While I would like my CI build to closely resemble a non-CI build, supporting
the `all-the-icons-install-fonts` call is a low priority with a medium amount of
work required.
After my CI build for Emacs failed because the .local/share/wallpaper directory
was missing I had two options:
A. include .local/share/wallpaper in default.nix, which is cumbersome
B. drop support for managing system wallpaper from Emacs
I chose option B.
I'm starting to prefer the `inherit (builtins) path` pattern in my Nix
expressions. I know this is idiomatic, so even if I don't like it, I am trying
to learn to like it.
These were hard-coded as $HOME/BRIEFCASE, which won't work in CI, since CI runs
as the user buildkite-agent-socrates, whose $HOME directory doesn't exist.
Instead of manually maintaining the list of directories that I expose to
readTree, I'm using `builtins.readDir` to get a list of all non-hidden top-level
directories.
TL;DR:
- Define runEmacsScript to emacs/default.nix for ci/pipelines/post-receive
- Write script.el to call (load init.el) and catch any errors
- Lint Elisp with gonewest818/elisp-lint
Also nice how Buildkite supports :gnu: emojis!
- Prefer prepending wpcDir, vendorDir to EMACSLOADPATH instead of using the
--directory flag
- Remove --load ${wpcPackageEl} because init.el calls (require 'wpc-package)
- Surround $@ in 2x-quotes
Following the advice of Domen's nix.dev anti-patterns, I'm preferring something
like...
```nix
builtins.path { path = /path/to/some.where; name = "some.where"; }
```
...to
```nix
/path/to/some/where
```
While the former is more verbose, it will fail to build when the path doesn't
exist, which I prefer.
I would prefer to define constants/briefcase in terms of `(getenv "BRIEFCASE")`
and assert that `(f-exists? (getenv "BRIEFCASE"))`, in one location:
constants.el
TL;DR:
- Prefer `(getenv "BRIEFCASE")` to `(f-expand "~/briefcase")`. I should audit my
Emacs for references to ~/briefcase and replace those calls with `getenv`.
- Remove calls setting <nixpkgs> and <depot> and rely exclusively on <briefcase>
- Prefer ~/nixpkgs-channels to ~/nixpkgs.
Notes:
- I need a better way of calling `home-manager switch` that resides within my
briefcase
I wanted Gitea to call Buildkite's pre-receive pipeline and either accept or
reject the incoming code depending on the outcome. The problem is that I can
only *create* builds from Gitea's pre-receive hook.
Now I'm left with two options:
1. run the lint-secrets step in post-receive
2. run `/nix/store/<hash>/git-secrets --scan-history $REPO_PATH` in Gitea
As far as I can tell, I cannot define Gitea hooks in Nix, which is unfortunate;
otherwise, option 2 would appeal more.
I'm doing option one for now.
So it turns out that I was wrong and that .git/config is stateful. Multiple
calls to --add-provider will append the same provider each time...
Instead I'm defining secret-patterns.txt and version-controlling it.
Then:
- dev-side: I'm adding `providers = cat ci/secret-patterns.txt` to .git/config
- ci-side: I'm adding `providers = cat ci/secret-patterns.txt` to .git/config
Unfortunately this is ad-hoc configuration ci-side, which I would like to
avoid. The good news is that my pre-commit hooks and failures from git-secrets
should now align with my CI, since they're both reading from
secret-patterns.txt. One step backwards... two steps forwards?
I'm also `cat .git/config` because I think the Buildkite destroys the
.git/config file for each build, but I want to verify that. If it does, I prefer
that because it seems to share the spirit of the "Destroy Your Darlings" essay.
Problem: my dev machine returns a different value for `git config --get-all
secrets.patterns` than my CI machine... I ran `git-secrets --register-aws` to
get additional coverage, but it's still not the same. I created an issue on the
git-secrets GH repo to get better troubleshooting advice, but I don't need the
logging info. anymore, so I'm removing it.
Somehow `git-secrets --scan-history` is exiting non-zero, when I don't think it
should. Logging some environment information to get a better idea of what's
going on.
After a handful of failed attempts to run lint-secrets.sh due to a missing
`git-secrets` executable on my git server, I decided that now was a good time to
use Nix to define my BuildKite pipelines.
TL;DR:
- Delete ci/scripts directory
- Define ci/pipelines/{briefcase,socrates}.nix
Outside of this repository:
- I logged into my admin account at git.wpcarro.dev and changed my Gitea
post-receive hook to trigger the briefcase pipeline
- I logged into my BuildKite account, deleted my build-briefcase pipeline,
created a new briefcase pipeline that called:
```shell
nix-build -A ci.pipelines.briefcase -o briefcase.yaml
buildkite-agent pipeline upload briefcase.yaml
```
One day I will audit all of my ad-hoc, non-mono-repo activity (like the steps I
listed above) and attempt to fit everything herein... one step at a time,
though!
Adding a fake secret to test to the new CI build step. I'm not sure I expect
this to fail the step because it relies on a pattern that I defined in
.git/config... let's see!
Having `git secret hide` as a pre-commit hook doesn't make much sense to me. I
will detail why when/if I write a blog post on briefcase's secret mgt setup.
The problem is, if I change secrets.json and then run `git status`, I won't see
any pending changes. This is because secrets.json is gitignore'd. If I run `git
secret hide` everytime I save secrets.json, I can rest assured that my `git
status` will be consistent with any updates to secrets.json.
I'm attempting to maintain a top-level secrets.json that defines all of the
sensitive data that I'd like to version-control without exposing everything in
cleartext to the world. To that end, I'm using `git secret`, which will use
`gpg` to encrypt secrets.json everytime I call `git secret hide` and decrypt
everytime I call `git secret reveal`.
I'm going to try this until I don't like it anymore... if that day comes...
I should write a blog post about my setup to solicit useful feedback and share
my ideas with others.
