I am currently working on an offline version you can download and run yourself, probably going to be called searchcode server with a pricing model similar to octopus deploy. You can register your interest here https://searchcode.com/product/
Well, but couldn't you just don't offer any support? If you would publish it on Github, you could just 'unwatch' it and maybe let the community support itself. Of course, bug reports and contributions are nice, so you could have a few trusted colloborateurs, who forward those few useful 'issues' to you
EDIT: I meant my comment more in the way of "Do you actually have an obligation to support your free software?". It's understandable, that, with open-sourcing code, you will get a lot of support load, but it will only affect you, if you actually plan to respond to them (That's why I proposed 'unwatch'ing it. It may seem stupid, but I thought this could be a way out for OP. Just code-dump the project, add a commit here and then, but never try to offer support/respond to questions etc.
But even then, I'm perfectly fine with not open-sourcing it. It's still an awesome and useful tool. Great work OP!
I am going to assume you are referring to why its not free software.
Yes I could. I can guarantee I would still get a deluge of email asking about how to set things up. The amount I deal with from my Decoding Captcha's article is more than I can bear.
However it is something I may do in the future. Certainly if I ever shut it down I will release all source.
I can't believe the first comment here is making a problem out of it not being free. Sadly the vocal majority of HN will berat you for not making all your code open source and the product free; but don't think for a second they know anything about what they're talking about. As a sole developer you have to support yourself somehow. This project is full of high-quality features so good job! Your effort is to be commended.
Unable to edit so adding a new comment. Yes that's possible but I think the result would be the same. I freely publish my details around so I think I would still get quite a lot of support requests.
Yes there are a few reasons for that. Mostly cost (it runs on 2 servers) and it can only index things that are publicly exposed. Since the majority of code released is in github it tends to skew towards it as the main source.
It does however index the following (and more) repository locations, Github, Bitbucket, Fedora Project, Google Code, CodePlex, Sourceforge, Tizen Project, Google Android, Minix3, Seek Quarry, Gitorious and a collection of sources from GNU Savannah.
Im always looking to add more, so if you know of say a github for fossil projects (tarpit would be an excellent name for this BTW) let me know so I can add and index them.
It actually does to an extent (its an offline process I need to run again actually).
Have a look at this search https://searchcode.com/?q=jquery+mobile notice that after a bit next to each result the text "Show 100 matches" appears? These are the duplicated identified by the system (limited to 100 results but all dupes should be gone).
The matching is a little fuzzy as well to try and clear out minor revisions of the same file or simple whitespace replacements etc...
Why it is not 100% free software http://www.boyter.org/2014/10/searchcode-com-100-free-softwa...
My responses to the previous discussion on HN http://www.boyter.org/2014/07/feedback-loop/ and the last discussion https://news.ycombinator.com/item?id=7947075
I am currently working on an offline version you can download and run yourself, probably going to be called searchcode server with a pricing model similar to octopus deploy. You can register your interest here https://searchcode.com/product/
Some information about how searchcode works (Django MySQL Sphinx Nginx Celery), http://www.boyter.org/2014/06/searchcode/ and http://www.boyter.org/2014/06/sphinx-searchcode/
You can read more about its development on my personal blog http://www.boyter.org/category/searchcode/
Feel free to ask any questions or provide feedback.