I think that it greatly depends on the company and its needs. If you think that there is a great chance that someone will simply ape your code and use it to com

I think that it greatly depends on the company and its needs. If you think that there is a great chance that someone will simply ape your code and use it to compete directly with you, it might be wise to hold back. If, on the other hand, you a likelihood of people getting involved and helping to improve the code, that might outweigh the risks. I'm not so worried about the former seeing as you're already competing against some very powerful players, but that's just me. It really depends on whether you think that your algorithms important enough to keep secret, or whether you thing that an open source search engine would really add something.