It certainly seems so. For example:
(:rolleyes:) produces a malformed token (:rolleyes with a grin smilie at the end. Given the left-associative nature of English (we read left-to-right), this is the Wrong Thing. My example should produce a rolleyes smilie encased in parenthesis, as (: is not a valid token.
Is this universally true given all combinations of smilies? Is there an easy way to fix this?
It’s not that it’s right-associative, it’s just that any combination of characters in your post that can produce a smiley will produce a smiley. The lesson to be learned from your example, I think, is that the ascii-graphic type smileys take precedence over the smileys with codes that are words, and the vB software checks for them first.
Test: (:rolleyes:)
Test: (:rolleyes:**)
Two ways that I can tell (tested on IE 5.5 in Windows 95), neither of which is particularly easy:
(:rolleyes:)
(:rolleyes:**)
a) Don’t know.
b) Not without modifying the source code, which we don’t do.
:rolleyes:rolleyes:
Looks like they are left-associative, but 
has higher presidence than :rolleyes:.
I think this has been done before, and a precedence was worked out. Probably goes through the text searching for one token at a time.
OK. Left-associative, but with precedence rules.
Interesting enough to keep a geek happy.
Thanks.
Test zone:
:wally:rolleyes: