#942 Overhaul fedora-websites generation and hosting
Opened 3 years ago by ryanlerch. Modified 7 months ago

Currently, we use a bunch of scripts and Make to generate the websites. Furthermore, we use genshi tempates as the basis of the sites.

This system is quite old, and has some challenges. From a template / theming point of view, there are many times where templates contain duplicated code, and other issues that make it hard for community members to jump into helping out on the websites.

The idea of this ticket is to facilitate discussion on issues of the current process, and progress through to thinking up solutions.

Ultimately, hopefully, we can use this ticket as a tracking ticket for tasks as we progress.


I'd love to see us move to one of the templating/site building system (hugo, jekyll, etc.) It would make it easier to maintain sites like flocktofedora.org amongst others. I have no real preference for the system used as long as it supports the theming requirements of design to ease delivery and can be translated (even if done via po4a or the like).

Using something like Hugo for the sites that remain in this repo would be amazing.

I agree that using a static site generator would be a great approach for these sites.

Personally I've been using eleventy and nuxt a lot. Eleventy is somewhat limited featurewise but is great for sticking to a vanilla JS approach. I've really liked nuxt because it gives the ability to build SSG, SSR, and PWA and it's component system is excellent. However it would pose a bit of a learning curve for familiarity with vue syntax for some things, but the documentation is really good and it is pretty easy for new devs to get into.

I'm hesitant to be supportive of using Jekyll or Hugo specifically because they are Ruby and Go based so we would need front end devs that are experienced in those languages.

I'm not sure if it's related to this ticket but in current initiative of main Fedora Website Revamp we decided that in first step because of limited time and resource we won't change CMS / tools we already use unless new design will require this.
+1 to above ideas however we are not able to implement them in first approach for Fedora website which also will give directions for other smaller projects.

I believe we were focusing on the question of whether or not to start applying a frontend framework to our websites.

Regarding the change of CMS, it would be great to adopt a headless CMS as it would make editing content easier, reduce barriers to participation from new contributors, and make it easier to change our frontend much easier.

While I agree that it's too early in the process to make a shift like this, I also do think that it's something worth planning and preparing for so that our websites can also reflect the innovative spirit of our operating system (and so our web devs can take advantage of the benefits of newer web technologies).

Login to comment on this ticket.

Metadata