +doCheck +doInstallCheck +dontAddDisableDepTrack +hardeningDisable +mesonFlags +name +nativeBuildInputs +out +outputs +patches +phases +preferLocalBuild +propagatedBuildInputs +propagatedNativeBuildInputs +shell +shellHook +stdenv +strictDeps +system ~PATH ~XDG_DATA_DIRS

Where my mkShell function looks something like:

{
  pkgs.mkShell {
    hardeningDisable = ["all"];
    shellHook = ''
      export PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1
      export PLAYWRIGHT_BROWSERS_PATH="${pkgs.playwright-driver.browsers}"
      ${pre-commit-check.shellHook}
    '';
    buildInputs = pre-commit-check.enabledPackages;
    packages = with pkgs; [
      # TODO: workout how to use go env
      # goEnv
      gomod2nix
      go_1_22
      playwright-test

      goose
      air
      golangci-lint
      gotools
      gotestsum
      gocover-cobertura
      go-task
      go-mockery
      goreleaser
      golines

      tailwindcss
      templ
      sqlc
    ];
  }
}

You can see I have certain tools I want other developers to have access to like go-task, which they will automatically get when they load into the devshell.

Issue

This is all great; however, I have noticed that shell completions stopped working. Where I can normally do task <TAB>

❯ task tests:e2e
build:dev  (Build the app for development, generates all the files needed for the binary.)  generate:sqlc  (Generates the code to interact with SQL DB.)
coverage                            (Run the integration tests and gets the code coverage)  lint                                      (Runs the linter.)
dev                                       (Start the app in dev mode with live-reloading.)  release                              (Release the CLI tool.)
docker:build                                            (Builds a Docker image using Nix.)  tests                                  (Runs all the tests.)
docker:load                                (Loads the Docker image from tar (in results).)  tests:e2e                  (Runs e2e tests with playwright.)
docker:publish                                                (Publishes the Docker image)  tests:integration          (Runs all the integration tests.)
format                                                               (Runs the formatter.)  tests:unit                        (Runs all the unit tests.)

However, in the devshell this does not work until I load into the shell manually nix develop. Which does defeat the purpose of using nix-direnv as you need to manually run a command now.

I don’t know what causes this and will do a deeper dive at some point 🤔.

Recently, I was trying to pull docker images using podman, on an Ubuntu laptop and was getting an error which looked something like:

Error: error creating build container: short-name "node:18.17" did not resolve to an alias and no unqualified-search registries are defined in "/etc/containers/registries.conf"

This is because Podman doesn’t allow us to use short names, by default we need to specify the registry i.e. docker.io/node:18.17. But for existing docker-compose.yml files, this would be a pain to edit. Especially because most people use Docker not Podman. So, in the end, you can edit a config file in home directory to enable the old way of pulling images.

nvim ~/.config/containers/registries.conf

# Add the following
unqualified-search-registries=["docker.io"]

You could also do it machine wide by editing the file /etc/containers/registries.conf.

That’s it! Real short one today! This just stumped me today for like 20 minutes.

Recently, I was trying to set up the kubernetes-dashboard, to make it easier to monitor my k8s cluster. I however noticed I was getting the following error:

… > add-kubernetes-dashboard via 🐹 v1.22.2 via   via ❄  impure (nix-shell-env)
 k$ kubectl logs -n monitoring kubernetes-dashboard-kong-75bb76dd5f-b27ll

2024/05/05 20:55:21 [error] 1319#0: *274054 [lua] init.lua:371: execute(): DNS resolution failed: failed to receive reply from UDP server 10.43.0.10:53: timeout. Tried: nil, client: 127.0.0.1, server: kong, request: "GET / HTTP/2.0", host: "localhost:8443", request_id: "af30b3162db70e2de8f1073a40f7d865"

Weird the DNS resolution seems to be failing, so I decided to follow the Kubernetes guide to try to debug issues.

I found out I couldn’t even do a nslookup:

kubectl exec -i -t dnsutils -- nslookup kubernetes.default

;; connection timed out; no servers could be reached
command terminated with exit code 1

I had set up k3s on NixOS (as per the previous posts in this series). In the ended, the issue ending up being not open the correct ports, we are using the Flannel VXLAN ¹. UDP 8472 All nodes All nodes Required only for Flannel VXLAN The GitHub issue that helped me solve my problem.

We need to open up port 8472 in our firewall rules, on my NixOS server we can do something like so.

{
networking.firewall = {
  allowedUDPPorts = [
    # ...
    8472
  ];
}

So in my case, since my k8s cluster was multi-node, the nodes couldn’t communicate with each other properly. That’s about it, it took me a while to figure out, basically RTFM 😭😭😭.

Recently, I was trying to create a derivation which needed to pull git submodules as well. I was getting an error which look something like this:

data/meson.build:76:0: ERROR: Nonexistent build file 'data/submodules/meson.build'

It was coming from this derivation https://gitlab.com/hmajid2301/dotfiles/-/blob/c153de146a3bf9339cbef013ac65bc32e6305c8e/packages/gradience/default.nix for building the latest version of gradience.

It turns out when we do a fetchFromGitHub we need to explicitly tell it to also pull submodules, which makes sense. So I was able to fix this. I had to add the fetchSubmodules argument:

{
  pkgs,
  lib,
  fetchFromGitHub,
  ...
}:
pkgs.python3Packages.buildPythonApplication {
  pname = "gradience";
  version = "0.8.0-beta1";

  src = fetchFromGitHub {
    owner = "GradienceTeam";
    repo = "Gradience";
    rev = "90b774174da0e3c6b5314e38226bea653a5bf57a";
    sha256 = "sha256-C0GV6vOEZ0wTaKO7BgGuFvHsHeaVwH0W1U8yKUMrO9c=";
    fetchSubmodules = true;
  };
}

That’s it! A simple fix, but took me longer than it should’ve done to figure it out.

If you are like me, you may be configuring your Neovim configuration using NixVim. Which is a “framework” making it easier to configure Neovim using mostly Nix configuration. Allowing us to keep most of our config in one language, at least in my case.

I started using Neorg to manage my notes and to-do lists in my “second-brain”, but noticed that the metadata part was not being highlighted as I expected.

My config looked something like:

{
  programs.nixvim = {
    plugins = {
      neorg = {
        enable = true;
        lazyLoading = true;
        modules = {
          # ...
        };
      };

      treesitter = {
        grammarPackages = with config.programs.nixvim.plugins.treesitter.package.builtGrammars; [
          norg
        ];
      };
    };
  };
}

Turns out there is another tree sitter package for the n Neorg metadata, but this is not available with nvim-treesitter. Hence, also not available with Nixvim. I was able to resolve my issue my using the tree-sitter specific grammar like so:

{
  programs.nixvim = {
    plugins = {
      neorg = {
        enable = true;
        lazyLoading = true;
        modules = {
          # ...
        };
      };

      treesitter = {
        grammarPackages = with pkgs.tree-sitter-grammars; [
          tree-sitter-norg
          tree-sitter-norg-meta
        ];
      };
    };
  };
}

It looks like this. That’s it!

I have been using Zellij for a while now. I tried to set up layouts for one of my personal projects. So that we could have tests and linting running and any other tasks we may want whilst doing development ¹.

However, I had an issue working out how to call commands that required direnv and nix to set up development environments. In my case, my nix dev shell installed the go-task tool to run tasks such as task lint or task tests. However the shell that zellij was running the commands in for the layout did not load my Nix dev shell, so the go-task binary was not available.

So to fix this, we need to do something like this layout.kdl:

layout {
    pane size=1 borderless=true {
        plugin location="zellij:tab-bar"
    }
    pane {
        command "fish"
        args "-c" "direnv exec . task lint"
    }
    pane {
        command "fish"
        args "-c" "direnv exec . task tests"
    }
    pane {
        command "docker-compose"
        args "up" "--build -d"
    }
}

Where the key being we need to execute the direnv environment i.e. direnv exec ., then we can run commands like normal. Where direnv .envrc file contains use flake. That’s it! If you want Zellij to use direnv and load your nix flake dev shell, you have to add a bit of boilerplate, which is not ideal but

I added Sway Notification Center as my notification manager and added a small “widget” to my Waybar, so I can see how many notifications I have and silence notifications by clicking on it. However, I found when I opened the swaync sidebar, in my case by right-clicking on the icon. I found that I could not click on anything else on my Waybar like workspaces. Now I know I should be using my keyboard, but sometimes it’s just easier to use a mouse.

The fix I found was on Reddit. I use nix and configure Waybar using home-manager, the on-click actions now have a small sleep, which I don’t even notice, and this resolves the above issue.

