Ask HN: how can we establish prior art to fend off future patents?

As part of building a spreadsheet, we've come up with a pretty neat algorithm to run a recalculation in parallel, taking advantage of multiple CPU cores etc. We're don't like software patents, but have had some experience with the patent process in the past. Looking at the algorithm, our gut feeling is that clever enough to be patentable but not so clever that we don't think other people aren't quite likely to come up with similar ideas should they put their minds to the same problem.

If we don't patent it, our worry is that someone else will independently come up with the idea, patent it, discover us, then sue us. The best defence against this that we can see is to make sure that we can prove that we had the idea before they patented it -- that is, to be able to establish priority. What we're wondering is, what's the best way to do that?

We've no worries about making the algorithm public -- it's pretty clever, but not that clever. But proving the date on which we made it public might be hard.

Any ideas?

[EDITED TO ADD] It occurs to me that there's another question here -- do we need to prove prior art, or just priority? That is, do we need to show that the algorithm was already public and known at the point that someone else tried to patent it, or do we just need to show that we had created it (even if we'd not shared it with anyone).

I guess being able to prove that we'd published the algoritm and thus it was prior art is a subset of the case where we can prove that we had thought up the algorithm, so making it public is safer.