[r6rs-discuss] [Formal] Requirement to detect circular lists

[Dan Muresan]

> 1.  The claim of "adverse global effects on performance"
> needs to be supported.  Having used and implemented these

I agree that there don't seem to be many instances where this makes a huge
difference. Certainly nothing becomes O(n) instead of O(1). For (length),
for example, there's just an extra check at each step -- no need for
hare-and-tortoise.

> 2.  The claim of circularity bugs being uncommon may be
> true, though there is no evidence of it.  I personally

I have never seen a circularity bug, or heard of anyone who has seen one.
Has anyone here ever had a circularity bug?

> so what?  Making Scheme easy to implement is not a goal
> of R6RS as far as I can tell.

Scary quote. When I first saw the spec size, I had the feeling that making
Scheme *harder* to implement was a goal of R6RS (though some of my
concerns have been subsequently alleviated).

Best,
Dan Muresan
http://alumnus.caltech.edu/~muresan