{
"custom/notification" = {
  tooltip = false;
  format = "{} {icon}";
  "format-icons" = {
    notification = "󱅫";
    none = "";
    "dnd-notification" = " ";
    "dnd-none" = "󰂛";
    "inhibited-notification" = " ";
    "inhibited-none" = "";
    "dnd-inhibited-notification" = " ";
    "dnd-inhibited-none" = " ";
  };
  "return-type" = "json";
  "exec-if" = "which swaync-client";
  exec = "swaync-client -swb";
  "on-click" = "sleep 0.1 && swaync-client -t -sw";
  "on-click-right" = "sleep 0.1 && swaync-client -d -sw";
  escape = true;
};
}

That’s it!

Recently I stopped using tmux to try zellij, however I noticed when I removed tmux from my nix config. I was getting the following error, when rebuilding my home-manager config:

The user systemd session is degraded:
  UNIT         LOAD   ACTIVE SUB    DESCRIPTION
● tmux.service loaded failed failed tmux default session (detached)

Legend: LOAD   → Reflects whether the unit definition was properly loaded.
        ACTIVE → The high-level unit activation state, i.e. generalization of SUB.
        SUB    → The low-level unit activation state, values depend on unit type.

I was wondering where this error was coming from, turns it the symlink to tmux.service has not been deleted that nix would create.

ls -al ~/.config/systemd/user/tmux.service
Permissions Size User        Group       Date Modified Name
.rw-rw-r--   453 haseebmajid haseebmajid 22 Aug  2023   /home/haseebmajid/.config/systemd/user/tmux.service

If we look at other systemd services we can see them symlinked to a nix package in the nix store.

ls -al ~/.config/systemd/user
lrwxrwxrwx     - haseebmajid haseebmajid 20 Feb 17:04   spotifyd.service -> /nix/store/7v52bpcyi0zqv012rrmd7s2r742hb2cy-home-manager-files/.config/systemd/user/spotifyd.service
lrwxrwxrwx     - haseebmajid haseebmajid 20 Feb 17:04   swayidle.service -> /nix/store/7v52bpcyi0zqv012rrmd7s2r742hb2cy-home-manager-files/.config/systemd/user/swayidle.service

So we can simply remove the tmux file, rm ~/.config/systemd/user/tmux.service. I should look into why this symlink is not removed.

However, I kept getting the following errors:

nix-shell -p vimPluginsUpdater --run vim-plugins-updater
error:
       … while calling the 'derivationStrict' builtin

         at /builtin/derivation.nix:9:12: (source not available)

       … while evaluating derivation 'shell'
         whose name attribute is located at /nix/store/whhzjfgalghpm34irclh01c0afynmyll-nixpkgs/nixpkgs/pkgs/stdenv/generic/make-derivation.nix:300:7

       … while evaluating attribute 'buildInputs' of derivation 'shell'

         at /nix/store/whhzjfgalghpm34irclh01c0afynmyll-nixpkgs/nixpkgs/pkgs/stdenv/generic/make-derivation.nix:347:7:

          346|       depsHostHost                = lib.elemAt (lib.elemAt dependencies 1) 0;
          347|       buildInputs                 = lib.elemAt (lib.elemAt dependencies 1) 1;
             |       ^
          348|       depsTargetTarget            = lib.elemAt (lib.elemAt dependencies 2) 0;

       error: undefined variable 'vimPluginsUpdater'

       at «string»:1:107:

            1| {...}@args: with import <nixpkgs> args; (pkgs.runCommandCC or pkgs.runCommand) "shell" { buildInputs = [ (vimPluginsUpdater) ]; } ""

Or I was getting 429 network errors, too many requests to GitHub.

Solution

I found this great person who was able to solve my issue on discourse. Assuming you have a fork of nixpkgs and have to clone it locally, i.e. git clone https://github.com/hmajid2301/nixpkgs.git

1. Update nixpkgs/pkgs/applications/editors/vim/plugins/vim-plugin-names
2. Can do this using the ./update.py add "gbprod/yanky.nvim" (where gbprod/yanky.nvim is the path to git repository)
3. Export your GitHub Personal Token
4. You can create one by following this link
5. export GITHUB_API_TOKEN=your_token
6. I will do a post in the future about how you can do this in fish shell and keep secrets out of your shell history.
7. nix-shell -I nixpkgs=channel:nixpkgs-unstable -p vimPluginsUpdater --run vim-plugins-updater
8. Make sure to run this in the repository root
9. Push your changes and create a PR with nixpkgs repo
10. You will likely need to rebase your commits to follow the contribution guidelines, check those when creating the PR

That’s it! I am not sure if this will be useful for others (I will most likely reference this myself in the future), but that’s the easiest way I found to add new vim plugins to nixpkgs.

Recently, I had an issue with my Ubuntu laptop running Hyprland, when running Firefox it had this ugly border around it. Which looked something like this:

You can see a weird border around the browser, and if you try to view some of the context menus, they appear small and difficult to read.

Fix

As per this Reddit thread ¹. There are two fixes, one upgrade to version 0.33 which I could not easily do with Ubuntu. So I used the other fix,

• Right click on title bar
• Select Customize Toolbar..
• Then check Title Bar
  • You may have to full-screen the app to make it easier to select the title bar

Hyprland Version

hyprctl version
Hyprland, built from branch HEAD at commit 5b8cfdf2efc44106b61e60c642fd964823fd89f3  (props: bump ver to v0.31.0).
Tag: v0.31.0

flags: (if any)
debug

Recently, I had to move one of my to-do files from its own repository, todo, to my second-brain repository. So I could better keep track of my ideas, and tasks to be done. However, I realised just copying the file over I would lose all the git history the file had.

Strictly speaking, I don’t need the git history of that file, as the most important thing is the current tasks to be completed. I could, also, go look in the older repository if I really needed to see the file history. However, I decided it would be a nice exercise to learn how to copy files over and keep their history.

How?

So how can we do just that? In this example, we will copy a file from todo to second-brain.

So clone both repos i.e.

git clone git@git.com/todo.git
git clone git@git.com/second-brain.git

Then we will use the git-filter-repo tool which we can use to rewrite our git history.

extract the history of a single directory, src/. This means that only paths under src/ remain in the repo, and any commits that only touched paths outside this directory will be removed. - git filter repo README

If you are using the Nix package manager, you can do something like:

nix-shell -p git-filter-repo

Then let’s create a new branch on our source repo (todo), then find all commits that are related to the todo.norg file.

cd todo
git checkout -b filter
git filter-repo --path todo.norg --refs refs/heads/filter --force

git filter-repo --path todo.norg --refs refs/heads/filter --force

# Parsed 69 commits
# New history written in 0.02 seconds...
# HEAD is now at 21e6ae4 chore: Update ToDos
# Completely finished after 0.03 seconds.

git log

# commit 21e6ae42c550b2fdd76a1a7c34f8574eb529175d
# Author: Haseeb Majid <hello@haseebmajid.dev>
# Date:   Sun Feb 4 22:46:36 2024 +0000
# 
#     chore: Update ToDos
# 
# commit c0e125be9bddf6666fc321ad3ea1b51e437e4bfc
# Author: Haseeb Majid <hello@haseebmajid.dev>
# Date:   Fri Feb 2 22:53:15 2024 +0000
# 
#     chore: Update ToDos

Then go to our target repo second-brain, we will add the todo repo as a local source to git. Then merge in the changes on the filter branch. We also need to pass the –allow-unrelated-histories flag tells Git to allow the merge, even if the histories of the branches don’t share a common commit.

cd second-brain
git checkout -b filter
git remote add todo-source ../todo
git fetch todo-source
git branch todo-source remotes/todo-source/filter
git merge todo-source --allow-unrelated-histories

# Merge changes into main branch
git checkout main
git merge filter

# Delete reference to remote and branch
git remote rm todo-source
git branch -d filter

Now we will have the todo.norg in our second-brain repo with the git history of the file. If we now look at the git log, we would see all those commits updating the to-do file.

Appendix

• Useful Links:
  • https://www.johno.com/move-directory-between-repos-with-git-history
  • https://blog.billyc.io/how-to-copy-one-or-more-files-from-one-git-repo-to-another-and-keep-the-git-history/

invalid argument for type conversion: missing custom lists settings

it turned out to somehow a mismatch in versions where everything was running 2023.6 but my mullvad cli was using 2023.5. So I ended up fixing this by changing my config to:

{
  services.mullvad-vpn = {
    enable = true;
    package = pkgs.mullvad-vpn;
  };
}

Where setting the package to mullvad-vpn, which then installed the correct version of the mullvad cli tool and I was then able to change locations using the gui as per usual. No idea why there was a version mismatch.

