There was a question recently, about why you don’t improve the HTML of HN, where you said "When the HTML is the most important thing to work on."(http://news.ycombinator.com/item?id=4927231)
I agree that the markup is relatively unimportant compared to features that I think could really improve the functionality and quality of the site, and I bet you have a long wishlist of features yourself.
But between the lines I interpreted that that list might be way longer, than the time that is available to you, allows you to work on it.
So why don’t you open source HN? I get that with a project that is important to someone, it’s hard to give away control. But you can still be the project lead, you could still have the last call and I feel to open up the project will lead to great feature discussions and ultimately a better hn.
Have you ever considered open sourcing it? And what’s the thought process on your decision?
Considering this, it's probably not a service that can be "open sourced" as the number of people who can actually work on it appears to be PG. And PG is probably too busy to even act as the project lead.
[0]: It's not so experimental anymore...