"Correlation is not causation" per se, but...
It does make me wonder even more about the impact of companies' tunnel focus on engagement metrics.
Interesting correlation between time spent in apps and user's (self-reported) happiness.
Relatively small scale study. But certainly a perspective I hadn't thought of explicitly before.
The long and short of how to *really* optimize SVGs.
This is a good lesson in performance. I was researching whether doing blurhash in WASM would be faster than the JS version.
This is 3 iterations against 16 images. If I had only measured the first 4 images, or taken the average, then I may have concluded that WASM is way faster. But in fact the JS version starts out slow but then quickly catches up (probably because it gets JITed). So there's really very little point in using WASM for this. https://github.com/nolanlawson/pinafore/pull/1781#issuecomment-630562314
Mozilla wrote a balanced and nuanced overview of the dot ORG issue.
In short: We're not out of the woods yet.
"Server-side rendering is not a fallback; client-side rendering is an enhancement."
I would further and say client-side rendering can be (not "is"), a performance optimization for subsequent page loads (and offline). It's only a net-win if you invest non-trivial effort (SW, streams, race network, handle long-lived tabs and cache misses to old JS, etc).
In almost all other cases, it merely produces a slower, less usable, less available version of itself.
Think twice before disabling `user-select` on (part of) a web page.
#a11y tip via Remy Sharp (🐦@rem)
"Measuring the performance of Wikipedia visitors’ devices" by Gilles Dubuc https://techblog.wikimedia.org/2020/05/07/measuring-the-performance-of-wikipedia-visitors-devices/
This is super fascinating: Wikipedia ran a microbenchmark on 0.1% of visitors, and tracked the performance over time. Looks like devices have been getting faster over the past year (or maybe browser updates improved the benchmark time… it's not clear).
WebKit's developer tools measure energy impact!
Great song, brought to you live from the living rooms of Jagger, Richards, Wood and Watts.
"Just in case you need a bit more *wat* in your life..."
– Ross Kirsling
The above post was from January, and in recent weeks several discussions took place on birdsite involving some big companies expressing (or finally making public) similar viewpoints.
My favourite one:
"For the record, at Uber, we're moving many of our microservices to what some call macroservices (or wells-sized services)."
Tech has a tendency to take cosmetic details of a thing that works well, and turn it into a trend. This is then avocated for with the expectation it'll make other things work well.
There no doubt exist principles that help build good software. But many "best practices" are just cosmetic. This is mostly harmless if they go with well-understood (theoretically) orthogonal principles.
"Microservices" not only lacks strongly related principles, it can make matters worse.
"My favorite part of every brand guidelines is the precious Logo Misuse section."
"Imagine if bands started making these!"
What does it take to put together a special issue in a student newspaper? I broke down our award-winning issue from October: https://blog.legoktm.com/2020/04/07/inside-scoop-breaking-down-our-award-winning-special-issue.html
Dutch expat living in London.
The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!