How This Website Works

We’re often asked how the website is put together and where it “lives”, so here’s an explanation — starting in plain English, with more technical detail further down for anyone who wants it.

In short

This website is a set of simple web pages, automatically assembled from plain text files and published to the internet using GitHub, a well-known, free hosting service used by millions of software and website projects worldwide. There’s no traditional database and no server for a committee member to maintain — updates are made through an easy, browser-based editing tool, and the live site rebuilds itself automatically within a couple of minutes of any change being saved. The photo gallery, archive of past documents, news and events are all part of the same site, built the same way — nothing is hosted separately or bolted on from elsewhere.

One of the charity’s volunteers maintains a backup copy of the site on their own equipment, in the unlikely event of loss of this site.

How updates get published

1. Editing — Committee members and other volunteers who have been given authorised access make changes using a straightforward, browser-based content editor. No coding knowledge or special software is needed.
2. Review — Each change goes through a short review step before it’s published, and is recorded as a discrete, timestamped entry in a version-control system that keeps a complete history of every edit ever made — so nothing is ever permanently lost, and any change can be reviewed or reversed.
3. Building — Publishing a change automatically triggers a build process that turns the underlying text files, and any photos or documents attached to them, into the finished web pages you see.
4. Publishing — The finished pages are automatically published to the live site, typically within a minute or two. The site also rebuilds itself automatically every night, so that dates — “upcoming” versus “past” events, for instance — are always current even if nobody has made any edits.

Where things live

Keeping the site resilient

The live site depends on one hosting provider, which is reliable but not something the charity controls directly.

One of the charity’s volunteers separately keeps a mirror of the site’s source and edit history on Forgejo, an open-source, self-hosted alternative to GitHub, along with a standby copy of the built website, both on their own equipment. If it were ever needed, this could give a head start on bringing a working copy of the site back online elsewhere.

Who can make changes

Access to edit the website is restricted to committee members and other volunteers who have each been individually authorised, with their own login approved by the organisation’s account owners. Access to the underlying technical infrastructure (hosting accounts, source repository, DNS) is limited to the IT volunteer role.

Accessibility

The visual design was chosen carefully with accessibility in mind, including for visitors who use screen readers or other vision-support technology, and every photograph on the site has “alt text” describing what it shows. A search result that points to a specific gallery photo moves keyboard/screen-reader focus straight to that photo, so it’s clear which one matched. Full details, including where we currently fall short, are on our Accessibility Statement page.

Privacy by design

No contact details appear anywhere on the site as plain text — enquiries go through the site’s own contact forms instead, to discourage automated harvesting. The site uses no cookies, no visitor tracking or analytics, and no third-party scripts or content-delivery networks beyond the hosting platform itself. What happens to the data you do send us, through those forms, is covered on our Privacy & Cookies page.

Data and hosting location

No confidential or sensitive personal information is published on this public website. Most of the hosting and automation services we use — GitHub chief among them — are, like the overwhelming majority of services used by UK charities, businesses and public bodies, run by companies with a base of operations in the United States. The service behind our two forms is the exception: it’s a small server we run ourselves, hosted in the UK, which sends submissions on by email through Fastmail — the same provider that already runs the charity’s other email and this site’s domain, not a new company. Either way, this means the data involved could, in principle, be subject to US as well as UK/EU legal frameworks, in addition to wherever the physical servers happen to sit. Because everything on this site is intended to be public, this has no practical effect on what’s published here. Our approach to genuinely confidential material — committee papers, correspondence, and similar — is handled through separate arrangements, not covered by this page.

For the technically curious

• Static site generator: Hugo (extended edition) — free, open-source software, version pinned for consistent builds
• Theme: Hextra (MIT-licensed), a vendored, version-pinned copy with our own overrides layered on top rather than edited in place — chosen partly for a clean accessibility audit result (zero WCAG 2.x A/AA violations found in testing) and a small code footprint
• Hosting: GitHub Pages — free to use
• CI/CD: GitHub Actions — free to use (within its included allowance); builds on every content push, plus a nightly scheduled rebuild so date-dependent content (events, news) stays accurate
• Content management: Sveltia CMS — free, open-source software; a Git-backed headless CMS working through an editorial review step — content edits are committed directly to the repository, with no separate database layer
• Version control: Git — free, open-source, and an open standard in its own right; hosted on GitHub. The repository’s visibility (public or private) may vary over time — currently public, since GitHub Pages on the free tier requires this — but the workflow described here is unaffected either way. A published site stays public regardless of whether its source repository is public or private
• Forms: a small relay we wrote and host ourselves, replacing a third-party service we used at first — plain Python, using only its own standard library, so nothing is pulled in from anyone else’s code and there’s nothing for a dependency-vulnerability scanner to ever need to flag. It runs on a small free-tier server we host in the UK, with Caddy (MIT-licensed, open source) handling HTTPS in front of it, and forwards each submission on by email through Fastmail rather than storing it. It’s a small, self-contained piece of the site — just two forms — so it could be changed again without touching anything else
• Map: an OpenStreetMap embed, with no API key, account or cookies required
• Photo gallery: a section of the same Hugo site, with a lightbox viewer via PhotoSwipe (MIT-licensed, vendored locally rather than loaded from elsewhere)
• Search: a site-wide search box (press / or Ctrl+K) using FlexSearch (Apache-2.0, vendored locally rather than loaded from elsewhere) to search page content entirely in the visitor’s own browser, with nothing sent anywhere
• Images: automatically resized into multiple resolutions and converted to modern formats at build time, so visitors never download a full-size original unnecessarily
• DNS: the domain’s DNS records point to GitHub Pages; these are standard, publicly resolvable records that can be checked independently with tools such as dig, nslookup, or any online DNS lookup service
• TLS/HTTPS: certificates are issued and renewed automatically by the hosting platform
• Backup: one of the charity’s volunteers keeps a mirror of the repository on Forgejo — a fully open-source, self-hosted Git platform comparable to GitHub — plus a standby build of the site, both on their own infrastructure

Questions

If you have questions about how the site works, or would like to get involved in maintaining it, please get in touch via our contact page.