In 2017, Parimal Satyal took one newspaper article apart and found a useful summary of what had gone wrong with the modern Web: the words were no longer the expensive part.
Read Against an Increasingly User-Hostile Web
Satyal’s essay, Against an Increasingly User-Hostile Web, was published in November 2017. It is partly a history of why he fell in love with the Web, partly a privacy argument, and partly an audit of what web developers had started doing to ordinary readers.
The personal history matters. Satyal writes about growing up in Nepal and making his first website at age eleven. The older Web he describes was not perfect, but its basic machinery was unusually generous: somebody could publish a document, link it to somebody else’s document, and make the result available without asking a platform for permission.
Then he measures the newer version.
For one Le Monde article, Satyal recorded a page weight of about 3.1 MB and 459 HTTP requests. Of those requests, 436 went to third parties. The page contacted 118 third parties and stored 100 cookies. He then rebuilt increasingly stripped-down versions of the same article. The text-only version was 17 KB, made one request, contacted no third parties, and set no cookies.
That comparison is still nasty because it does not require much interpretation. The article itself had become a small passenger inside a much larger tracking and advertising machine.
A webmaster essay that aged annoyingly well
Satyal was not arguing that websites should freeze themselves in 1996. He explicitly acknowledged faster networks, better standards, stronger security, richer media, and the practical reasons modern sites became more complicated. His complaint was about whose interests that complexity served.
The latter part of the essay gets practical. He asks web professionals to question automatic third-party requests, unnecessary social widgets, oversized pages, dependence on JavaScript, and whether a site remains usable on a slow connection. HTTPS, accessibility, open tools, and direct publishing all fit into the same argument: treat the visitor as a person instead of raw material.
There is another nice old-Web detail here. The essay itself is openly published, linkable, readable without an account, and its source is available for people who want to suggest changes. Satyal later turned it into a Paris Web talk, but the original page still works perfectly well as a document.
CacheRat’s larger 1,967 Ancient Web Domains research list contains hundreds of examples of the sort of independent Web Satyal was defending: strange personal pages, specialist references, hobby projects, and documents that exist because somebody decided to publish them.
The essay is worth rereading because its strongest evidence is not philosophical. It is a browser inspector, a request count, and one brutally small 17 KB HTML file.