profile
viewpoint
Charles charles-l Allentown, PA https://c.har.li/e programming rascal. more code at: http://git.sr.ht/~nch

akkartik/mu 730

Soul of a tiny new machine. More thorough tests → More comprehensible and rewrite-friendly software → More resilient society.

charles-l/dirt 24

x86 assembler in scheme

charles-l/dbg 8

a linux debugger

charles-l/gamedev_tutorial 8

Code for my game development series on youtube.

charles-l/i 7

minimal chicken scheme irc client

charles-l/dotfiles 6

all my dots

charles-l/intro_to_c_slideshow 6

Intro to C for the High-Level programmer

charles-l/capstone 2

a programming language implementation guide (see release page for final pdf)

charles-l/cforpyprogrammers 2

A series of assignments for Python programmers to learn C.

charles-l/crystal-gl 2

OpenGL bindings for the Crystal language

startedcreme-ml/creme

started time in 15 hours

startedyeraydiazdiaz/lunr.py

started time in 16 hours

startedaidenybai/lucia

started time in 5 days

startedEmbarkStudios/rust-gpu

started time in 6 days

startedftilde/ugdb

started time in 6 days

startedyrp604/rappel

started time in 6 days

startedbadtuple/strats

started time in 9 days

issue openedanordal/selfdock

segfaults qira docker container

selfdock is awesome! Works great as a lightweight sandboxing for binaries on my system, and I'll probably be using it a lot as a replacement to docker. Thanks for building a delightfully simple piece of software like this :)

I tried to really see how far I could push it when running a docker container for qira.

After building the image, I ran:

$ docker save qira | undocker -i -o /tmp/qira
$ HOME=/root selfdock --rootfs qira run qira /bin/ls
*** program is /bin/ls with hash b79f70b18538de0199e6829e06b547e079df8842
** found 130 names
*** elf loaded
**** using /qira/tracers/qemu/qemu/x86_64-linux-user/qemu-x86_64 for 0x3e
no qira server found, starting it
*** deleting old runs
**** running /bin/ls
****** starting WEB SERVER on 0.0.0.0:3002
qemu: uncaught target signal 11 (Segmentation fault) - core dumped
on 0 going from 1 to 86529...*** using base 0 for 0
*** WARNING, changing segment 0x400000 4315748
Failed to get [vvar] offset 0 : [Errno 2] No such file or directory: '[vvar]'
Failed to get [vdso] offset 0 : [Errno 2] No such file or directory: '[vdso]'
*** started analysis_thread 0
qira: Bus error

I haven't had a chance to dig into what's segfaulting yet, but I wanted to at least mark down the issue. If this is something you think should be possible to run with selfdock I could take a look at debugging it a bit further.

created time in 10 days

issue openedanordal/selfdock

Environment variables are inherited by container

For my usecase at least, it would be useful to have an option to clear the environmental variables for the container like docker containers do...

This is probably considered an enhancement (but could also be considered a security improvement if users are carrying around secrets in their envs). I might take a crack at implementing it if I have a bit of time soon.

created time in 10 days

push eventcharles-l/undocker

nc

commit sha 8707e7dc4bb5ccc6883af156571e2e177afdf09f

Add message to clarify when non-existent docker image archive is given

view details

nc

commit sha 53c062300aef3dbd1a56b468394f3947c558898e

Update README with latest parameters and fixed examples.

view details

push time in 10 days

issue commentlarsks/undocker

Getting IsADirectoryError when trying to run examples

Ohh just checked the code. It looks like no argument is required when reading from stdin anymore? I've just tested it and it seems to work. I'll update the error message to clarify that a bit more...

charles-l

comment created time in 10 days

issue openedlarsks/undocker

Getting IsADirectoryError when trying to run examples

undocker -i -o seems to be broken.

