Marcel van Lohuizen

We are thinking about whether we should bump the priority of the implementation of `numdefined`, which will simplify working with "oneofs" patterns like this a whole lot simpler as well...

Upon closer inspection, this seems more like a disjunction problem than a cycle problem per se.

Cycle errors (not to be confused with structural cycles) are a kind of incomplete error. So this is correct. The spec mentions that cycles can be broken. So this is...

I can confirm this is fixed in the cycle/comprehension rework I have locally in my client.

Almost done with the cycle rework. And now I noticed it is _not_ fixed. Not sure why it was working before, but upon further investigation, this is actually working as...

Note: this is marked as comprehension as it has overlapping logic with the implementation of comprehensions and this will ensure this is validated against the upcoming changes for this.

> About not finding list concatenation in the spec, I instead found #1677 , which suggest removing list concatenation using the + operator, and mentions that it is already removed...

This has indeed been fixed by the recent changes. It may be underreporting error positions. A todo is still outstanding for this.

Analysis: it seems that when there is a "cross product" of a disjunction with itself, the cycle detection allows progressing one level deeper, which has the effect of causing quite...

This will fix the performance issue by detecting cycles on the earliest possible point, preventing the combinatorial computation in the first place: https://review.gerrithub.io/c/cue-lang/cue/+/543675/1