Recently, I tried to boot with my Ventoy USB on my new AMD motherboard Framework. However, I was getting a which looked like Verification failed:(0x1A) Security Violation. It turns out this was because secure boot was turned on. So we needed to turn it off initially to boot off the Ventoy. You can follow the instructions here.

We can then turn on secure boot after, I will do a future post how we can do this with NixOS. When I get it working.

That’s it! Super short post this time.

Appendix

• SO Post

Recently, I started using a ZSA Voyager split keyboard, moving to this keyboard has some advantages but the last thing I felt I was missing from my old keyboard (which has a volume knob) was being able to control the volume. So I set up the key maps in their software (Oryx) however, I noticed that binding were not working.

Where I had something like this:

bind=,XF86AudioRaiseVolume,exec, volume --inc
bind=,XF86AudioLowerVolume,exec, volume --dec
bind=,XF86AudioMute,exec, volume --toggle
bind=,XF86AudioMicMute,exec, volume --toggle-mic

This worked with my old keyboard just fine, however with my new keyboard, I realised I was pressing a key to change the layers to access the media keys. I press what would normally be a Shift key to change to layer 3, where my media keys are. As I don’t have enough keys to do that all in one layer, like I could do on my other keyboard.

I think since I was holding down a key at the same time as pressing the media keys, I ended up using bindi to fix my issue.

i -> ignore mods, will ignore modifiers - ¹.

So now my config looks like:

bindi=,XF86AudioRaiseVolume,exec, volume --inc
bindi=,XF86AudioLowerVolume,exec, volume --dec
bindi=,XF86AudioMute,exec, volume --toggle
bindi=,XF86AudioMicMute,exec, volume --toggle-mic

That’s it!

Sometimes we want to use env variables in our docker-compose files like so:

services:
    client:
        image: ${CI_DEPENDENCY_PROXY_GROUP_IMAGE_PREFIX}/nginx
        ports:
            - 8000:80

Here we are going to use the GitLab CI dependency proxy to pull our Nginx image, so we can speed up our pipelines but also avoid being rate limited by docker hub. However, when running this locally, we will need to make sure the CI_DEPENDENCY_PROXY_GROUP_IMAGE_PREFIX env variable is set. Which just adds a bit more work, instead we can leverage some of the special syntax docker-compose provides ¹, which I think it inherits from bash.

Where we can use interpolation to set default values if the env variable is not set:

services:
    client:
        image: ${CI_DEPENDENCY_PROXY_GROUP_IMAGE_PREFIX:-docker.io}/nginx
        ports:
            - 8000:80

In this case, :- will use docker.io if the env variable is not set or is empty. There are several other variations you can find in the footnote below.

I had an issue that whilst debugging my Golang tests in Neovim, the buffer I was focusing on would always change to the breakpoint. Even when the buffer was open and visible already, say I had the code with the breakpoint on the left buffer and the tests I was starting the debugger from on the right buffer.

You can see the example belows:

When I start the DAP debugger on the right hand side, by using it to debug the nearest tests, the buffer will change to the breakpointed line. However I want it to use the open buffer on the left because it is already open.

To do this we need to update our Neovim options: vim.o.switchbuf = "useopen,uselast". This will now behave as we want. Note uselast, here is the default value for this option. You can read more about by typing :help 'switchbuf.

That’s it! We can now debugger our code and keep the layout we have.

Recently, I was trying to open an NTFS drive on my NixOS machine; however, the drive was corrupted. So I did the following to fix the drive.

nix-shell -p ntfs3g
ntfsfix /dev/sda1

Where /dev/sda1 is the broken drive. This was enough for me to be able to mount the drive and access the files on it. I didn’t need to fix it on a Window machine.

Recently on my laptop I had to add some extra config settings to my Hyprland config. Rather than polluting my hyprland.nix file with if, else depending on the host. I wanted to add the extra config in the home.nix of the host. So it’s contained within that host. Where my home.nix is the entry point for my home-manager config for that host.

My hyprland.nix looks like this:

{
wayland.windowManager.hyprland = {
  enable = true;
  extraConfig = ''
     input {
        kb_layout = gb
        touchpad {
            disable_while_typing=false
        }
     }
  '';
};
}

I want to add some extra config to the extraConfig option/attribute of this attribute set. Then in my home.nix I added the following:

{
wayland.windowManager.hyprland.extraConfig = lib.mkAfter ''
  exec-once = /usr/libexec/geoclue-2.0/demos/agent
  exec-once = warp-taskbar

  bind=,XF86Launch5,exec,/usr/local/bin/swaylock -S
  bind=,XF86Launch4,exec,/usr/local/bin/swaylock -S
  bind=SUPER,backspace,exec,/usr/local/bin/swaylock -S
'';
}

In this case we want to start some processes on this host, like showing warp in my taskbar. The key line here being lib.mkAfter. Which will append the content after the existing config options we defined in hyprland.nix

As per some of my recent articles, you will be aware I am using Hyprland (tiling manager) on Ubuntu and managing the config using nix (home-manager). I was having issues where for some reason it wouldn’t set the PATH variable correctly.

On my NixOS machine, the following would be fine:

bind=,XF86AudioRaiseVolume,exec, volume --inc
bind=,XF86AudioLowerVolume,exec, volume --dec

However, on Ubuntu I needed to provide the full path:

bind=,XF86AudioRaiseVolume,exec, ${volume}/bin/volume --inc

So in the Hyprland config it would look like:

bind=,XF86AudioRaiseVolume,exec, /nix/store/q1wm84g65smaq4agq7zp40q57x3534ni-volume/bin/volume --inc

It was because the session was being started by GDM (gnome), and the PATH variable was not being set properly. In Wayland GDM picks up environment variables from ~/.config/environment.d/envvars.conf ¹ so we can add it here.

❯ bat ~/.config/environment.d/envvars.conf --plain
PATH="$PATH:/home/haseebmajid/.nix-profile/bin"

Replace haseebmajid with your username (or I guess $HOME should work). This means when we run scripts now, it will source them from this nix-profile folder. Where the binaries get symlinked, including the volume script above. So now you don’t need to specify the full path (though maybe you should 🤷)

Appendix

• Hyprland Config
• Reddit Thread
• Discourse Thread

I have been tmux with the foot terminal and when trying to share my screen I couldn’t work out which terminal to share based on the name. Since they are running tmux, all the terminals were titled terminal - t.

To fix this and name it after what is running (and where) in tmux, i.e. foot blog/nvim. We can do this by adding the following to our tmux config:

set-option -g set-titles on
set-option -g set-titles-string "#S / #W"

Here, the #S is the session name and #W the window name; hence, we get something like the above. Though you could use lots of other different formats ¹.

This is enough for me to figure out which window to share.

$ cp /usr/share/applications/org.gnome.Shell.desktop ~/.local/share/applications/org.gnome.Shell-Wayland.desktop

$ less ~/.local/share/applications/org.gnome.Shell-Wayland.desktop
...
Exec=/usr/bin/gnome-shell --wayland
...

$ cp /usr/share/gnome-session/sessions/gnome.session ~/.config/gnome-session/sessions/gnome-wayland.session

$ less ~/.config/gnome-session/sessions/gnome-wayland.session
...
RequiredComponents=org.gnome.Shell-Wayland;gnome-settings-daemon;
...

#!/bin/sh

[ -r /bin/gnome-session ] && {
	export \
		EGL_PLATFORM=wayland \
		GDK_BACKEND=wayland \
		XDG_SESSION_DESKTOP=gnome
	/bin/gnome-session --session=gnome-wayland
}

$ gnome

Recently, I moved to Hyprland on Ubuntu. I wanted to start gnome in another TTY (teletype). It was more effort to find than I expected:

# Go to teletype
CTRL+ALT+1

dbus-run-session -- gnome-shell --display-server --wayland

That’s it, short and sweet, this post! You now started gnome in a Wayland session.

I was having issues with my debugger today (well technically yesterday because I am publishing this a day later to spread out my blog posts but same difference) and it took me a few hours to realise what was going on. In my case, I was trying to debug a test written in Golang using nvim-dap-go on Neovim.

The reason the test was failing because the test file had build tags so delve (the debugger) couldn’t compile the file, i.e. using the unit build tag in the example below.

