The degree to which desktop load times are stable over 10 years is in itself interesting and deserves more curiosity than just saying "javascript bad"
Plausible alternate hypotheses to consider for why little improvement:
* Perhaps this is evidence for control theory at work, ie website operators are actively trading responsiveness for functionality and development speed, converging on a stable local maximum?
* Perhaps load times are primarily determined by something other than raw bandwidth (e.g. latency, which has not improved as much)?
* Perhaps this is more measuring the stability of the test environment than a fact about the wider world?
If this list of changes is accurate, that last point is probably a significant factor -- note that e.g. there's no mention of increased desktop bandwidth since 2013.
https://httparchive.org/reports/loading-speed?start=earliest...
The degree to which desktop load times are stable over 10 years is in itself interesting and deserves more curiosity than just saying "javascript bad"
Plausible alternate hypotheses to consider for why little improvement:
* Perhaps this is evidence for control theory at work, ie website operators are actively trading responsiveness for functionality and development speed, converging on a stable local maximum?
* Perhaps load times are primarily determined by something other than raw bandwidth (e.g. latency, which has not improved as much)?
* Perhaps this is more measuring the stability of the test environment than a fact about the wider world?
https://httparchive.org/faq#what-changes-have-been-made-to-t...
If this list of changes is accurate, that last point is probably a significant factor -- note that e.g. there's no mention of increased desktop bandwidth since 2013.