Reading about the Fresh framework lead me to an article about islands architecture and progressive hydration. “Island based client hydration” looks like a type of progressive enhancement.

Wonder how an adjustment to that theory pans out? Might be enough of a golden hammer for most use cases but imo, scaling/speed issues are more fundamental/situational. You could swap between or hybridize client/server all you want and still mess up.

Hugo has template performance metrics with the --templateMetrics flag. Here’s a snippet of the metrics for this site.

$ hugo --templateMetrics

Start building sites … 
hugo v0.101.0+extended linux/amd64 BuildDate=unknown

Template Metrics:

 cumulative       average       maximum         
   duration      duration      duration  count  template
 ----------      --------      --------  -----  --------
7.278303484s   49.177726ms  237.175705ms    148  partials/generate-feeds.html
7.22797312s   112.93708ms  246.988674ms     64  _default/single.html
6.982028147s   83.119382ms  245.262274ms     84  partials/navigator-right.html
6.683942678s   47.070018ms  237.223569ms    142  partials/web-ring.html
1.685089741s    5.561352ms   28.254398ms    303  _default/summary.html

The static versus dynamic site wars are mostly conflicts over where discomfort and computation should happen at increasing scale. You’ll either be lagging/computing on the server side, the client or, maybe somewhere in between.

Users are psychologically primed to hate vertical scrollbars, but what happens when you remove them? Bring in the horizontal “indicators” duh… Oh you designers. ‘Tis probably wise for browsers to make that a user option and default to vertical scrollbars before scroll accessibility goes up in flames.