#MacOS: Some design missteps in Tahoe can be remedied as I learned from c't today: Smaller corner radii: `defaults write -g NSConvolutionOverride1 -float 10` Wasted space around the floating sidebar: `defaults write -g NSSplitViewItemSidebarDefaultsToFloatingAppearance -bool false` Much better!

A new addition to my “Visual Phenomena and Optical Illusions” collection: “Mainz-Linez”. Here's the live demonstration and explanation of the somewhat mysterious title: michaelbach.de/ot/mot-Mainz.... Enjoy!

michaelbach.de/ot/aud-Shepa... My demo of the ever rising scale, invented by Roger Shepard, had failed due to new security restrictions in browsers. I rewrote it and added plot showing amplitude vs. log(frequency), hopefully making it easier to understand this amazing effect. Enjoy! #illusion

Bluetorial-A dream and a bit of a nightmare Serving as Editor-in-Chief at Science was fascinating. I greatly enjoyed working with talented and committed editorial, news, graphics, and production staff. But the inside look into scientific publishing and AAAS was also deeply disillusioning.

a cartoon says hey everybody an old man 's talking while bart simpson looks on

ALT: a cartoon says hey everybody an old man 's talking while bart simpson looks on

media.tenor.com

Quitting the Washington Post: anntelnaes.substack.com/p/why-im-qui.... Impressive, my admiration! Good read. Hope not harbinger of worse things to come. I learned of it via daringfireball.net.

Bicycling through the nearby woods, I happenchanced upon this frosty gem.

How are people that used to be evidently smart so bad at understanding obvious truths that they actively destroyed the value of their core product?

melomys@melomys.bsky.social · 2y ago

Not that Google was ever good, but this cartoon is a good summary of now.

Open the present by shaking the tree? Or is this Schrodinger's gift?

🧪In our journal club: Just had a paper where the authors needlessly made their figures rectangular, although the content was quadratic (same range on x- and y-axis). It really should be quadratic to understand the content better, also the identity line is then at 45°. Old gripe of mine, sorry 😎.

A new paper on FrACT michaelbach.de/fract/: “Freiburg Vision Test (FrACT): Optimal number of trials?” (open access) dx.doi.org/10.1007/s004.... Mysterious “kink” at 18 trials (figure): 18 trials a good compromise between fatigue and precision for Sloan letters or 8-alternative Landolt-Cs.

I just finished an experiment: For three weeks I responded to every "invitation" from a predatory journal with "STOP PESTERING ME". Results: I now receive about twice as many per day than before… Time to end the experiment and conclude that I'm just validating my email for them 😎. #science