When I try to run the example from the README:

% docker pull busybox
Using default tag: latest
latest: Pulling from library/busybox
Digest: sha256:a9286defaba7b3a519d585ba0e37d0b2cbee74ebfe590960b0b1d6a5e97d1e1d
Status: Image is up to date for busybox:latest
docker.io/library/busybox:latest
% docker save busybox | undocker -i -o busybox busybox
Traceback (most recent call last):
  File "/home/nc/.local/bin/undocker", line 8, in <module>
    sys.exit(main())
  File "/home/nc/.local/lib/python3.8/site-packages/undocker.py", line 99, in main
    open(args.image, 'rb') if args.image
IsADirectoryError: [Errno 21] Is a directory: 'busybox'

Running undocker 7:

% pip3 show undocker
Name: undocker
Version: 7
Summary: Unpack docker images
Home-page: http://github.com/larsks/undocker
Author: Lars Kellogg-Stedman
Author-email: lars@oddbit.com
License: UNKNOWN
Location: /home/nc/.local/lib/python3.8/site-packages
Requires:
Required-by:

created time in 10 days

startedyoshinorim/quickstack

started time in 10 days

startedanordal/selfdock

started time in 10 days

starteddandavison/delta

started time in 10 days

startedGloriousEggroll/proton-ge-custom

started time in 10 days

startedelucent/jasmine

started time in 10 days

startedzyantific/zydis

started time in 10 days

startedcmfcmf/MagicMouse

started time in 10 days

startedConnorNelson/pypreload

started time in 10 days

startedalexheretic/glyph-brush

started time in 10 days

startedpyglet/pyglet

started time in 15 days

PR opened pyglet/pyglet

Fix data type for text_input.py example.

The (older?) vertices2i and colors4Bn type specifiers were causing this example to blow up for me. This patch fixes the text_input.py script.

+2 -2

0 comment

1 changed file

pr created time in 15 days

push eventcharles-l/pyglet

nc

commit sha 659f04eff93ce248ebe68ea3b5d48c69f94c1611

Fix data type for text_input.py example.

view details

push time in 15 days

fork charles-l/pyglet

pyglet is a cross-platform windowing and multimedia library for Python, for developing games and other visually rich applications.

http://pyglet.org

fork in 15 days

startedr-lyeh/FWK

started time in a month

startedrbanffy/3270font

started time in a month

startedfenollp/reMarkable-tools

started time in a month

startedkisonecat/lerminus

started time in a month

startedparasyte/pixels

started time in a month

startedlibrespot-org/librespot

started time in a month

startedshellphish/driller

started time in a month

startedCrushedPixel/Polyline2D

started time in a month

startedmeshtastic/Meshtastic-device

started time in 2 months

startedcytecbg/gpucanvas

started time in 2 months

startedgeohot/qira

started time in 2 months

startedzardus/shelltools

started time in 2 months

startedcanselcik/libremarkable

started time in 2 months

startedruscur/snowpatch

started time in 2 months

startedorndorffgrant/bnhook

started time in 2 months

startedzetzit/zz

started time in 2 months

startedIAmStoxe/wirehole

started time in 2 months

startedDataDog/scipio

started time in 2 months

startedChevyRay/crunch-rs

started time in 2 months

startedmooman219/fontdue

started time in 2 months

startedflorentc/xob

started time in 2 months

startedalandefreitas/matplotplusplus

started time in 2 months

startedtjammer/raylib-ocaml

started time in 2 months

startedDiggsey/spanr

started time in 2 months

startedfonsp/Pluto.jl

started time in 2 months

startedu2zv1wx/neut

started time in 2 months

startedbpatrik/pigallery2

started time in 2 months

startedbettercap/bettercap

started time in 2 months

startedevilsocket/pwnagotchi

started time in 2 months

startedtiffany352/rink-rs

started time in 2 months

startedluvsound/pippi

started time in 2 months

startedkquick/Thespian

started time in 2 months

startedept221/tinySoC

started time in 2 months

starteddbandstra/zang

started time in 2 months

push eventcharles-l/kiss3d

nc

commit sha 73d325d271c6b0e8cecca736b3b8726c086f9202

remove completed FIXME

view details

push time in 2 months

pull request commentsebcrozet/kiss3d

Add basic cubemap support.

@sebcrozet some quick questions for you when you get a chance:

  1. Is WASM support ready? I couldn't get the example project to build, and I saw some recent activity around WASM support so I'm not sure if it's still in progress. But if it's ready, I'll dig into my issue further and make sure this PR supports WASM. Otherwise, I think it makes sense to make web cubemap support a secondary PR later once WASM stuff is fully fleshed out.
  2. Currently, the Cubemap texture is stored in the CubemapMaterial structure rather than the ObjectData.texture field (https://github.com/sebcrozet/kiss3d/pull/240/files#diff-d6cc39930e971e874f1b95f5064596cbR31). This is certainly at odds with the rest of the API design, but I'm not sure of a better way of implementing this at the moment. I wanted to avoid slapping a new member on the ObjectData field, or making the ObjectData.texture field a variant type. Besides, cubemaps are generally bound to cube meshes, so I'm not sure putting cubemap info in the generic ObjectData struct which can be associated with other meshes quite makes sense either. Maybe it would be worth creating an entirely new type with a hardcoded cube mesh and cubemap texture?
  3. The build is failing on the rustfmt step. I tried running rustfmt across the project, but that changed almost every file, and I didn't want to submit a noisy diff. Before merging, I can run it across the new files I've added, but I'm not sure that'll fix the build.

