Two things arrive here by pull request: articles and calendar events.
Some lists on the site are not edited here at all, because a community project already maintains them and foojay.io refreshes its copy every day. A fix made in this repo is overwritten at the next sync, so send it to the source instead:
| What | Where it is maintained | Shown on foojay.io at |
|---|---|---|
| Java User Groups (and their meetup calendars) | World-Wide-JUGs/GlobalWWJugs — one small file per JUG | /jugs/, /calendar/ |
| Java Champions | aalmiray/java-champions — java-champions.yml |
/java-champions/ |
| Java conferences, dates and CFPs | javaconferences/javaconferences.github.io — README.md |
/java-conferences/, /calendar/ |
Everything under data/ that these produce is generated. Never edit it by hand.
Everyone who contributes here agrees to follow our Code of Conduct: communicate in a friendly way, respect other opinions, and keep feedback constructive.
The full, up-to-date guide for writing and submitting an article lives on the site itself:
→ How to submit your next article on foojay.io
It walks through creating your author profile, preparing your post folder, adding images, and delivering it (pull request, fork, or zip).
-
Posts are contributed via pull request (fork the repo if you don't have write access). You don't have to clone the whole repository (about 2 GB with history): see Submitting without a full clone.
-
Create your post in a folder under
draft/<your-slug>/with the text inindex.md(orindex.adoc) and its images in the same folder. Copytemplate/post.md(ortemplate/post.adoc) as your starting point. A maintainer moves the folder intocontent/posts/<year>/<month>/<day>/<your-slug>/when it's published — you never need to pick that path yourself unless you want the post to be published on a specific date. In that case, clearly specify the preferred date in your pull request. -
Markdown or AsciiDoc. Name the file
index.adocand write AsciiDoc instead, starting fromtemplate/post.adoc. Everything else is the same — same folder shape, same frontmatter, same checks. Markdown is shorter for ordinary prose and what most articles use; AsciiDoc is worth it for captions on code blocks, numbered callouts inside a snippet, admonitions, and tables that span cells. -
Only the article and its images go in the PR:
index.mdorindex.adoc, plus images. Example code, scripts and project files don't belong in this repository: publish them in a repository of your own and link to it from the article. Videos too: host them on YouTube or another external service and embed or link them ({{< youtube VIDEO_ID >}}), so the repository stays small. -
Company-based and commercial articles are welcome if they are Java/OpenJDK/JVM-oriented — they need real, working Java example code or developer tooling, not just a product description. Marketing-only submissions are declined.
-
New author? Add yourself as
content/authors/<your-slug>/_index.mdin the same PR (note the underscore — seetemplate/author.md). An author is a real person: a real photo of yourself as the avatar (no logos) and at least one link we can verify you by (GitHub, LinkedIn, Bluesky, Mastodon, website). Company or team profiles are not accepted. -
Leave
date:as it is in the template. A maintainer sets it and moves your folder to match when the article is published. Only set it yourself if you need the article out on a particular day: it's a plain day (date: "2026-01-01"), with no time — the site is rebuilt once a day, at 07:00 UTC, so that is when articles go out and a time cannot make it earlier, only later, by missing that build. A future date schedules the post (it stays unbuilt and unlisted until then, and appears in "Coming soon" on the home page). -
You don't need to build anything: the PR check validates your frontmatter and builds the site, so it catches a bad author slug or a missing image for you.
-
If you do want to preview it, run
hugo serverand openhttp://localhost:1313/website/today/<your-slug>/.draft/is deliberately not built, so the post has to be somewhere Hugo looks. Either:- copy your folder to
content/posts/<year>/<month>/<day>/<your-slug>/, with today's date (or an earlier one) in both the path anddate:— a future date is a scheduled post and is not built; or - keep a future
date:and start the server withhugo server --buildFuture.
Don't commit the copy: a maintainer does the move. Previewing an AsciiDoc post locally also needs the
asciidoctorbinary (gem install asciidoctor); without it Hugo stops withfailed to execute binary "asciidoctor". Nothing extra is needed for Markdown, and the PR check installs it either way. - copy your folder to
See the guide linked above for the details.
The repository is around 2 GB with history, and an article only adds a few files. Two lighter routes, both ending in a pull request from your own fork:
In the browser
- Open the
draftfolder on github.com/foojayio/website and choose Add file → Create new file. GitHub offers to fork the repository for you. - Name the file
your-slug/index.md(the/creates the folder) and paste your article in, starting fromtemplate/post.md. - Commit changes… → Propose changes → Create pull request.
