Another way tracked links will break (hard) if you don’t consider %-encoding
There are so, so many ways to break links. I’ve seen more encoding disasters than most, but this one was new to me. →
There are so, so many ways to break links. I’ve seen more encoding disasters than most, but this one was new to me. →
More dynamic date magic that can be done simply using Velocity and Java (after a little head-scratching). →
A subtle mistake in a link sent to 100,000 people was about to wreak havoc on attribution. Here’s how we avoided that with an emergency JS intervention. →
Chances are your site is only usable if multiple 3rd-party services are up and running — part of the cost/benefit of not building stuff in-house. But you definitely don’t want to rely on seemingly developer-only services being up! →
If one group is making guesses about another group’s datatypes, your project wasn’t ready for go-live. →
There’s no One Best Way™ to store complex data. To find the best fit, consider what — and who — will be doing the encoding and decoding. →
> Update 2022-03-18: The prize goes to Shivam, who DM’d me the correct answer: 31 bytes. I’ll explain in the next post! Another encoding-related quiz, this time with →
Last night we heard from a user that 90%+ of their Marketo emails were flagged as spam. They’d only made a minor DNS change — or rather, what they thought →