Nothing is more annoying than stale documentation of indeterminate vintage.

We’ve all been there - you’re 3 hours into solving some problem, you’re out of ideas, and it’s time to hit Google and see if anyone else has seen anything relevant. The official documentation to the thing you’re using is crap, so that’s no good, but what’s this! Someone wrote a blog post describing what looks like this very issue, and they found a workable solution! Eureka!

You happily skim their procedure, and while it doesn’t look precisely familiar, you think you understand the gist enough to give it a go. A little bit of copy/pasting, a little bit of find/replacing, one last read through for anything that stands out as dangerous and you’re ready to give it a shot.

Unfound method: bumblethwarp() on line 23,323 of main.lang.
ERROR: thrown YouCopiedShitOffTheInternetAgainDidntYouException:
  - lang.exception.fml

What the hell? That article looked legit! Several people even commented on it saying that they tried it, and it worked fine, and thanks man you’re a life saver! Where did you go wrong?

You skim the page again for any explanation, and then you see it. There, hidden there at the top, in gray text on a slightly-lesser-gray background, guiltily hiding next to the tag cloud in hopes it won’t be seen is the publication date. And it says “03/02”. Son of a bitch.

You may never know which decade that article comes from. Was that March 2002? March 2nd? February 3rd? It doesn’t matter; you already dug up the even-crappier documentation from v0. of the library you’ve been banging your head on, and sure enough, that bumblethwarp() thing was a huge controversy back in the Clinton era, and it resulted in a full on rewrite. Which invalidated all the advice that guy’s blog post was attempting to deliver.

You’re back to square one, and you’ll never get that time back.

I get that it’s embarrassing when you haven’t updated your blog in 3 years; I’m surely worse than most. And I get that it might seem like a good idea to spackle over the evidence with an ambiguous publication date, but please don’t. Having out of date documentation on the internet not only wastes everyone’s time, but it can end up perpetuating bad advice long after it stopped being best practice.

Do your part: make sure your published literature has a clearly visible and complete publication timestamp with year, month, and date.