cafkafk/nix-weather
cafkafk/nix-weather
1
1
Fork 0
Code Issues 5 Pull requests Projects Releases 3 Packages Activity Actions

docs: move usage before how it works
Some checks failed
Conventional Commits / Conventional Commits (pull_request) Has been cancelled
Details

Browse source 
Signed-off-by: Christina Sørensen <christina@cafkafk.com>
This commit is contained in:
Christina Sørensen 2024-09-05 09:44:01 +02:00
parent cc785d16d4
commit 95d53c5c9a
Signed by: cafkafk
GPG key ID: 26C542FD97F965CE
1 changed files with 13 additions and 14 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

27
README.md
View file

@ -21,20 +21,6 @@ weather`](https://guix.gnu.org/manual/en/html_node/Invoking-guix-weather.html).
</div> </div>
## How It Works
The basic idea is that we construct a set of all requisites to build the
top-level of a NixOS configuration, and then query Nix cache(s) for the narinfo.
By doing this in a high concurrency, parallel task runner (i.e. tokio async
runtime), and only querying the headers for status codes, we can reach
impressive speeds, typically around 45~ network time.
One of the biggest limiting factors regarding speed is building the
`config.system.toplevel`, and finding the necessary requisites with `nix-store`.
Caching the requisites is a future goal, so that we only have to build the
`toplevel`, and then match against its derivation in cache, which should cut
down the nix part of the runtime by ~80%.
## Usage ## Usage
> **Note** > **Note**
@ -49,3 +35,16 @@ nix-weather --name myhost --config ~/git/my-nixos-config
Here, we specify the name of the host, as specified at the flake output Here, we specify the name of the host, as specified at the flake output
`nixosConfiguration`, and a path to the NixOS configuration flake. `nixosConfiguration`, and a path to the NixOS configuration flake.
## How It Works
The basic idea is that we construct a set of all requisites to build the
top-level of a NixOS configuration, and then query Nix cache(s) for the narinfo.
By doing this in a high concurrency, parallel task runner (i.e. tokio async
runtime), and only querying the headers for status codes, we can reach
impressive speeds, typically around 45~ network time.
One of the biggest limiting factors regarding speed is building the
`config.system.toplevel`, and finding the necessary requisites with `nix-store`.
Caching the requisites is a future goal, so that we only have to build the
`toplevel`, and then match against its derivation in cache, which should cut
down the nix part of the runtime by ~80%.