Thanks for the feedback. I figure it can’t hurt to experiment. Every time the machine starts coughing and sputtering, I’ll make a couple passes with used descaling solution. Guess I should keep track of how many iterations I use it. When it fails to work, I’ll make a fresh batch. I might experiment with doing the 1st pass with used solution and a 2nd pass with new (or newer) solution.
BTW, that Cloudflare link you gave is a booby-trap of dark patterns. It gave a blocking cookie popup with no “reject all” option to Tor Browser and an endless list of cookie switches to toggle, so I tried lynx. It gave a 403 claiming “enable javascript and cookies to continue” to Lynx. Then I loaded the archive version in Firefox with js and animations both disabled, and I could finally read the text (very useful and far deeper than I can comprehend). But then an animation at the bottom played anyway. I had to disable still images to stop the animation (guessing the ad is an animated GIF).
What a disasterous display of web enshitification.. anyway, I hope this helps someone.
One of the interesting points that was made is different machines have different components, so it’s best to use the manufacturer recommended descaling solution. Which I assume would also be commercially biased. I hope one day the #rightToRepair will require a disclosure of the materials in the machine and chemical requirements (as opposed to brand requirements).
update
That page inspired me to create a new community: “Asshole Design: Web Edition” :).
I just have to say that must be the shittiest font I’ve ever seen used for academic research. I might still read it since it’s only a few pages!