calebmkim

(Edited to a more tricky example) Trying to implement this, and I'm trying to think of possible edge cases. Consider the following example: ``` component child (in: 32) -> (out:...

> correct me if I'm wrong, but don't continuous assignments have the "magical" property that they are always active, even when the component that contains them isn't running (i.e., its...

After thinking about it, the general `with` statement seems like it is (probably?) the most natural way to solve our problem. Here is a summary of my thinking: 1) We...

# Synchronous Meeting Summary ## Why this problem matters 1) Inlining is a very important optimization, since it is what makes other optimizations possible (e.g., the graphs at beginning of...

After looking through the codebase for a bit, I think a generalized `with` would be doable. From a quick glance, the things that need to change are the following. 1)...

Yep, this all makes sense-- it seems like optimizations to reduce the scope of `with` would help. > If we have with c { x } where x is a...

> The high-level problem is that, in general, we cannot transition from a to b in the same cycle when generating dynamic FSMs because the done condition of b could...

During a synchronous discussion, we were trying to generalize the problem expressed in the original post: in other words, come up with other situations in which we need cycle-precise behavior...

Just took the next open slot

Based on the synchronous discussion today, it seems like we probably do not want to introduce a zero-cycle transition as a guarantee of the language, as it would infect the...