- Add images by opening your branch in your fork, going to
draft/your-slug/and choosing Add file → Upload files. They join the same pull request. A new author profile goes in the same way, undercontent/authors/<your-slug>/.
If you already have an old fork, use Sync fork first.
With a shallow, sparse clone (your own editor, a few MB instead of 2 GB)
No write access: fork the repository and clone your fork.
git clone --depth 1 --filter=blob:none --sparse https://github.com/<you>/website.git
cd website
git sparse-checkout set draft template content/authors
git checkout -b my-article
mkdir draft/your-slug
cp template/post.md draft/your-slug/index.md
# edit, add images, then:
git add draft/your-slug
git commit -m "Add article: <title>"
git push -u origin my-articleInvited as a writer on foojayio/website: no fork needed, clone the
repository itself and push your branch to it.
git clone --depth 1 --filter=blob:none --sparse https://github.com/foojayio/website.git
cd website
git sparse-checkout set draft template
git checkout -b my-article
mkdir draft/your-slug
cp template/post.md draft/your-slug/index.md
# edit, add images, then:
git add draft/your-slug
git commit -m "Add article: <title>"
git push -u origin my-articleIn both cases open the pull request from the link git push prints. Add
content/authors to the sparse-checkout set line if you are also creating
an author profile.
Only part of the tree is checked out, so hugo server won't work from this
clone; the PR check builds the site for you. To preview, use a full clone.
foojay.io/calendar/ carries three kinds of event, and they arrive by different routes. Pick the right one:
JUG meetups are pulled automatically, once a day, from the calendar each group
already publishes (its own .ics, a Google Calendar, or Meetup's export).
Nothing in this repo is hand-edited for them.
If your JUG's events are missing, the fix belongs upstream, in the
community-run World Wide JUGs directory:
add or correct your group's file there, giving it a calendar: (any iCal URL)
or a meetup: entry. It shows up on foojay's calendar at the next daily sync,
and on every other site that reads the directory too.
Conferences are pulled in daily from javaconferences.org,
the community list of upcoming Java conferences. If yours is missing, or a date,
location or call for papers is wrong, send a pull request to the
README.md of that repository
(one table row per conference). It appears on foojay.io's
/java-conferences/ map and on the calendar
at the next daily sync, and on javaconferences.org itself. Don't add a file to
data/events/ for a conference that belongs on that list.
These have no list or feed to join, so they are added by hand:
- Copy
template/event.yamltodata/events/<event-slug>.yaml. - Name the file after the event including its year —
ai-coding-workshop-berlin-2026.yaml,ocx-webinar-week-2026.yaml. Lowercase letters, digits and dashes only. Next year's edition is a new file, not an edit of this one. - Fill it in and open a pull request. Nothing else needs touching.
name: "AI Coding Workshop Berlin 2026"
type: "Workshop"
url: "https://example.org/ai-coding-workshop/"
start: "2026-09-15"
venue: "Example Venue"
city: "Berlin"
country: "Germany"| Field | Required | Notes |
|---|---|---|
name |
yes | Shown on the calendar. Include the year, like the filename. |
url |
yes | The event's own site — where the calendar sends a reader. |
start |
yes | YYYY-MM-DD. Add a time only when it matters: "2026-09-15T19:00:00+02:00", in the event's own UTC offset. Without one the event is shown as running all day. |
city |
yes unless online | |
country |
yes unless online | Feeds the "N countries" count. |
type |
no | "Conference" (the default), "Workshop", "Community Day", "Hackathon", "Webinar"… Shown next to the event. |
end |
no | YYYY-MM-DD, the last day, inclusive. Leave it out for a one-day event; a multi-day one is drawn as a band across those days. |
venue |
no | The building, if it's worth naming. |
online |
no | true for an online-only event; then city/country can be left out. |
Everything else is derived, so there is nothing to keep in step: the colour on the calendar comes from the filename, the month band from the dates, the counts in the page header from the entries themselves.
Two things worth knowing:
- Never add an event to
data/jug-events.json. That file is regenerated from the JUGs' calendar feeds and committed every day — a hand-written entry in it is gone within 24 hours. That is exactly whydata/events/exists. - You never have to come back and delete the file. An event drops off the calendar on its own the day after it ends. Removing it afterwards is welcome tidying, never urgent.
One file per event, so two people adding two conferences in the same week never
touch the same file and never get a merge conflict. The PR check
(jbang scripts/validate/Frontmatter.java) validates your entry — it will tell
you about a misspelled field, a date it can't read, or an end before the start.