Meta Sites and Pages

My 1994 web presence was hosted by first httpd and soon after the legendary apache. Wiki was about 300 lines of perl registered as CGI, short for common gateway interface. wikipedia

I wrote many dozens of cgi scripts in the following years and visit a small handful of them almost daily. As a first experiment with modularity in deno I suggested we recreate that experience if not the actual interface. The question became, what api do we expect of our server when we want to compute up some pages.

We started by inheriting a handful of helper functions that would emit pages and page parts.

This moved to an abstraction that would be passed in when a rendering was requested.

The abstraction became a module named "wiki".

The server could discover metasites in the filesystem just as our legacy farm mode discovers virtual sites.

Our experimental scraper was the first metasite to migrate to its own repo. See Region Scraping and Indexing

Our definition of what is a metasite and how they are coded remains fluid. This could still be the most important, or the least important, of the new thinking we are doing.

We've built several and deployed one production service for Content Transformation.

Title Space in File Browsing sounded good but broke our notion of Titles that can be reduced to slugs.