3 ms·
Well, yeah. But because most people don't really check /new on a regular basis, if you're on /new when a lot of people are on it's more likely somebody will see
by qwertyuiop924 10y ago
Well, yeah. But because most people don't really check /new on a regular basis, if you're on /new when a lot of people are on it's more likely somebody will see you.
But to get to front, you need at least 9 other people to see your post, and think it isn't junk. Whether or not this happens is controlled by so many variables that it's essentially random: Whether or not your post is good for HN, whether the specific audience (HN carries a few) that would like your post is on at the time, how much they like it, how many people are looking at /new, how many new posts are coming, and so on.
I've gotten a few post to the front, and from personal experience I can say it's hard to know what will make it.
A blog post from Andy Wingo about concurrency and programming language design? Nope.
A talk from Bryan Cantrill at USENIX about how the conference system is broken, and how we can fix it? Nada.
An anecdote on mobile app development? Nothing.
Javascript framework? Forget it.
A garbage collector, an R5RS Scheme->JS compiler, and R7RS Large Red Edition? Zip, Zero, Zilch.
An excerpt from John Carmack's blog about the design of QuakeWorld, Two different Lisp implementations, and a blog post about how CHICKEN Scheme does garbage collection (it does, Cheney on the MTA by the way, which is a pretty cool hack to get cheap continuations, GC, and tail recursion at the same time)? Suddenly, HN cares.
So yeah, it's pretty random. Unless you post an article about an implementation of Lisp. HN really likes Lisp.
- angersock 10y agoOne could observe that the common theme in your list of submissions there is technical content, not politics or valleywagging or entrepreneurship porn.
- qwertyuiop924 10y agoWell yeah. I submit what I come to HN to read. And I don't come to HN to read that.