RX-wiki

Differential Perma-Cache

It consists on using a tool (such as Git) to download a project (either completely or partially), using the same tool for delta-updates, and using a light-weight IDE (such as Helix or NeoVim) to browse the project (of course, having a text shell (such as Bash) and Unix-like utilities (such as GNU coreutils) is implied).

The purpose of such a setup is to enable an efficient workflow for general-purpose information retrieval:

• Most files are available locally, which allows fast opening and saves network bandwidth
• User can use whatever search-engine they want, or whichever is most appropriate, such as: grep, Ripgrep, fzf, find, etc…
• If the VCS is Git, blobs will be cached permanently

All of this is in contrast to web browsers:

• Data might not be available locally, because the browser deletes old cache
• User is at the mercy of whatever search-engine is provided by a website, or the browser’s own “find in page” (ctrl+f) feature
• Cache is ephemeral, even if a server never changes its response

Overall, a much better experience!