//go:build unit
// +build unit

package options_test

When we set up the dag-go plugin, we need to pass an extra field called build_flags:

lua require('dap-go').setup {
  delve = {
    build_flags = "-tags=unit,integration,e2e",
  },
}

Replace the above with the tags you use! If you are using NixVim you should be able to do something like this ¹:

{
  programs.nixvim = {
    plugins = {
      dap.extensions.dap-go = {
        enable = true;
        delve = {
          path = "${pkgs.delve}/bin/dlv";
          buildFlags = "-tags=unit,integration,e2e";
        };
      };
    };
  };
}

Then I have a keybinding to debug the nearest test like so, again using NixVim.

  {
    action = "<cmd> lua require('dap-go').debug_test()<CR>";
    key = "<leader>td";
    options = {
      desc = "Debug Nearest (Go)";
    };
    mode = [
      "n"
    ];
  }

in lua:

vim.keymap.set("n", "<leader>td", "<cmd>lua require('dap-go').debug_test()<CR>")

Recently, I wanted to install the nightly version of Neovim (version 0.10) because it supports inlay hints. However, on nixpkgs the latest version of Neovim as of writing is 0.9.4. So how can we get the nightly release? Using nix/home-manager.

Simple, we can use an overlay that will add the Neovim nightly package with the nightly. Assuming we are using nix flakes.

Add the nightly Neovim as an input:

 inputs.neovim-nightly-overlay.url = "github:nix-community/neovim-nightly-overlay";

Then add the actual overlay, I do this in my home manager config ¹:

  nixpkgs = {
    overlays =  [
      inputs.neovim-nightly-overlay.overlay
    ];
  };

Then in another module we can do something like:

home.packages = with pkgs [neovim-nightly]

or if you are using NixVim ² like me:

{
  programs.nixvim = {
    enable = true;
    extraPlugins = with pkgs.vimPlugins; [ plenary-nvim ];
    package = pkgs.neovim-nightly;
  };
}

If you use network manager on Linux and have both Wi-Fi and wired connection. You may want to prefer using a wired connection over Wi-Fi, due to stability. To do open the nm-connection-editor, if you are using Nix, you can download it from nixpkgs like usual.

Higher number means higher priority.

So for our wired connection I set the priority to 100 (it was previously -1) and then for our Wi-Fi connection I set the priority to 1.

That’s it! We should now see network manager prefer to use Ethernet over Wi-Fi.

Recently, I wanted to run a bash script where it needed to receive a JSON file as input. However, the JSON I had was taken from somewhere on the internet. In this case, it was taking a JSON blob and converting it to a nix attribute set. However, I didn’t want to save the contents to a file beforehand.

This is the command here as an example, note the /dev/stdin which acts as file like object for us.

wl-paste | json2nix /dev/stdin

So wl-paste returns the contents of our clipboard (our JSON). Then we pipe that over to our json2nix script which expects a file. We can pretend to have a file by using the /dev/stdin ¹

Recently, I was trying to run a libvirt virtual machine, created using virt-manager. The VM was running NixOS and somehow no matter what I did it would not connect to the internet.

I was running Mullvad as my VPN, and it turned out this was the app causing issues, I had to turn on Local network share ¹. Like in the photo below:

That’s It! Simple fix that took me longer to work than I’d like to admit.

Today (for once), I ran golangci-lint run and it failed on CI with the following error:

internal/options/parser_test.go:13: Function 'TestParse' is too long (69 > 60) (funlen)
func TestParse(t *testing.T) {
task: Failed to run task "lint": exit status 1

Where my .golangci.yml file looked like this:

run:
  timeout: 5m
  skip-dirs:
    - direnv

linters:
  enable:
    - bodyclose
    - dogsled
    - dupl
    - errcheck
    - exportloopref
    - funlen
    - gochecknoinits
    - goconst
    - gocritic
    - gocyclo
    - gofmt
    - goimports
    - gomnd
    - goprintffuncname
    - gosec
    - gosimple
    - govet
    - ineffassign
    - lll
    - misspell
    - nakedret
    - noctx
    - nolintlint
    - revive
    - staticcheck
    - stylecheck
    - typecheck
    - unconvert
    - unparam
    - unused
    - whitespace

So basically we want the ability to turn off funlen for our tests (and other linters). Because we don’t mind if some of our test functions get a big long with different sub-tests.

So to disable some linters for our test files we can do something like this:

linters:
  enable:
    - bodyclose
    # ...

issues:
  exclude-rules:
    - path: _test.go
      linters:
        - funlen

That’s it! For once a TIL, that was written on the same day as I learned it (or re-learnt it)

I wanted to change the background colour of the line when it was stopped during debugging using DAP with neovim. To make it easier to see where we are currently breakpointed.

We already have a highlight group called DapStopped. Which in my case is defined as: DapStopped = { bg = C.grey }.

Then we need to assign the custom highlight group, the key bit being linehl for our use case.

require("dap")

local sign = vim.fn.sign_define

sign("DapBreakpoint", { text = "●", texthl = "DapBreakpoint", linehl = "", numhl = ""})
sign("DapBreakpointCondition", { text = "●", texthl = "DapBreakpointCondition", linehl = "", numhl = ""})
sign("DapLogPoint", { text = "◆", texthl = "DapLogPoint", linehl = "", numhl = ""})
sign('DapStopped', { text='', texthl='DapStopped', linehl='DapStopped', numhl= 'DapStopped' })

Then we start the debugger and it stops at a breakpoint line it looks something like in the photo below.

Recently I wanted to use GNU parallel, a nifty little tool we can use to run tasks in parallel, shock horror I know. I already have the moreutils package installed using nix (home-manager).

So I added this:

home.packages = with pkgs; [
  parallel
  moreutils
]

Then I ran the home manager switch like usual, I got the following error.

error: builder for '/nix/store/vh6i81xhf6pvybdpall8z8l8y0i6mr8p-home-manager-path.drv' failed with exit code 25;
       last 1 log lines:
       > error: collision between `/nix/store/slkylri9sbn4w7paaixzc5wj6cwfk83m-moreutils-0.67/bin/parallel' and `/nix/store/fjpnw99zvx2f910s016jrmybc5jxirpn-parallel-20230822/bin/parallel'
       For full logs, run 'nix log /nix/store/vh6i81xhf6pvybdpall8z8l8y0i6mr8p-home-manager-path.drv'.
error: 1 dependencies of derivation '/nix/store/m4p56x5lgjfpvdy0xjrbk981sfchpl8c-home-manager-generation.drv' failed to build

Because moreutils also have a binary called parallel. How do we tell Nix to overwrite the moreutils binary and use the one in the parallel packages? Simple by using lib.hiPrio like so:

home.packages = with pkgs; [
  (lib.hiPrio parallel)
  moreutils
]

This tells Nix to increase the priority of this package you can see the code here.

That’s it we can still have access to the other binaries moreutils provides like sponge and parallel from parallel package.

I use Fish Shell with fzf and fish fzf plugin. I installed fzf using nix and home-manager:

{
  programs.fzf = {
    enable = true;
    enableFishIntegration = false;
  };
}

It adds fzf-history-widget script, which is bound to ctrl+r, shell reverse history search. Which we can see when we run this: bind | grep -e fzf-history-widget -e fzf-file-widget

Whereas I wanted to use the _fzf_search_history which is made available by that fish fzf plugin. Now PatrickF1 recommends running the uninstaller script from fzf, but I found another way we can overwrite the binding by adding the following lines to our fish.config:

bind \cr _fzf_search_history
bind -M insert \cr _fzf_search_history

So now when I press ctrl+r in my shell I used to get this (before):

but now I get this (after):

Appendix

• [My nix fish shell config](https://gitlab.com/hmaji](https://gitlab.com/hmajid2301/dotfiles/-/blob/a34417ed452fe068ec4a91fba7ef0a14fca31b76/home-manager/shells/fish.nix#L48-49)

Recently I was adding sway to my nix config (setup via home-manager). I already had Hyprland config, I wanted both sway and Hyprland to use my waybar config with some slight differences. So basically I want to check if the current host machine is using Sway or Hyprland (I am assuming we will only use one).

The main reason for using Sway is my work laptop uses Ubuntu 22.04, It’s not easy to run Hyprland on a “stable” distro like Ubuntu better suited to Arch or NixOS.

If sway is enabled include some specific config in my Waybar config else include the Hyprland config. So first of all I have a file called sway.nix which looks something like this:

wayland.windowManager.sway = {
  enable = true;
  package = pkgs.swayfx;
  # ...
}

I then have a host file for my work laptop called curve/home.nix which imports this sway expression:

imports = [ ../../home-manager/desktops/sway.nix ];

Finally, the main part of this article, let’s look at my waybar.nix which is imported by sway.nix (and hyprland.nix) expression, which Check if sway is enabled (which happens if the sway.nix is imported).

