Yeah, a forward goto is one that takes you later down the same code block, to some code the program would most likely have hit eventually anyway, just skipping over some redundant work/checks. It doesn't jump into the body of a loop, or backward in such a way that it could produce a new loop. I'll use them occasionally when a loop could terminate in multiple ways, and each way demands a different finishing step to cap off the iteration.

(Though a function call is usually better for that — just gets awkward if the finishing step has to work with and update lots of local variables we'll need later)

Yeah; I've not dared to use it, but it could be used to finish/return from inside multiple scopes.

Exiting scopes is usually easier to reason about than entering or re-entering them via an unexpected route.

In Dijkstra's "Go To Statement Considered Harmful", he mainly argues against cases where a goto confuses the mental coordinate system we use to think about points in a program's execution.

For a goto that takes you forward without jumping into a loop or conditional body, the coordinate of an action in execution is pretty much just its coordinate in the text, maybe with an asterisk to denote some steps we skipped over. So it doesn't add too much extra overhead.

I personally never had a problem with them and dont find them confusing as well. It seems logical from just looking at the code. On the other hand, I still can't wrap my head around shader code

@LudoProf Your link requires Sheridan UPN login credentials. Here's a non walled version that the wikipedia article.

As a bonus, TIL the term snowclone.

Whoops, that'll teach me to share articles I looked up on my work machine. Fixed the link above — thanks for the callout!