Mark S. Miller
Mark S. Miller
https://github.com/endojs/endo/pull/2356 is a failed experiment to make some progress on this.
> `[x--][0] > y` I like it!
I think the term "completion value" may be causing confusion. In `(x--, x > y)`, the value being compared to `y` is the value of `x` after decrementing. In `x--...
I suspect that a lot of this is due to the old assert pattern. I'd wait and re-measure after we're cleanly switched to the `|| Fail...` assert style. Unless of...
Yes, this change looks good to me.
> @erights To be concrete, I'm talking about, not this PR, but instead a PR to say that there's no stability for NaN (since this is the implementation reality). Thanks...
We discussed my major concern: That the security purpose of addressing your threat model is only well served by a per-realm (in particular, per-compartment) check. However, we did not get...
> Almost nothing using internal slots operates in a membrane-transparent way. Almost all use of internal slots is membrane transparent. See the Date example at https://github.com/tc39/proposal-array-is-template-object/issues/10#issuecomment-524248326 above. > Private fields...
> I'm not in favor of making a habit of piercing through Proxies in cases like this. Neither am I. I prefer not to create the problem in the first...
The problem is that ```js aDate.getFullYear() // is normal usage Date.prototype.getFullYear.call(aState) // is unusual, so breakage less painful Array.isArray(x) // is normal usage, so breakage would have been painful Array.isTemplate(x)...