Gabriel Nordeborn

Coercion for records also break this assumption (more fields can be on the object than what's represented by the record). Honestly I think these are just quirks we'll have to...

> If so, we should deprecate supporting comparison operations (`==`, `>`, `=`, `

Actually, let's enable project cache config too by default. Updating the PR.

I think so. But it'd be good to give it a spin in your project(s) too of course.

We can do pre-release for a while, sure. Maybe a week or so. But then the big benefit comes from shipping this to stable.

Another observation - it's not related to the new pattern match variant type spread functionality. The same code but without the variant type spread also has the same behavior: https://rescript-lang.org/try?version=v12.0.0-alpha.10&module=esmodule&code=C4TwDgpgBAhgxsAlgewHYBVwQIxQLwBQUUAPlAOqLAAWyArsAAowgA2yMAJtkaRVdWZsO3ABQBvAB4AuKIlTAAvgEoCBUJFgIUGLACZ8vMpRr0mLdlz1H+NIZc56JIWfKWrixgfZEBmZ64KKmoa0PBIaJiQvoaetoIWIgAsEjJyQR58JgnCXACsAenuNtk+XABsqYHFcaWJXADshW7Btd71nAAcVUXB6lhaEaixfAB04+E6UTg246OTkfqzE9qL0WqsEMBQAGb4UKILqLJHyvgAfFDivADOAO5UcNSDOlcl7bmcKQD6mV52HQKv3eAM+lWBbVBDiaEKyHwc3V+F14xAAwmgbshNqN2ABzUQAImoEFY7AJfyg3wuB0yil4SWsdLUBCAA...

To summarize this (and please @cristianoc say if you think differently): 1. The 2 different code snippets produced are functionally equivalent, so there's no _bug_ in the sense that something...

@DZakh thanks for the feedback! The current object is bloated, I agree, but I don't see where runtime performance would be an issue here...? Could you expand on that part?...

So, how about this: * We select a few of these constants we want to make possible to inject. `__VALUE_PATH__` and `__POS__` sounds reasonable to me, that should cover most...

I made the change I talked about above, and updated the example accordingly. What do you think @cknitt @DZakh ?