Jeremy W. Murphy

Fortunately, that was an easy bug to squash: I had forgotten to put the single-character-pattern edge case back in. Thanks again for picking it up.

Essentially. I'm really just waiting for a sign from Marshall that it's an algorithm he would accept before I do any more work on it. On 27 Oct 2016 8:30...

Actually, that does remind me that I started benchmarking for sequences of types larger than char, but every algorithm was worse than std::search, at least on random data. MN was...

Glad to (finally) hear from you, Marshall. Ah, search test 3 and 4, yes, with strings as the search elements. I don't remember exactly why I avoided them last year,...

I do have a list of things at the bottom that need finishing, most essentially of all is the documentation, which is in progress. But more feedback is always appreciated....

Thanks, I just really need to motivate myself to finish the documentation and trim the fat on the search test 5 data, as I suspect that is the only real...

@ZaMaZaN4iK it's OK, I'll call for Marshall's attention when it's finished. The list of caveats/unfinished things is real; I'll tick them off or remove them when they're done or remove...

> Very, very odd. I _just_ got an email about three new commits. But when I come here, I see they were made 16 days ago. Some of the commits...

@Kojoley I think you need to resolve the merge conflicts for this to ever have a hope of being merged.

The IAR compiler for Arm uses Dinkumware and if I disable exceptions, which is admittedly I guess a non-standard extension, then it fails to compile this injection because it does...