The calendar page brought its own ICS parser and its own RRULE expansion. Both
only covered the cases that happened to be needed when they were written, and
the calendar has moved on since. Against the published calendar, for September
2026:
Spieleabend FREQ=WEEKLY;INTERVAL=2;BYDAY=SA shown 05. 12. 19. 26., correct 05. 19.
CCCB Plenum FREQ=MONTHLY;BYDAY=TU;BYSETPOS=2 shown 01., correct 08.
CCCB Plenum FREQ=MONTHLY;BYDAY=TU;BYSETPOS=4 shown 01., correct 22.
INTERVAL was only read for monthly rules, so the Spieleabend was shown twice as
often as it takes place. BYSETPOS was not implemented at all, and since
parseInt("TU") is NaN the fallback turned both Plenum rules into "first
Tuesday", putting two Plenums on a day without one and none on the two days
with one. UNTIL, COUNT, EXDATE, RECURRENCE-ID, BYMONTHDAY and a BYDAY listing
more than one weekday were not handled either.
The text was no better. Content lines longer than 75 characters are continued
on the next line, of which there are 477 in the calendar, and the parser did
not join them, so it cut values off in the middle of a word. It also split
every line at the first colon, which lands inside the parameter of
DESCRIPTION;ALTREP="data:text/html,...". And it never resolved the escaping, so
"\n" was shown as those two characters. 30 of 31 descriptions were wrong:
before: "Der Club Discordia ist ein öffentliches Treffen in den Clubr"
after: "Der Club Discordia ist ein öffentliches Treffen in den Clubräumen des CCC Berlin"
Hand the parsing and the expansion to ical.js, which is vendored for the start
page anyway. The month view now asks the library for the occurrences that touch
the month, which removes the reimplementation along with all of the above.
While the events are being reduced to what the view needs:
- An event is entered on every day it covers, so the Amateurfunk trip from
30.10. to 01.11. is no longer marked on 30.10. alone. The end of an event is
not part of it, so one ending at midnight stays on the day before.
- Days and times are derived in Europe/Berlin instead of from the digits of the
ICS string. Every event currently carries TZID=Europe/Berlin, so the wall
clock time shown does not change, but a UTC timestamp would have been shown
in UTC.
- The URL of the event is used for the link in the detail panel. Events without
one are shown without a link, as on the start page. This replaces
createEventLink(), which guessed URLs from the title and was never called,
and the panel no longer builds an <h> element, which is not an element.
Descriptions may contain line breaks, so keep them in the panel.
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The upcoming events table pulled its ICS parser straight from a CDN with
`import ICAL from "https://unpkg.com/ical.js/dist/ical.min.js"`. That sends
every visitor of the start page to unpkg.com, which hands their IP address and
user agent to a third party before any of our own code runs. The URL is not
even pinned to a version, so whatever ical.js publishes next is executed on our
site without anybody looking at it, and the start page silently breaks when the
CDN is unreachable.
Check the parser into `assets/js/vendor/` and let Hugo bundle it. This is what
`js.Build` is for: it runs the esbuild that is built into Hugo, so it resolves
the import at build time and needs no node_modules and no extra tooling in the
build environment. The result is minified and fingerprinted like the other
scripts of the site, and the script tag carries a subresource integrity hash.
Since the bundle now has a content hash in its name, its URL cannot be written
by hand in the markdown any more. Move the table and the script tag into an
`upcoming` shortcode, which is the same pattern `calendar.html` already uses,
and move `upcoming.js` from `static/` to `assets/` so Hugo can process it.
The vendored file is the unminified `dist/ical.js` of the pinned release: it
carries the MPL-2.0 header and is the source form of what we ship, and Hugo
minifies it for delivery anyway. `assets/js/vendor/README.md` records the
version, where it came from and how to update it.
The built bundle renders the same table as before, checked against
https://berlin.ccc.de/calendars/all.ics.