{ config, ... }: {
  programs.waybar = {
    enable = true;
    settings = [
      {
        modules-left = [
          "custom/launcher"
          (
            if config.wayland.windowManager.sway.enable == true
            then "sway/workspaces"
            else "hyprland/workspaces"
          )
          "custom/currentplayer"
          "custom/player"
          "custom/audio_idle_inhibitor"
        ];
    }
  }
}

In the example above config.wayland.windowManager.sway.enable == true, we just do a simple if statement check to Decide which Waybar module to show (sway or Hyprland).

That’s it! How we can check if other expressions can use attributes set in other expression files. If you have suggestions to make this cleaner I’m all ears! This is not cleanest way to do something like this. But since these are just my own dotfiles I know I will only use one of Sway or Hyprland on my devices.

Recently I needed to install devenv using flakes in home-manager. One of the things I needed to pass to was the type of system to install on i.e. "x86_64-linux".

So as I temp hack I had something like: inputs.devenv.packages."x86_64-linux".devenv.

However I was able to access the system using the pkgs attribute like so:

{
  inputs,
  pkgs,
  ...
}: {
  home.packages = [
    inputs.devenv.packages."${pkgs.system}".devenv
  ];
}

Where my flake.nix looks something like:

  outputs = {
    self,
    nixpkgs,
    home-manager,
    ...
  } @ inputs: let
  in {
    inherit lib;
    homeConfigurations = {
      # Desktops
      mesmer = lib.homeManagerConfiguration {
        modules = [./hosts/mesmer/home.nix];
        pkgs = nixpkgs.legacyPackages.x86_64-linux;
        extraSpecialArgs = {inherit inputs outputs;};
      };
    };
  };

You can see here the system is defined with the pkgs, as I guess we need to know for which architecture we should build for i.e. amd, arch etc. But yeh thats it! For use within home-manager it seems we can use pkgs.system, to access the architecture, so we don’t need to hardcode it.

When I moved to NixOS I noticed that tmux-resurrect stop restoring some applications such as man and nvim. Like it used to on my Arch machine. I recently found a solution to my problem (thanks to a lovely chap on the nixos discourse).

By adding the following lines to our tmux config:

resurrect_dir="$HOME/.tmux/resurrect"
set -g @resurrect-dir $resurrect_dir
set -g @resurrect-hook-post-save-all 'target=$(readlink -f $resurrect_dir/last); sed "s| --cmd .*-vim-pack-dir||g; s|/etc/profiles/per-user/$USER/bin/||g; s|/home/$USER/.nix-profile/bin/||g" $target | sponge $target'

What this does is it edits the tmux-resurrect file from this:

# bat ~/.tmux/resurrect/last --plain

pane    dotfiles    0   1   :*  0   nvim ~/dotfiles :/home/haseeb/dotfiles  1   nvim    :/home/haseeb/.nix-profile/bin/nvim --cmd lua vim.g.loaded_node_provider=0;vim.g.loaded_perl_provider=0;vim.g.loaded_python_provider=0;vim.g.python3_host_prog='/nix/store/4n4d0f1xd14gl4pfymdcqb9pmagcyyfj-neovim-0.9.1/bin/nvim-python3';vim.g.ruby_host_prog='/nix/store/4n4d0f1xd14gl4pfymdcqb9pmagcyyfj-neovim-0.9.1/bin/nvim-ruby'

to this:

# bat ~/.tmux/resurrect/last --plain

pane    dotfiles    0   1   :*  0   nvim ~/dotfiles :/home/haseeb/dotfiles  1   nvim    :nvim

NOTE: we are just left with :nvim as the binary to run (from tmux-resurrect perspective). Then to restore my neovim session we could use a plugin which creates a Session.vim file in the local directory of that project, that tmux-resurrect can use to restore the actual neovim session rather than just running neovim.

However I don’t want to manage session files like that. So I use auto-session which checks if a session exists in a specific directory and if it does it automatically restores the last saved session when neovim opens.

Appendix

• Discourse Thread
• My tmux config

Recently I was trying to configure my own neovim config. I wanted the current line I was onto be coloured slighty differently and also the line number to be white. So it’s easier to see the line number and which line I was on.

It was not immediately obvious how to do this and took me long to work out than I’d like to admit 😅.

First we need to create to two highlight groups:

-- Highlights the line
CursorLine = { fg = 'NONE', bg = c.black2 },

-- Highlights the line number
CursorLineNr = { fg = c.white },

Then in our vim options set the following vim.o.cursorline = true (this is what I was missing, in my case).

Anyways that’s it! A really quick post, on something stupid I got stuck on!

I have been recently using NixOS/home-manager and I have been writing a lot of nixlang. To have my system state declaratively set up. I have been doing most of this editing in neovim. It took me a bit of time to work out how to get it set up so there is some basic LSP support and auto-formatting. I created a file called nix.lua and it looks like this:

return {
  -- Correctly setup lspconfig for Nix 🚀
  {
    "neovim/nvim-lspconfig",
    opts = {
      servers = {
          -- Ensure mason installs the server
          rnix = {},
      },
      settings = {
          rnix = {},
      },
    },
  },
  {
    "jose-elias-alvarez/null-ls.nvim",
    opts = function(_, opts)
      local nls = require("null-ls")
      if type(opts.sources) == "table" then
          vim.list_extend(opts.sources, {
              nls.builtins.code_actions.statix,
              nls.builtins.formatting.alejandra,
              nls.builtins.diagnostics.deadnix,
          })
      end
    end,
  },
}

This config will make sure we have some basic LSP i.e. missing pkgs definition in our expressions. It will auto-format our code as well. Which is very helpful.

P.S. I still don’t have any form of auto-complete setup. If you have any ideas I’m all ears.

Appendix

• My nix.lua

Recently I have been trying to configure my newish Keychron keyboard using VIA. To add a key to be able to take print screens, making use of the layers provided, as my Keychron Q1 PRO doesn’t have a dedicated print key 😓.

However, when I opened the app or the website, I would get the following error.

If we go to chrome://device-log you should be able to see something like this: Failed to open '/dev/hidraw9': FILE_ERROR_ACCESS_DENIED.

hidraw enables raw access to USB and Bluetooth Human Interface (hidraw) devices.

This means Chrome cannot access our keyboard, so to fix this we can do something like:

sudo chmod a+rw /dev/hidraw9

Then open the app/go to the website and you should see the errors are gone.

To set the original permissions you can do:

sudo chmod 600 /dev/hidraw8

Appendix

• Arch Forum Post which helped me

I have been using Kanshi to setup my monitor setups automagically depending on which monitors are plugged i.e. if my laptop is docked or not. If it is docked I want my laptop display to be off, when not docked I want it to be on. So my kanshi config file ~/.config/kanshi/config to look something like:

I use the name of my monitors as the ports they are plugged into my vary.

profile home_office {
  output "GIGA-BYTE TECHNOLOGY CO., LTD. Gigabyte M32U 21351B000087" mode 3840x2160@60Hz position 3840,0
  output "Dell Inc. DELL G3223Q 82X70P3" mode 3840x2160@60Hz position 0,0
  output "eDP-1" disable
}

profile undocked {
  output "eDP-1" enable scale 1.100000
}

To set this up using NixOS/home-manager we can do something like this say a file called kanshi.nix:

{
  services.kanshi = {
    enable = true;
    systemdTarget = "hyprland-session.target";

    profiles = {
      undocked = {
        outputs = [
          {
            criteria = "eDP-1";
            scale = 1.1;
            status = "enable";
          }
        ];
      };

      home_office = {
        outputs = [
          {
            criteria = "GIGA-BYTE TECHNOLOGY CO., LTD. Gigabyte M32U 21351B000087";
            position = "3840,0";
            mode = "3840x2160@60Hz";
          }
          {
            criteria = "Dell Inc. DELL G3223Q 82X70P3";
            position = "0,0";
            mode = "3840x2160@60Hz";
          }
          {
            criteria = "eDP-1";
            status = "disable";
          }
        ];
      };
    };
  };
}

I was unable to get it work until I added this line systemdTarget = "hyprland-session.target";, this is the Systemd target to bind to. Since we are using hyprland we need to attach to hyprland the default is sway.

NUR is the Nix user repository like the Arch user repository (AUR). It exists so that:

The NUR was created to share new packages from the community in a faster and more decentralized way. - https://github.com/nix-community/NUR

If we want to install packages from NUR in our home manager config which is set up using Nix Flakes. Assuming you build your home manager like

To do so first add it as an input in our flake.nix file:

  inputs = {
    nixpkgs.url = "github:nixos/nixpkgs/nixos-unstable";
    nur.url = "github:nix-community/NUR";
  };

Then we can use with home-manager by importing inputs.nur.hmModules.nur, for example in your home.nix file:

  imports = [
    inputs.nur.hmModules.nur
  ]

Then we should be able to import packages like so:

  home.packages = with pkgs; [
    nur.repos.peel.rofi-wifi-menu
    nur.repos.peel.rofi-emoji
  ];

Note these packages won’t work because they are out of date, and they caused me a lot of grief 😅, thinking it wasn’t working.

