IA S/W Emulation

How does this all work

The files are grouped and split up into 3 GitHub repositories (linked below) and corresponding deploys.

GitHub Actions

Internet Archive is using docker OCI containers to serve the repo files as static files, with a lightly customized nginx httpd server. Each deployed container is running on a nomad cluster.

Each git push to this repo will cause a GitHub Actions CI/CD pipeline to run. Presently, it does a [build] and [deploy] pair of jobs. (See CI/CD GitHub Action for more info.)

loader.js

The main ringleader from an archive.org/details/ page for a s/w emulation "item" will load loader.js.

browserFS

There is a parallel browserfs.min.js file that gets loaded alongside loader.js. It is used for node backend-like FileSystem operations like reading and writing files, in the context of the browser. Tracey thinks this came from here:

• https://www.npmjs.com/package/browserfs and notes the browserfs.min.js* file pair have a fix from db48x after the v1.4.3. Likely worth looking into the near future if/as needed. browserFS can use "ES Modules" for JS import/loading via something like:

import BrowserFS from 'https://esm.sh/browserfs'

nginx

Each repo has the same layout, with all files starting at the top dir:

• README.md - this file
• default.conf - nginx configuration
• Dockerfile - container build instructions
• .dockerignore - files to skip copying into the deployed container
• .gitignore - any files to ignore and not checkin
• .github/workflows/cicd.yml - GitHub Actions setup We aim to keep all 3 repos with the same 6 files.

the nginx has a few critical changes for the handling of .gz files -- since nginx can also "gzip on the fly" text files as well. Thus, the browser needs to know if the file is being dynamically turned to a .gz file by nginx -- or if the file is being served as an already compressed .gz as a static file.

example items

• typical type of item: https://ia-petabox.archive.org/details/gg_Star_Wars_1993LucasArts_Tiertex_Design_Studio_U.S._GoldUS

• most challenging new ruffle type of item: https://ia-petabox.archive.org/details/flash_loituma

current deployments - production, uses the production branch:

• https://internetarchive-emularity-engine.dev.archive.org/
• https://internetarchive-emularity-config.dev.archive.org/
• https://internetarchive-emularity-bios.dev.archive.org/

current deployments - testing, this one is for the main branch:

• https://emularity-engine.ux-b.archive.org/
• https://emularity-config.ux-b.archive.org/
• https://emularity-bios.ux-b.archive.org/

there are 3 related git repos & deployments:

• https://github.com/internetarchive/emularity-engine
• https://github.com/internetarchive/emularity-config
• https://github.com/internetarchive/emularity-bios

prior archive.org source items with setup files:

• https://archive.org/serve/emularity_engine_v1
• https://archive.org/serve/emularity_config_v1
• https://archive.org/serve/emularity_bios_v1