HN&&LO is a web page that scrapes the APIs of Hacker News and Lobste.rs and collates those entries that share a URL.
This lets you easily follow the discussion on both sites for these links.
I wrote this tool because I’m much more active on Lobste.rs than on HN, but I would like to keep “tabs” on both sites. I’m especially interested in how the information is disseminated between the sites.
“Slow webapps”
I’ve long liked to grab stuff via a web API, stuff it into a DB, and output a web page based on the data. This project was a first attempt to use a templating engine in Perl, and I’m ashamed to say I’ve taken so long to understand the greatness of this approach.
The data updates hourly, there’s rarely more than a couple of new entries on Lobste.rs per hour.
In fact, this can illustrate the difference in scale between the two sites.
Time period: 5 Jul 2019 11:19 to 8 Aug 2019 23:59 (both times in UTC).
- Number of submissions with URLS on Hacker News: 26,974
- Number of submissions on Lobste.rs: 899
Average number of submissions per hour: 33 for HN, 1 for Lobste.rs.
Once updated, a static web page is generated. This keeps overhead low.
Differences between Hacker News and Lobste.rs
Coincidentally, while I was writing this post, an article was published in the New Yorker about the moderators (all two of them) of Hacker News. This sparked a discussion on HN (obviously) but also on Lobster.rs (see both linked here).
To me, what makes Lobste.rs better is that it’s much smaller (one can easily catch up with a day’s worth of comments), the editing tools are more sophisticated, and the tag system allows one to determine whether a post is interesting or not. HN is much harder to grasp.
I think the tendencies noted in the article are common to both sites - male, youngish software developers with a rationalist bent. But I think Lobste.rs is a bit more self-aware of this fact (or rather, critiques are more visible there). In some ways, Lobste.rs, being a reaction to HN, defines itself culturally both for and against the bigger site.
Source and TODO
If you’re interested in pedestrian Perl code, source is here.