So NASA pushed the James Webb launch again. Word came down today that it's slipping from December 22nd to "no earlier than" December 24th, and this time the excuse is almost funny: a communication cable between the observatory and the Ariane 5 rocket wasn't talking right, and the team down at the launch site in Kourou has to physically reconnect it and re-verify every single connection point before anyone's comfortable putting a $10 billion telescope on top of a rocket.
A cable. After twenty-five years of development, a decade of "launching next year," and something like ten billion dollars spent, the thing that's holding it up this week is a cable that wasn't seated right.
I've been following this telescope since I was in grad school, which tells you something about how long its been in the pipeline. There used to be a running joke in one of my old physics mailing lists that Webb's launch date moved backward roughly one year for every year that passed, like it was on a treadmill. 2011. 2013. 2015. 2018. Every time, some new problem: a torn sunshield during testing, a dropped bolt that rattled around inside the thing and needed weeks to locate, a software glitch during a simulation. At some point you stop being surprised and start almost expecting it. Today's delay barely even registered as news to me until I saw it pop up in three different feeds within an hour, and even then my first thought was "of course."
And yet. This is still, genuinely, the most important piece of scientific hardware most of us will see launched in our lifetimes. Hubble looks at visible and near-UV light. Webb is built almost entirely for infrared, which means it can see through the dust clouds that block Hubble's view, and it can see light that's been redshifted so far by the expansion of the universe that its basically invisible to anything else we've built. The mirror alone is 6.5 meters across, made of eighteen gold-coated beryllium segments that have to fold up like origami to fit inside the rocket fairing and then unfold, one hinge and latch at a time, over about two weeks once its actually in space, a million miles from Earth. There is no repair mission possible out there. Not like Hubble, where astronauts could fly up and swap out a gyroscope. If something goes wrong during that unfolding sequence, that's it. Which is, I think, the real reason NASA keeps pushing the date back a day or two at a time instead of just YOLO-ing it onto the pad. Nobody wants to be the person who rushed the last step.
Anyway. If the cable checks out, current plan is Christmas Eve morning, launching from French Guiana on an Arianespace rocket, which is its own bit of trivia I like: the most American of American space projects is riding to orbit on a European rocket because NASA doesn't have anything in its own fleet with the right combination of payload size and orbit precision at the moment. I'll probably have the livestream open in a background tab whatever day it actually goes, half working and half just watching a countdown clock, the way I did for the Falcon Heavy demo flight back in 2018.
Also, completely unrelated except that its been eating my week: if you're running any Java services with an old log4j library anywhere in your stack, and haven't patched past 2.17 yet, go do that before you do anything else today. I spent about six hours on Saturday going through an old side project's dependency tree that I honestly hadn't looked at since 2019, finding log4j buried three layers deep in a transitive dependency I didn't even know I had. Not the most fun way to spend a weekend. There's been a new patch practically every few days this month as people keep finding fresh angles on the same underlying mess, so "patched" from two weeks ago might not mean patched now. I'm not going to write the whole explainer here because half the internet already has, but if this is the first you're hearing of it, that's your prompt.
Back to the telescope though, because that's the story I actually want to be thinking about this week. There's something nice about a piece of hardware that takes this long and still, somehow, gets built. Most of what we cover on a blog like this moves fast and gets forgotten fast. Webb has been under construction since before some of you reading this were using the internet regularly. It'll be nice to see it actually leave the ground.