On Sat, Apr 16, 2022 at 3:22 PM Andrey Semashev via Boost
+1. I very much like that the current website can be cloned, with history, and updated locally or online, and then the changes can be merged upstream with PRs. The fact that you're using the usual toolset for this, nothing more than git and your favorite text editor, is a huge plus. Using a separate interface for that, or a database, would be a step in the wrong direction, IMO. Your only database should be git.
Yeah static content would stay in the GitHub repository. But dynamic content such as the list of libraries, the list of boost.org accounts corresponding to library maintainers, the tags for libraries, the index for the live news feed on the front page, the forum data, information for user accounts, the results of every review, the review calendar, and so on - those are going into a database. The website we are building is not a static site it is a social media portal for Boost and C++. It cannot be represented merely as a collection of html and css. It is software (i.e. a web application). Django/Python to be precise. I was hoping to have this discussion in the other mailing list, because I didn't want to start a useless discussion here where everyone offers ideas that contradict one another and nothing gets done, which is typically what happens when big line-item things like new websites are discussed. Thanks