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The response of the fetch went to the parser without ever looking at it. When
the calendar could not be loaded the error page of the web server was parsed as
a calendar, which threw inside a promise nobody was waiting on. The result was
an empty table, an unhandled rejection in the console and an error message
about broken calendar syntax that says nothing about the actual problem, a
calendar that is not there.
Refuse a response that is not ok, naming the status, and log failures in the
chain, like the calendar page already does.
before: Uncaught (in promise) Error: invalid line (no token ";" or ":")
"<html>404 Not Found</html>"
after: Fehler beim Laden der Termine:
Error: /calendars/all.ics: 404 Not Found
The table stays empty either way, there is nothing to show without a calendar.
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
`toLocaleString()` was given the German locale but no time zone, so it printed
the time in whatever zone the browser of the visitor is set to. The events
happen in Berlin, so this is only correct for visitors who are in Berlin.
Somebody reading the start page from Sydney was told the Plenum of Tuesday
20:00 takes place on Wednesday at 04:00.
Format in Europe/Berlin explicitly. Only the printing was wrong, picking and
sorting the events works on absolute points in time and was not affected.
Drop the two replacements around the formatted date while touching it. The
first removes the comma after the weekday and the second puts it back, so they
cancel each other out:
"Samstag, 22.08., 17:00" -> "Samstag 22.08., 17:00" -> "Samstag, 22.08., 17:00"
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
An occurrence of a recurring event that was changed on its own carries a
RECURRENCE-ID and is stored as an additional VEVENT next to the event it
belongs to. `getAllSubcomponents("vevent")` returns those components as well,
and since they have no RRULE of their own they were handled as separate single
events. The occurrence therefore ended up in the list twice: once from
expanding the recurring event, which resolves the modified time through the
exception, and once more from the extra VEVENT.
To make it worse the two rows disagreed, because name and URL were taken from
the recurring event while the time came from the modification, so the first row
showed the new time under the old name.
Skip components that are a recurrence exception, they are already covered by
the event they modify, and take name and URL from the occurrence details, which
point at the modification where there is one and at the event itself otherwise.
Events whose RECURRENCE-ID refers to an event that is not in the file are
dropped by this, which cannot happen in a full calendar export.
The published calendar currently contains no RECURRENCE-ID at all, so nothing
changes for it today. It is exported from a CalDAV server though, and moving a
single Club Discordia or Plenum out of the way of a holiday is exactly what
creates such a modification.
With a recurring Plenum whose 25.08. occurrence is moved two hours earlier and
renamed:
before: 25.08. 18:00 CCCB Plenum
25.08. 18:00 CCCB Plenum (verschoben)
after: 25.08. 18:00 CCCB Plenum (verschoben)
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The "Nächste Veranstaltungen" table selected events with `start > now`, so an
event disappeared from the list the moment it began. Someone looking at the
start page at 20:30 no longer saw the Plenum that had started at 20:00 and ran
until 22:00, which is exactly when that information is most useful.
Select on the end of the event instead, so an event stays listed for as long as
it is still running. A currently running event sorts first, because the list is
ordered by start time.
For recurring events the end of the individual occurrence is needed, not the
end of the series, so go through `getOccurrenceDetails()`. That also resolves
occurrences overridden via RECURRENCE-ID. The chronological break out of the
iteration keeps using the raw occurrence time, which stays monotonic even when
an override moves a single occurrence. `ICAL.Event.endDate` falls back to
DURATION and, for all-day events, to the following day, so events without an
explicit DTEND keep working.
The `maxDays` window still applies to the start of an event, so a long running
event does not extend the window.
This is not a regression from the commit below, the Python generator it
replaced filtered on `dtstart >= start` in the same way.
Checked against https://berlin.ccc.de/calendars/all.ics: the Plenum (20:00 to
22:00) is now listed at 20:30 and 21:59 and gone at 22:01, and the multi-day
Amateurfunk trip (30.10. 12:00 to 01.11. 18:00) stays listed throughout.
