---

1.5.2 Elisp Interpreter

The cornerstone of Emacs though is the elisp interpreter — without it, there is no Emacs; speed and asynchrony are the two main issues. The interpreter runs in a single thread and intensive tasks will lock the UI thread. But there are workarounds.

When you write elisp you are not just writing snippets of code run in a sandbox, isolated from everything — you are altering a living system; an operating system running on an operating system. Every variable you alter and every function you call is carried out by the very same interpreter you use when you edit text.

Emacs is a hacker’s dream because it is one giant, mutable state. Its simplicity is both a blessing and a curse. You can re-define live functions; change variables left and right; and you can query the system for its state at any time — state that changes with every key stroke as Emacs responds to events from your keyboard to your network stack. Emacs is self-documenting because it is the document. There are no other editors that can do that. No editor comes close.

And yet Emacs never crashes — not really, anyway. Emacs has an uptime counter to prove that it doesn’t (M-x emacs-uptime) — multi-month uptimes are not uncommon.

So when you ask Emacs a question – as I will show you how to do later – you are asking your Emacs what its state is. Because of this, Emacs has an excellent elisp debugger and un-limited access to every facet of Emacs’s own interpreter and state — so it has excellent code completion too. Any time you encounter a lisp expression you can tell Emacs to evaluate it, and it will: from adding numbers to setting variables to downloading packages.