I’ve got this automation to get my favorite bookmarks from Pocket to my website. When I’ve got like five good ones, I know it’s time to send another newsletter.
This didn’t happen for a long time. Because the automation was broken.
It’s supposed to work like this:
I bookmark something with Pocket.
I read it and decide it’s so good, I want to share it with you, so I give it a little yellow star.
Every 15 minutes (kind of wasteful, if you ask me, but there’s no better setting for it), Zapier checks if there are any new yellow stars.
If there’s a new one, Zapier adds it to an Airtable.
The rest of the flow involves a JavaScript package that connects my website with Airtable, and some code that shows the bookmarks on my site. But that doesn’t matter, because the Pocket-Airtable connection is broken.
I've been trying over and over again to make my Zapier ‘zap’ copy my favorites from Pocket to Airtable, but it looks like it’s just broken now. The issue could lie with the Pocket API, Zapier or Airtable. None of them show any other indications of things being broken and the test runs work fine.
This is exactly the kind of thing I’m really unsure about that whole no code/low code thing. It sounds so nice, no programming, just clicking here and there to make complex things work.
But my super simple bookmark thing failed. Silently. Weeks ago. I can manually copy some bookmarks, but if my business would run on such a zap, it’d have cost me badly.
Now it’s not like the things I’ve coded myself always work. Especially things like user-facing forms can fail badly and silently. Like that time when I deployed a contact form. I tested it locally, on staging and in production. My coworkers tested it extensively too. Turned out it only worked if the sender’s e-mail address contained our company domain.
But at least there I could do something about it!
Old, but good: the inventor of hero images wrote about hero images.
Bounded rationality: no matter how many slides and spreadsheets are involved, business decisions are eventually based on gut feeling.
For your inspiration:
This one already got enough attention, but it's very good:
I felt sloppy for waiting four years to run a survey on our users to get some proper demographic data. But when we had it, there was nothing actionable in it. I should have listened to Heydon: The Folly of Chasing Demographics.
Not at all design-related, but good exactly because of that: this course teaches the basics of geology. I wanted to know more about rocks (I always wondered why they look differently in different places) and found this MOOC. Rocks are just a tiny bit of it, because really it’s about how the earth works. I know—it’s in the name. But this planet is wild.
Here’s a cool personal website:
I recently finished the Complete CSS course. Which isn’t a complete overview of CSS selectors, properties and values, but instead a complete approach to writing CSS, from talking to designers to creating composable layouts and finished components. Compared to the abundance of free resources on the topic, it looks expensive, but found it was worth every penny:
And unlike when learning how to use Zapier, this stuff will be relevant years, if not decades from now!