Show HN: Hitcount.me – Live Website Popularity | Hacker News Reader

Show HN: Hitcount.me – Live Website Popularity

benwilber.github.io

67 pointsbenwilber010 comments

Cool project! Some suggestions:

1) Add rel="nofollow" to your links on https://hitcount.me ... or say hello to SEO spammers!

2) http://hitcount.me isn't forwarding to https://hitcount.me properly. I get the default nginx page. Don't forget HSTS headers and other friends.

3) If you're going to have keepalive on, make sure your sysctl settings are tuned to handle it! Having worked on pixel servers for an ad network, I recommend sending a "Connection: close" but enabling SSL session caching/resumption. (see #4) This doesn't really matter that much until you become popular.

4) SSL could use some tweaks. Initial handshake is quite slow. Caching/resumption should be on, too. https://www.ssllabs.com/ssltest/analyze.html?d=hitcount.me

Also, looks like you're triggering a pattern in uBlock. Not sure about other adblocker plugins. I guess "hitcount" has been used by other networks/stats systems in the past: http://i.imgur.com/C47gHwD.png

Hey thanks for the tips! Yeah I need to tune ssl and a few other things. I fixed the https redirect and nofollow, thanks!

Looks like you're getting spammed now by that OHFB site. :-(

Technically uBlock is working perfectly.

I’m sure you can figure out ways to break it, but it’s a waste of your time (and you need to get a life).

If it's not worth gaming then why waste time making it in to a game by producing a leaderboard? If your answer to making a leaderboard is 'because it's fun' or 'to learn something', then those reasons apply equally well to working out a reason to game it so you're at the top.

Yeah that was meant mostly tongue-in-cheek. There are plenty of ways to break it, I just didn't imagine that anyone would bother.

Just hitcount doesn't seem to be a great measure of quality. What I think would be more interesting would take into consideration time spent on page. That could really help kill clickbait that has nothing to do with the actual content.

For example, if I could find where people are spending more than 2-3 minutes on a page, those are probably much more valuable reads than just hits.

This is really cool and I appreciate the ease of implementation. I feel that the best content on Reddit, HN, and the like is often not intentionally seeking traffic but is rather unexpectedly shared. Unfortunately, that content is often hidden amongst articles that are specifically posted on those sites to increase traffic. Since you have to implicitly embed the Hitcount.me .gif file on your website, Hitcount.me could act as an aggregator for finding the best traffic-seeking content, while hopefully leaving Reddit and HN to power the community-driven content and hidden gems.

does it pass the `ab -n 100000` test?

One more rule added to my filter. Thanks.

Reply on news.ycombinator.com