Sorry for the big PR -- this ended up unfurling into more changes than I was expecting. When I convert it from a draft PR, I can create a few stacked PRs for more sane chunks of work (i.e. make bind_texture only use TEXTURE_2D, add cubemap support to the context, add the Cubemap material, etc).

charles-l

comment created time in 2 months

push eventcharles-l/kiss3d

nc

commit sha c8238bac75fc3ea9b6aa9ff84d6d194ab1de0185

tighten up the bind_texture/bind_cubemap functions

view details

push time in 2 months

push eventcharles-l/kiss3d

nc

commit sha 3712c412e920c6fba1d08c7d5de2194d6e873d8c

add docs

view details

push time in 2 months

startedbintracker/bintracker

started time in 2 months

pull request commentsebcrozet/kiss3d

Add basic cubemap support.

Ran a quick profiling test on the cube_map.rs example to try and figure out what the startup time was coming from. Apparently, the PNG decoder is pretty slow (according to my profiling results 96% of the startup time is spent in the png decode function). Probably not worth digging into at the moment. I just noticed that a lot of time was being spent in the InflateStream code, so this issue might be related https://github.com/image-rs/image-png/issues/114

Anyway, it goes beyond the scope of this PR and the only short-term modification I can think of making is multithreading the cubemap loading function which is probably more trouble than it's worth.

charles-l

comment created time in 2 months

startedgwihlidal/meshopt-rs

started time in 2 months

push eventcharles-l/kiss3d

nc

commit sha f6a924f307a0248a1acff0e9e00b44595621ef7a

make cubemap material builtin

view details

push time in 2 months

push eventcharles-l/kiss3d

nc

commit sha 626234b4ab14413f6d151574aabe3be2853c0b55

add cubemap images

view details

push time in 2 months

startedhoffstadt/DearPyGui

started time in 2 months

startedolive-editor/olive

started time in 2 months

startedfcpg/vim-waikiki

started time in 2 months

startedraphlinus/font-rs

started time in 2 months

push eventcharles-l/kiss3d

nc

commit sha bceb5e3e85a0ba2ab57c11808b6b7496179ca563

make cubemap API more typesafe

view details

push time in 2 months

startedmackron/dr_libs

started time in 2 months

startedzoidbergwill/awesome-ebpf

started time in 2 months

startedGaijinEntertainment/daScript

started time in 2 months

startedhaze/ambrosia.vim

started time in 2 months

startednschloe/pygmsh

started time in 3 months

startedAirGuanZ/imgui-filebrowser

started time in 3 months

startedredox-os/orbtk

started time in 3 months

startedNatronGitHub/Natron

started time in 3 months

startedjesseduffield/horcrux

started time in 3 months

startedh3rald/min

started time in 3 months

pull request commentsebcrozet/kiss3d

Add basic cubemap support.

Hmm... WASM support might be tricky -- I'm having difficulty getting the example WASM project to compile...

charles-l

comment created time in 3 months

startedbevyengine/bevy

started time in 3 months

startedbjorn3/rustc_codegen_cranelift

started time in 3 months

startednot-fl3/macroquad

started time in 3 months

startedfacebookincubator/propfuzz

started time in 3 months

startedstateright/stateright

started time in 3 months

starteddr-soft/glbind

started time in 3 months

startedsomerandomdev49/card

started time in 3 months

startedcisco/openh264

started time in 3 months

startedRadivarig/UvSquares

started time in 3 months

startedEmbarkStudios/puffin

started time in 3 months

startedJoshua-Ashton/libjobs

started time in 3 months

startedfacebook/pyre-check

started time in 3 months

startedservo/rust-smallvec

started time in 3 months

startedphaazon/luminance-rs

started time in 3 months

more