Noam Postavsky writes: > Oh, yes, I was confused by your docstring. By "if no match found" you > meant when the line doesn't match ^foo^bar^ at all; I had somehow got > the impression you meant that there was no match for "foo". Ah, yes, I'll try to make the docstring a bit more clear! > This eq test will always be nil, right? Because the only time it's t > is when you pass something that's not a history reference, but the thing > we passed is a history reference by construction. So this could be > simplified to Yup, that's true, I'll simplify that down. Here's a patch which tries to fix those issues. Thanks again, -Jay