But that’s it! You should be able to use NUR packages with home-manager setup via a flake.

I have recently moved to NixOS, one of the great features of NixOS is that you can set up your entire machine from a single git repo. We can do this declaratively, what we mean by this is we tell nix what we want the state to be and nixos will work out how to get there.

For example, we can install Mullvad set various options already. This means on a new machine we don’t manually have to setup these mullvad settings.

We can set it up by doing something like mullvad.nix:

{pkgs, config, ...}: {
  environment.systemPackages = [ pkgs.mullvad-vpn pkgs.mullvad ];
  services.mullvad-vpn = {
    enable = true;
  };

  systemd.services."mullvad-daemon".postStart = let
    mullvad = config.services.mullvad-vpn.package;
  in ''
    while ! ${mullvad}/bin/mullvad status >/dev/null; do sleep 1; done
    ${mullvad}/bin/mullvad auto-connect set on
    ${mullvad}/bin/mullvad tunnel ipv6 set on
    ${mullvad}/bin/mullvad set default \
        --block-ads --block-trackers --block-malware
  '';
}

The main part here is we set up mullvad and then in the service we set various options. That can be set via the GUI app.

${mullvad}/bin/mullvad set default \
        --block-ads --block-trackers --block-malware

One improvement we could make is to use a secret manager like nix-sops and even login automatically using our account id [1].

Appendix

• Dotfiles

FZF is an amazing fuzzy finder tool, that is super flexible and you can create some cool cli one-liners with it. For example, I use it with the fish shell plugin to search through my command history (CTRL + R).

In this post, I will show you how vim style navigation to the FZF pop-up preview. All we need to do is add the following to the relevant file i.e. fish.config, .bashrc or .zshrc.

export FZF_DEFAULT_OPTS="--bind 'j:down,k:up,ctrl-j:preview-down,ctrl-k:preview-up'"

This will mean we can move and up down using the normal Vim binding j and k

In this post I will show you how you can change your fish shell greeter from the default

Welcome to fish, the friendly interactive shell
Type `help` for instructions on how to use fish

To something custom, this will run every time you open a new shell!

To do this, go to your fish_greeter.fish file which can usually be found at ~/.config/fish/functions/fish_greeting.fish. Where mine looks like:

function fish_greeting
    fortune | lolcat -f | chara say -c kitten
end

fortune generates a random quote, then lolcat colours that quote as a rainbow and finally chara prints out the kitten.

Nix (Optional)

If you are using fish to configure your fish shell you can do something like (I do this in a file called fish.nix):

{ lib, pkgs, ... }:

{
  programs.fish = {
    enable = true;
    functions = {
      fish_greeting = {
        description = "Greeting to show when starting a fish shell";
        body = "fortune | lolcat -f | chara say -c kitten";
      };
    };
  }
}

If you open your settings.json, you can add a section like this:

{
  "eslint.validate": [
    "javascript",
    "svelte",
    "javascriptreact",
    "typescript",
    "typescriptreact"
  ],
  "[typescript]": {
    "editor.defaultFormatter": "esbenp.prettier-vscode",
    "editor.tabSize": 2,
    "editor.codeActionsOnSave": {
      "source.organizeImports": true,
      "source.fixAll": true
    }
  },
}

Now this format our typescript files automatically when we save it, including organising our imports. Now this is all well and good, except when we add the following ESLint plugin ESLint-plugin-import.

This plugin lets us set our own rules for how to sort plugins and how to organise them into groups. So our ESLint config may look something like this:

module.exports = {
  // ...
  extends: [
    // ...
    "plugin:import/errors",
    "plugin:import/warnings",
    "plugin:import/typescript"
  ],
  rules: {
    // ...
    "import/order": [
      "warn",
      {
        groups: ["builtin", "external", ["sibling", "parent"], "index"],
        pathGroups: [
            {
                pattern: "$app/**",
                group: "external"
            },
            {
                pattern: "~/**",
                group: "sibling"
            }
        ],
        alphabetize: {
            order: "asc",
            caseInsensitive: true
        },
        "newlines-between": "always"
      }
    ]
  }
};

Now this will conflict with the way VS Code will format our imports, it will format them incorrectly according to the settings we have just set up in our ESLint config. To stop this from happening it’s very simple we need to remove the "source.organizeImports": true, line from our config. Now VS Code will use just ESLint to format our imports 🎉.

There is one downside to this approach I have noticed, which is unused imports no longer get removed automatically, which used to happen. But this will get fixed by pre-commit hooks.

Appendix

• Example Repo

At the moment all the slides are RevealJS “sites”, which means it’s a presentation built in HTML. Now I would like to have all my talks deployed to single GitLab pages site. For example something like:

• https://hmajid2301.io/talks/an-intro-to-pocketbase/
• https://hmajid2301.io/talks/docker-as-a-dev-tool/

So how can we do this ? Simple! In our job to deploy the site we just need to add public/docker-as-a-dev-tool and public/an-intro-to-pocketbase respectively. Which ever path we provide is that path we will be able to access those slides on.

Here is an example .gitlab-ci.yml:

# .gitlab-ci.yml
stages:
  - pages

pages:
  stage: pages
  only:
    - main
  scripts:
    - mkdir -p public/docker-as-a-dev-tool
    - mv docker-as-a-dev-tool public/docker-as-a-dev-tool
  artifacts:
    paths:
      - public

Voila that’s it! Now we have multiple slides deployed to the same GitLab pages site. So I can keep all of my talks nice and neat in a single repo.

Appendix

• My Talks Project

Recently I have been creating a SvelteKit app, when creating a new SvelteKit app you get a choice of different things you can add. Such as using vitest for unit testing.

I needed to spy on/mock a method in a class, to see if it was called when a button was pressed and, it was called with the correct arguments.

Let’s say we have a Button component which looks like this:

<!-- button.svelte -->
<script lang="ts">
    import { API } from "./api"

    const api = new API()
</script>

<button on:click={api.create}>Press Me!</button>

Then the api.ts looks something like this:

export class API {
	async create() {
        // ... does something
	}
}

So how do we test that the create method is called? Let’s assume we will be using the svelte testing library ¹.

// __tests__/button.test.ts
import { render } from "@testing-library/svelte";
import userEvent from "@testing-library/user-event";
import { describe, expect, test, vi } from "vitest";

import Button from "../Button.svelte";
import { API } from "./api";

describe("Button", () => {
	test("Successfully render Button", async () => {
		const user = userEvent.setup();
		const mock = vi.spyOn(API.prototype, "create");

		const { getByRole, getByLabelText } = render(Button, { props: {} });

		const button = getByRole("button", { name: "Add Bookmark" });
		await user.click(button);
		expect(mock).toHaveBeenCalled();
	});
});

The key being this line vi.spyOn(API.prototype, "create"), where we need to use .prototype. So we can access the methods in the class. Without the .prototype we will not find the method on the object itself.

That’s it! Thanks for reading.

Appendix

• Example Test from my App

In this post, I will quickly show you how you can reduce boilerplate code to log in to your app in Playwright tests. Log in to our app is a very common action that will likely be required in most of our tests.

You can read this documentation here, which will explain how you can set this up. However, what do you do if you want to say test the login flow or the register flow? You cannot use the same auth file as all the other tests as you will already be logged in.

It is a great way to reduce boilerplate code from our tests, which would require us to write code to log in before each test. You could log yourself in the test but, then we are adding more boilerplate code again.

What you can do is create a 2nd file say called playwright/.auth/not_logged_in.json which looks like this:

{
	"cookies": [],
	"origins": []
}

In other words there is no state, so the user will not be logged in. Then we can do something like this in say our login.test.ts:

import type { Page, expect, test } from "@playwright/test";

test.describe(() => {
	let page: Page;

	test.beforeEach(async ({ browser }) => {
		const loginContext = await browser.newContext({
			storageState: "playwright/.auth/not_logged_in.json"
		});
		page = await loginContext.newPage();
	});

	test("Successfully login to app", async ({ baseURL }) => {
		await page.goto("/login");

		const email = "test@bookmarkey.app";
		await page.locator('[name="email"]').type(email);

		const password = "password@11";
		await page.locator('[name="password"]').type(password);

		await page.locator('button[type="submit"]').click();
		await page.waitForURL(`${baseURL}/my/collections/0`);
	});

	test("Fail to login to app using incorrect credentials", async ({ baseURL }) => {
		await page.goto("/login");

		const email = "test@bookmarkey.app";
		await page.locator('[name="email"]').type(email);

		const password = "wrong_password";
		await page.locator('[name="password"]').type(password);

		await page.locator('button[type="submit"]').click();

		const toastMessage = await page.locator(".message").innerText();
		expect(toastMessage).toBe("Wrong email and password combination.");
		await page.waitForURL(`${baseURL}/login`);
	});
});

Rather than using the “normal” page argument provided by Playwright, we will create our context. Using this stateless file, will mean the user is not logged in.

If we still want a test to use to automatically logged in we can do:

import { test } from "@playwright/test";

test.describe(() => {
	test("Successfully load collections in side bar", async ({ page, baseURL }) => {
		await page.goto(`${baseURL}/my/collections/0`);
		await page.getByRole("link", { name: "folder closed test" }).click();
		await page.waitForURL(`${baseURL}/my/collections/46lfmlwhymhv6xl`);
	});
});

That’s it!

Appendix

• Used in the Bookmarkey Project

Pocketbase is a backend as a service, it is very nice because it is written in Golang it can compile down to a single binary. It also allows us to extend it using Golang and use it as a framework. In my case, I needed some extra logic before adding data to the database so I decided to extend Pocketbase and write some of my business logic.

Part of the code involved adding some entries to the Database including a relationship. However, I noticed that the relationship looked different in the Pocketbase GUI as compared with the ones I created using the Pocketbase client in the SvelteKit part of my app.

The other relationships showed up a bit differently, they didn’t include the id of the relationship (as a foreign key) but rather a field on the other collection. In the example below the relationship field I am referencing is the bookmark_metadata. Which uses the title field.

Where the bookmark metadata may look like:

We can achieve this using the following code:

bookmarkRecord := models.NewRecord(collection)
bookmarkRecord.Set("bookmark_metadata", []string{metadataRecord.Id})
bookmarkRecord.Set("favourite", false)
bookmarkRecord.Set("collection", collectionID)
bookmarkRecord.Set("custom_order", math.MaxInt32)

The key being that we set bookmark_metadata as a slice (array). However, if we did:

bookmarkRecord.Set("bookmark_metadata", metadataRecord.Id)

In the GUI the record would look like this:

Both options will work for setting up a relationship between two collections. I just wanted to show you how you can set it so it shows up like relationship created via the JS SDK. You can read more about creating records with Golang here.

If you are familiar with Gitlab CI you probably know that jobs in the same stages will run in parallel. However you can also run jobs in different stages at the same time. Let’s see our .gitlab-ci.yml looks like this:

stages:
  - test
  - deploy

format:
  stage: test
  script:
    - task format

lint:
  stage: test
  script:
    - task lint

deploy:preview:
  stage: deploy
  only:
    - merge_request
  image: docker
  script:
    - task deploy

So for my use case when a user creates a merge request, I want to run some tests against the code, such as linting and formatting. But also deploy my app to a preview environment. Whilst these jobs belong to different stages.

The deploy job shouldn’t really depend on the jobs in the test stage as we still want to deploy our app regardless so we can test in the preview environment.

We can do that using needs ¹ keywords, so our deploy:preview job will now look like this:

deploy:preview:
  stage: deploy
  only:
    - merge_request
  needs: []
  image: docker
  script:
    - task deploy

Now our job will run at the same time as the test stage jobs. The needs keyword is used to run our job out of order.

An empty array ([]), to set the job to start as soon as the pipeline is created. - Gitlab CI Docs

Recently I was working on my SvelteKit site and added some code like so to my hook.server.ts:

try {
    if (event.locals.pb.authStore.isValid) {
        await event.locals.pb.collection("users").authRefresh();
        event.locals.user = structuredClone(event.locals.pb?.authStore.model);
    }
} catch (err) {
    console.log("failed to refresh auth", err);
    event.locals.user = undefined;
    event.locals.pb.authStore.clear();
}

I needed to add the structuredClone function so that only POJO (plain old javascript objects) would get stored in the event.locals.users variable. However when I deployed this to production in Netlify. This broke my site! Specifically, it wouldn’t let the user log in.

If you deploy your SvelteKit app on Netlify the server-side code gets turned into functions for us automagically! I was able to work out that structuredClone function broke my app because in my login function I could see something like:

ReferenceError: structuredClone is not defined

It seems that the structuredClone was only added in nodejs 17 ¹. However, Netlify functions default to nodejs 16.

For all Node.js functions deployed on or after July 11, 2022, Netlify uses Node.js 16 as the default runtime. - https://docs.netlify.com/functions/optional-configuration/?fn-language=js

Reading the docs we can see if we set the AWS_LAMBDA_JS_RUNTIME in the GUI to nodejs18.x this will force our functions to use nodejs 18.

And that’s it! By setting our nodejs version to 18, we will be able to use the structuredClone function.

In this post, I will show you how you can autofocus on input that is in a child component. So in my use case, I want the user to click a button to add a “collection” and then it will show the input and immediately focus on it. Which looks something like this:

So how can we do this?

Let’s say we have a child component called Input.svelte which looks like this:

<script lang="ts">
	export let type: string;
	export let name: string;
	export let ref: HTMLInputElement | undefined = undefined;
</script>

<input bind:this={ref} {name} {type} />

The key prop here is ref, which we will use to focus on our input. ref is a reference to the component itself. Then in our parent component say called AddCollection.svelte we can do something like:

<script lang="ts">
	import { tick } from "svelte";
    let ref: HTMLInputElement;
</script>


<button
	on:click={async () => {
		await tick();
		ref?.focus();
	}}>
	Add Collection
</button>
<Input bind:ref type="text" name="addCollection" />

That’s it!

Appendix

• Revelant SO Post

I have recently been creating an app with SvelteKit and Typescript. I noticed all of a sudden Typescript and VS Code not playing nice with each other. It wouldn’t show me the types of variables that I knew it was showing me before. So I started to investigate and work out what was wrong. I was getting locals with a type any:

Even though locals has a defined type in my app.d.ts

I used the documentation to create my SvelteKit app like so:

declare namespace App {
	type PocketBase = import("pocketbase").default;
	interface Locals {
		user?: import("pocketbase").Record | import("pocketbase").Admin | null | undefined;
		pb?: PocketBase;
	}
}

npm create svelte@latest my-app
cd my-app
npm install
npm run dev

This gives us a tsconfig.json file that looks like this:

{
	"extends": "./.svelte-kit/tsconfig.json",
	"compilerOptions": {
		"strict": true,
		"allowUnreachableCode": false,
		"exactOptionalPropertyTypes": true,
		"noImplicitAny": true,
		"noImplicitOverride": true,
		"noImplicitReturns": true,
		"noImplicitThis": true,
		"noFallthroughCasesInSwitch": true,
		"noUncheckedIndexedAccess": true,
		"types": ["vite-plugin-pwa/client"]
	}
}

My tsconfig.json looked like this:

{
	"extends": "./.svelte-kit/tsconfig.json",
	"compilerOptions": {
		"strict": true,
		"allowUnreachableCode": false,
		"exactOptionalPropertyTypes": true,
		"noImplicitAny": true,
		"noImplicitOverride": true,
		"noImplicitReturns": true,
		"noImplicitThis": true,
		"noFallthroughCasesInSwitch": true,
		"noUncheckedIndexedAccess": true,
		"types": ["vite-plugin-pwa/client"]
	},
	"include": ["./setupTest.ts"]
}

The line breaking my configuration was the last include ¹ it turns out it was overwriting the include within the ./.svelte-kit/tsconfig.json. That we were extending above. Which had defined its own include:

{
  "include": [
      "ambient.d.ts",
      "./types/**/$types.d.ts",
      "../vite.config.ts",
      "../src/**/*.js",
      "../src/**/*.ts",
      "../src/**/*.svelte",
      "../src/**/*.js",
      "../src/**/*.ts",
      "../src/**/*.svelte",
      "../tests/**/*.js",
      "../tests/**/*.ts",
      "../tests/**/*.svelte"
  ]
}

Turns out we were just replacing all these files above with just ./setupTest.ts, hence it couldn’t find the types that SvelteKit creates for us ² 🤦‍♂️.

We were of course now overwriting all these files with just the one we specified. Causing VS Code to throw errors and not type things that had types. Removing this line "include": ["./setupTest.ts"] fixed the issue more. In my case, I just moved this file into my tests folder.

