Local editor discovered 1,380 lost subheadings in ancient Signpost scrolls. And what he found was shocking.
Yeah, you got clickbaited. Anyway, here's the deal:
Recently, I wrote and deployed an argosy of scripts (covered in more detail here) to extract 1,380 lost subheadings from the revision history of the Signpost's main page. These are now in their respective articles' header templates (and from there, in the module indices that serve as an article database). While this allows for much broader flexibility in our display methods, that isn't very exciting (or at least not until these display methods are actually put into practice). What is exciting — or at least mildly amusing — is a whirlwind tour of the never-before-seen Signpost greatest-hits compilation.
Basically, the subheadings were introduced in July 2012, as part of the perpetual effort to keep the Signpost modern and bumpin' — they started out as simple excerpts from articles that were shown on the main page. In 2017, they started being incorporated by default into the RSS-description templates — these are invisible, they don't display anywhere on the article page, but they provide metadata in the HTML — and began to assume their current form (brief, couple-sentence-long hooks). Well, I went through and put all of them into RSS-description templates, so now there's 2,464 articles with machine-readable subheadings, out of 5,462 Signpost articles in total — i.e. there are precisely 1,998 articles from before July 2012 that just never had headlines in the first place. Well, whatevs.
Some were missing, some were messed-up, some were typos — honorable mentions to the 2018-02-20 humor column ("headline?") and the 2017-12-18 blog feature ("."). Among the rest, a few extremes stood out, which I had nothing better to do than put in tables for my own amusement — and maybe, dear reader, yours as well.