> It's also worth mentioning that Roy Fielding, one of the principal authors of the HTTP specification (along with Burners-Lee), characterized Code-on-Demand (t

> It's also worth mentioning that Roy Fielding, one of the principal authors of the HTTP specification (along with Burners-Lee), characterized Code-on-Demand (that is, scripting/VMs, e.g. Java and JavaScript) as a core constraint of Representational State Transfer (REST) [2]. The REST architectural style, as described in Fielding's thesis published in 2000, is intended to characterize the architecture of the Web. Not exactly. REST is only semantic addressing. It becomes an architectural principle when that is the primary basis for the organization of content. That said it is additive to the organizational concept of hyperlinks, but it does not replace or disqualify the original intention of content distribution. > Burners-Lee goes on to sketch out what such a system might look like, and characterizes the current browser/server architecture, emphasizing the separation between display logic and server logic, and anticipating the evolution of those display programs: That is how we came to something like CSS. I remember developing for the web in the 90s and browsers completely abandoned this concept. This concept of separation didn't become a practical reality (cross-browser) until about 2006 and then was still challenging in those (IE6) days. > Lastly, an excerpt from an interview with a Wired interview of Burners-Lee from 2014, characterizing JavaScript as part of the web's "fabric" and emphasizing its ability to evolve [3]: Yes, the current web cannot function without something like JavaScript. That is the web adapting to the technology opposed to the technology adapting to the web.