Skip to content

Instantly share code, notes, and snippets.

View wolfv's full-sized avatar
🚀

Wolf Vollprecht wolfv

🚀
532 followers · 85 following
View GitHub Profile
@wolfv
wolfv / recipe.yaml
Created April 1, 2024 18:06
DOSBOX recipe for rattler-build
package:
name: dosbox
version: "0.74.0"
source:
url: https://downloads.sourceforge.net/project/dosbox/dosbox/0.74-3/dosbox-0.74-3.tar.gz
sha256: c0d13dd7ed2ed363b68de615475781e891cd582e8162b5c3669137502222260a
build:
number: 0
@wolfv
wolfv / fetch.sh
Created September 8, 2021 13:08
Fetch from OCI registry (ghcr.io)
export TOKEN=$(curl --silent https://ghcr.io/token\?scope\=repository:wolfv/artifact:pull | jq -r .token)
curl \
--silent \
--request 'GET' \
--header "Authorization: Bearer $TOKEN" \
--header "Accept: application/vnd.oci.image.manifest.v1+json" \
'https://ghcr.io/v2/wolfv/artifact/manifests/1.0'
echo "\n\nFetching image content now:\n\n"
@wolfv
wolfv / github_actions.yaml
Last active March 10, 2024 15:28
micromamba usage
name: CI
on:
push:
branches:
- main
pull_request:
branches:
- main
@wolfv
wolfv / run_conda_builds.py
Created July 28, 2019 18:13
Run many conda builds in parallel, as fast as possible.
from pathlib import Path
import jinja2
import yaml
from conda.models.match_spec import MatchSpec
import networkx as nx
# import matplotlib.pyplot as plt
import os, random, time, json
import workerpool
NO_BUILD_LOCAL_PKGS = True
@wolfv
wolfv / benchr.R
Last active June 14, 2023 18:14
R benchmark RcppArmadillo vs RcppEigen vs xtensor-R
Rcpp::sourceCpp("benchr.cpp")
x <- matrix(as.double(1:10000), 10)
library("rbenchmark")
benchmark(
"arma" = {
z <- somefunc_arma(x)
},
@wolfv
wolfv / ROS Noetic packages on conda.html
Created February 13, 2021 09:57
Available ROS noetic packages on conda / robostack channel
<!DOCTYPE html>
<head>
<meta charset="UTF-8">
<style>
.r1 {color: #800080; font-weight: bold}
.r2 {color: #008000}
.r3 {color: #800000}
body {
color: #000000;
background-color: #ffffff;
@wolfv
wolfv / cmake_remove_duplicate_static_libs.py
Created March 12, 2023 19:14
from pathlib import Path
ltxt = Path("_rclpy_pybind11.dir/link.txt")
cmd = ltxt.read_text().split()
final_cmd = []
for c in cmd:
if c.startswith("@CMakeFiles"):
p = c.removeprefix("@CMakeFiles/")
@wolfv
wolfv / graphql_upload_file.py
Last active January 23, 2023 15:37
Upload a file with Python, Requests and GraphQL
import requests
import json
url = 'http://localhost:3001/api/graphql'
cookies = {}
mutation = '''mutation($file: Upload!) {
uploadEnv(environmentFileUpload: $file) {
id
@wolfv
wolfv / testdl.bat
Created January 21, 2023 23:33
mkdir -p ~/.local/bin
set PLATFORM=win
set ARCH=64
curl -Ls https://micro.mamba.pm/api/micromamba/%PLATFORM%-%ARCH%/latest | tar -xvj -C ~/.local/bin/ --strip-components=1 Library/bin/micromamba.exe
@wolfv
wolfv / benchmark.md
Created January 17, 2023 13:33
Raw benchmark values for `micromamba` 1.0 vs 1.2 vs `conda-package-handling`

Jaxlib

micromamba 1.2.0

➜ hyperfine "micromamba package extract jaxlib-0.4.1-cpu_py38h6beaf4d_1.conda outmmjaxnew" --warmup 3
Benchmark 1: micromamba package extract jaxlib-0.4.1-cpu_py38h6beaf4d_1.conda outmmjaxnew
  Time (mean ± σ):     227.2 ms ±  17.4 ms    [User: 158.4 ms, System: 52.8 ms]
  Range (min … max):   212.9 ms … 267.5 ms    13 runs