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
The "Nächste Veranstaltungen" table read the event URL via `ICAL.Event.url`,
but ical.js does not expose a `url` getter on `ICAL.Event` (it only has uid,
summary, description, color, location, sequence, the dates, organizer and
attendees). `event.url` was therefore always `undefined` and the `?? ""`
fallback turned it into an empty string, so every row rendered as
`<a href="">`, a dead link that just reloads the start page.
Read the URL from the VEVENT component instead. This also picks up the
`URL;VALUE=URI:` form used by most events in the published calendar, which is
exported from a CalDAV client and does not use a bare `URL:` property.
Not every event has a URL, so only wrap the name in a link when one is
present and emit plain text otherwise.
Checked against https://berlin.ccc.de/calendars/all.ics (31 events, 2 of them
without a URL):
before: <td><a href="">CCCB Plenum</a></td>
after: <td><a href="https://wiki.berlin.ccc.de/Plenum">CCCB Plenum</a></td>
after: <td>Aktionstag gegen Überwachung im Chaos Computer Club Berlin</td>
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
`build.sh` no longer post-processes anything: since the switch to ICS files it
just deletes `public/` and runs `hugo`. The `CALENDAR` placeholder, the Python
dependency on `icalendar`, `python-dateutil` and `pytz`, and the `de_DE.UTF-8`
locale requirement are all gone, and nothing in `packages.nix`, `devShells.nix`
or `flake.nix` pulls in Python any more. Drop those claims.
Describe instead where the two calendar views actually come from: the
"Nächste Veranstaltungen" table and the calendar page are rendered in the
browser from `/calendars/all.ics`, which is published next to the site and is
not built from this repository. That also replaces the old advice to preview
them via `./build.sh` plus a local HTTP server, which no longer helps.
Without that file the two of them stay empty. Hugo serves everything below
`static/` from the root of the site, so it can simply be put there, and the
command that downloads it is worth writing down. Checked from an empty state:
afterwards the file is served under `/calendars/all.ics` by `hugo serve` and the
table on the start page fills with the upcoming events. It is a copy of
published data and has no business in the repository, so ignore it.
Mention the per-section `.ics` feeds Hugo does generate, so the ICS templates
under `layouts/` are not mistaken for the source of `all.ics`.
Fixes: c28f04c6e8 ("switch to ics files; make calendars work; fix some minor issues")
Assisted-by: Claude:claude-opus-5
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Der DI.Day findet am 5. Juli 2026 nicht im Chaos Computer Club Berlin
statt. Der neue Blog-Eintrag informiert darüber und verweist auf die
alternativen Berliner DI.Day-Veranstaltungen (c-base, Stadtschloss
Moabit).
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Der Digital Independence Day findet ab Juli 2026 nicht mehr im
zweimonatlichen Turnus statt. Mit UNTIL in der RRULE endet die
Wiederholung nach dem letzten Termin am 3. Mai 2026, sodass keine
künftigen Instanzen mehr in den Kalendern (all.ics,
veranstaltungen/index.ics) und in der Terminübersicht erscheinen.
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
gen_upcoming.py normalisiert dtstart auf naive lokale Zeit, reichte die
laut RFC 5545 in UTC anzugebende UNTIL-Zeit der RRULE aber tz-aware an
dateutil weiter. Die Mischung aus naivem dtstart und tz-awarer UNTIL
lässt dateutil.rrule mit "RRULE UNTIL values must be specified in UTC
when DTSTART is timezone-aware" abbrechen -- der Build schlägt fehl,
sobald ein wiederkehrender Termin ein UNTIL enthält.
Mit ignoretz=True wird die UNTIL-Zeit ebenfalls naiv behandelt,
konsistent zur restlichen Verarbeitung. Für Termine ohne UNTIL ist das
ein No-Op.
Fixes: 2e3a02af0c ("refacture all the things!")
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
