# Time to Stop Saying "Pure/Virtual" Binding

**URL:** <https://revolt.metaeducation.com/forum/t/time-to-stop-saying-pure-virtual-binding/2441>\
**Category:** Naming\
**Created:** [May 11, 2025, 2:53am UTC](https://revolt.metaeducation.com/forum/t/time-to-stop-saying-pure-virtual-binding/2441 "2025-05-11T02:53:29Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![hostilefork](https://revolt.metaeducation.com/forum/user_avatar/revolt.metaeducation.com/hostilefork/32/421_2.png) [@hostilefork](https://revolt.metaeducation.com/forum/u/hostilefork)\
**Post date:** [May 11, 2025, 2:53am UTC](https://revolt.metaeducation.com/forum/t/time-to-stop-saying-pure-virtual-binding/2441/1 "2025-05-11T02:53:29Z")

</div>

Rebol's historical model of binding was to mutably paste hidden "binding" pointers onto WORD!s in code, that influenced the lookup... which isn't always bad. But what I think was definitely bad was that _this was the only binding there was._

There were big weaknesses. Such as that if WORD!s in a method in a function body had to resolve to a specific variable, you'd have to copy and re-paste bindings on bodies of methods to create a new instance of that object. **That was crazy.** 🤪

Yet sometimes WORD!s had to carry knowledge of where they looked up to. The trick became knowing when this binding should be carried by the word, and when it shouldn't.

- As these ideas evolved, I used **"virtual binding"** as a sort of blanket term to mean that the WORD!s themselves may be unbound, but it was things like BLOCK! instances that carried contextual/enviroment information... which the unbound words could be looked up in at arbitrary points, taking into account any new definitions vs. being stucked in a fixed concept from a prior "wave" of mutable binding.

- When this idea escalated to its logical conclusion, I called it **"pure virtual binding"** , to imply that when you start running module or script code from the top _none_ of the code carries a binding. When everything comes after the fact, you get things like:

## These Terms Have Become Meaningless

**"virtual binding"** and **"pure virtual binding"** are nothing but historical terms.

There is only **"binding"** , and everything else just a transient term used while trying to pin down how it should act.

I particularly want to reclaim the term **PURE** because we need to start marking functions that don't have side-effects. _(Oddly enough, ["Rye" has beaten me to the punch with actually deploying this](https://revolt.metaeducation.com/forum/t/rye-language/1768/6), though I've advocated for it a long time before that...just never got around to it, for whatever reason.)_

But reclaiming the term **VIRTUAL** is probably useful as well.

## Please Excuse WikiMan As He Edits Your Posts

As I Wiki-Edit this forum to try and make it non-useless for modern audiences, expect to see the terms **"virtual binding"** and **"pure virtual binding"** [be relegated to the Archive](https://revolt.metaeducation.com/forum/c/archive/61)

In spirit we _could_ still say that binding does have "virtual" and "non-virtual" components, but there's no need to waste the term "virtual". We just say that a BLOCK! might be bound, while the WORD!s in it could be bound--or unbound. If the words in it are BOUND! and you project the binding, then you're acting "virtually".

Anyway, I am going to stop using the terms...and so should you _(though really, has anyone else besides me ever uttered the phrases: "virtual binding" or "pure virtual binding"? probably not 🫤)_