I will show you in this post how you can integrate Gitlab, Netlify and Sentry. I may do a more detailed post in the future.

GitLab -> Sentry

To connect Gitlab and Sentry follow this guide. Then add the repos you want to monitor in Sentry. In my case, it is the Bookmarkey GUI.

Netlify -> GitLab

This integration is pretty simple just go through the normal Netlify setup process to add a new site. You can read more about that here and here.

Netlify -> Sentry

Finally, how can we connect Netlify and Sentry?

First, why do we want to do this? It is so we can link specific releases with Netlify deploys. Such as if it was a deploy preview or a deploy to production.

To do this we will be using the sentry-netlify-build-plugin ¹. First, let’s follow the guide in the README.

We can install this via the GUI in Netlify however I prefer to keep as much of my config in code. So we will use a netlify.toml file. Append the following:

[[plugins]]
package = "@sentry/netlify-build-plugin"

[plugins.inputs]
sentryOrg = "majiy"
sentryProject = "bookmarkey"
sentryRepository = "HaseebMajid Apps / Bookmarkey / gui"

Update the specific sentry-related variables to match your project. Remember since SENTRY_AUTH_TOKEN is a secret variable, created when we set up the internal integration with Netlify, you should add that to the Netlify GUI directly.

The other day (a few months ago) I was comparing Goland and VS Code for Golang development. I noticed Goland by default seemed to have nicer syntax highlighting, so I started looking at what I could do in VS Code to do this.

Before

After

It turns out we can do this with our own custom rules using editor.tokenColorCustomizations. To do this go to our settings (settings.json) and add something like:

{
  "editor.tokenColorCustomizations": {
    "[Dracula]": {
      "textMateRules": [
        {
          "scope": "source.go",
          "settings": {
            "foreground": "#6dd8ce"
          }
        }
      ]
    }
  },
}

Where:

• Dracula is the colour scheme
• scope is the type of token we want to colour

We can find out what we should set the scope as by going to the command palette and selecting Developer: Inspect Editor Tokens and Scopes.

Then go and hover over the different parts of our code and it will tell you what type of token it is ¹.

That’s it!

By default PocketBase uses the local file system to store uploaded files. If you have limited disk space, you could optionally connect to a S3 compatible storage. - PocketBase

Why BackBlaze ?

Simple the prices are generally cheaper as compared with other S3 compatible storage platforms. How can we set up BackBlaze as our S3 bucket to store uploaded files such as user avatars?

BackBlaze

Go to BackBlaze

• Create an account
• Create a new bucket
  • You should be able to make it public or private

• Take note of the endpoint i.e. s3.us-west-004.backblazeb2.com.
• Then go to App Keys
  • Create a new Application key

• Take note of the keyID and applicationKey

PocketBase

Go to your PocketBase instance:

• Go to Settings
• Go to File storage

Add the following details

• In ENDPOINT add the endpoint above https://s3.us-west-004.backblazeb2.com
• BUCKET is the name of the bucket we created before ArticleTest
• REGION you can find in the URL us-west-004
• ACCESS KEY is the keyID above
• SECRET is the applicationKey above

Then click Save changes and you should see S3 connected successfully.

That’s It!

In this post, I will go over how you can use docker-compose and Gitlab CI. In this example, we will be running playwright tests directly on the Gitlab runner. The tests will start a SvelteKit server also running on the Gitlab runner. The SvelteKit server will connect to PocketBase (backend) running in docker-compose.

So essentially we need a way for something running locally to connect to something running in docker in Gitlab CI (on a runner). This is a pattern I am using in my new app Bookmarkey ¹.

Let’s pretend we have a docker-compose.yml file which looks something like this:

services:
  pocketbase:
    image: ghcr.io/muchobien/pocketbase:latest
    ports:
      - '9090:8090'
    volumes:
      - ./pb_data:/pb_data
      - /pb_public

and our package.json scripts section looks like this:

{
    "scripts": {
        "test": "playwright test",
    },
}

Finally the most important file let’s look at the gitlab ci file .gitlab-ci.yml:

stages:
  - test

tests:e2e:
  stage: test
  only:
    - merge_request
  image: mcr.microsoft.com/playwright:v1.29.0-jammy
  services:
    - docker:dind
  variables:
    DOCKER_DRIVER: overlay2
    DOCKER_HOST: tcp://docker:2375
    VITE_POCKET_BASE_URL: 'http://docker:9090'
  script:
    - # ... Installing docker and docker compose
    - docker compose up -d
    - npm run test

Since all Gitlab CI jobs run in Docker. If we want to run docker-compose inside a job we need to use the dind service ². We can then use docker normally i.e. using docker compose to start PocketBase. The most important line here is normally to connect to PocketBase I would use http://localhost:8090. However, since we are using the dind service we need to use docker instead of localhost. Hence this VITE_POCKET_BASE_URL: 'http://docker:9090', passed to my SvelteKit app.

I spent about two days debugging this issue, even though I’d solved this problem before 🤦‍♂️. So I decided to make a quick post so hopefully you can avoid wasting your time.

If you try to use prettier on a (Hugo) Golang HTML template you may get something that looks like this:

{{- if or .Params.author site.Params.author }}
  {{- $author := (.Params.author | default site.Params.author) }}
  {{- $author_type := (printf "%T" $author) }}
  {{- if (or (eq $author_type "[]string") (eq $author_type "[]interface {}")) }}
    {{- (delimit $author ", " ) }}
  {{- else }}
    {{- $author }}
  {{- end }}
{{- end -}}

into this

{{- if or .Params.author site.Params.author }} {{- $author := (.Params.author |
default site.Params.author) }} {{- $author_type := (printf "%T" $author) }} {{-
if (or (eq $author_type "[]string") (eq $author_type "[]interface {}")) }} {{-
(delimit $author ", " ) }} {{- else }} {{- $author }} {{- end }} {{- end -}}

This of course something we don’t want. So let’s use a prettier plugin that can solve this problem for us. Let us install npm install --save-dev prettier-plugin-go-template.

Create a .prettierrc file with the following contents:

{
  "overrides": [
    {
      "files": ["*.html"],
      "options": {
        "parser": "go-template"
      }
    }
  ],
  "goTemplateBracketSpacing": true
}

Prettier will now format our HTML files correctly, in our Hugo project.

Optional Settings

By default the format of our files like so:

<a
  href="https://github.com/reorx/hugo-PaperModX/"
  rel="noopener"
  target="_blank"
></a>

If we add the following option to the .prettierrc, "bracketSameLine": true then our file will look like:

<a
  href="https://github.com/reorx/hugo-PaperModX/"
  rel="noopener"
  target="_blank"
></a>

Sometimes our file may look something like this:

<span>
  Analytics by
  <a href="https://{{ .Site.Params.goatcounter }}.goatcounter.com"
    >Goatcounter.</a
  >
</span>

To fix the dangling > we can add the following option to the .prettierrc file , "htmlWhitespaceSensitivity": "ignore" then our file will look like:

<span>
  Analytics by
  <a href="https://{{ .Site.Params.goatcounter }}.goatcounter.com">
    Goatcounter.
  </a>
</span>

So you have a SvelteKit site you want to deploy to Netlify how can we do that? First, go about setting up a site like normal! In my case, I am going to import an existing site from Gitlab.

Netlify

In the build settings we want the following settings:

That’s it now our site should start deploying, and it should also set up deploy previews for MRs automatically More about that here.

Our Project

Within our SvelteKit project if you look at svelte.config.js it should look something like this:

import adapter from '@sveltejs/adapter-auto';
import preprocess from 'svelte-preprocess';

/** @type {import('@sveltejs/kit').Config} */
const config = {
  // Consult https://github.com/sveltejs/svelte-preprocess
  // for more information about preprocessors
  kit: {
    adapter: adapter(),
  },
  preprocess: [
    preprocess({
      postcss: true
    })
  ]
};

export default config;

You do not need to use the Netlify adapter here unless we need to specify specific options such as edge functions etc. However to let SvelteKit know we are using the Netlify adapter we need to pass an environment variable ¹.

We can also create a netlify.toml file repository to keep the settings locally rather than just on Netlify. The advantage of this is keeping our settings in code (and version control).

[build]
publish = "build"
command = "npm run build"

[context.production.environment]
NETLIFY = "true"

Sometimes we want to be able to run our unit tests and integration tests separately. In Golang we can do this using build tags, build tags are used to tell the compiler important information when we run go build ¹.

Let say we have a file called package_test.go. By adding // +build integration to the top of the file without any whitespace. This test file will only be run when we specify the tags in our test command go test --tags=integration.

Our integration test file would look something like:

// +build integration

package mypackage_test