Sign inRegisterCheck connection

The Cumulus Default Template: Chrome Around Current Conditions

Cumulus main was the default template: site chrome around a current-conditions slot bound to logger tags, not a parked-domain page or the subdomain root.

Back to Site history

On a Cumulus website, main is not a parked-domain story and not the software-host essay that occupies this hostname’s root. It is the default template page: the chrome that identifies the station, plus a slot that displays current conditions from logger tags. Visitors treated that page as “the weather site.” Operators treated it as the file the station program processed and uploaded on a timer. Those are the same HTML document doing two jobs, and the jobs must stay labeled.

This URL, /main on cumulus.tnetweather.com, is that template idea. The original processed HTML is not rehosted, and no live Mesa (or other) readings are reprinted as current weather. The article explains how chrome and the conditions slot work, why units and time belong in both, and how this page differs from the subdomain homepage and from the main-host lapse article.

Historical context

Steve Loft’s Cumulus and later Cumulus MX both generate a public site by substituting web tags into HTML (and other) templates. The Cumulus Wiki web tags page is the living list of names such as <#temp>, <#timehhmmss>, and unit tokens (last checked 13 August 2026). A typical default site had a front page that looked finished in a browser: banner, navigation, a large temperature, wind and pressure nearby, maybe a small trend image, a footer with software credits. Under the skin it was a template with empty slots. Cumulus filled the slots and FTP’d the result, or MX served an equivalent through its own HTTP stack.

Community PHP often wrapped the same idea in a CMS-style filename: main, index, default. Incoming links copied /main because that is what the menu called the front page. This restoration keeps the path. It does not keep a live tag processor at this address.

TNET does not maintain Cumulus. Current behaviour is defined by the Cumulus Wiki and, for MX, by Cumulus MX on GitHub (last checked 13 August 2026).

Chrome is metadata, not decoration

Chrome is everything on the page that is not the latest sample: station name, location line, navigation, units legend, software version, “updated” wording, and footer credits. It looks like design. It is the metadata that makes the numbers in the slot interpretable.

Identity. A stranger who lands from a mesomap needs the station name in text, not only in a banner graphic. Graphics fail; crawlers and accessibility tools read the heading.

Location. City and region belong in chrome so the temperature is not an orphan number. Coordinates, if published, are configuration. They do not have to dominate the front page; they must not contradict the compact info card.

Units. Cumulus templates that print <#temp> without <#tempunitnodeg> (or an equivalent unit string) produce a bare “21.4.” Chrome should state the unit system once, and the slot should repeat the unit next to each scalar. A page that says “US units” in the footer and “°C” beside the thermometer has failed.

Software citation. “Generated by Cumulus” plus flavour and version is how a later reader knows the tag vocabulary. Classic Cumulus 1 and Cumulus MX do not share every field. Chrome that says only “weather station software” is not a citation.

Navigation. Links to mobile, webcam hub, graphs, and extra sensors are chrome. They should go to real resources. A nav item named “Live” that points at a stale JPEG is a conditions-slot problem wearing a menu label.

Chrome must not impersonate an official forecast office. A Zambretti-style or software forecast phrase in the header is a scenario from the station program, not an NWS product. Official U.S. forecasts and alerts remain at weather.gov.

The current-conditions slot is a contract

The slot is the region bound to the latest logger packet: outdoor temperature, humidity, wind, pressure, rain today, packet time. On a Cumulus site those values come from web tags or from a PHP consumer of realtime.txt. Either way the slot is observed (plus software-derived indices if you choose to show them). It is not a 24-hour graph, not a climate table, and not a forecast tab.

A complete slot states:

  1. Each quantity with its unit — temperature with °C or °F, wind with the station wind unit, pressure with hPa or inHg, rain with mm or in.
  2. Packet time with timezone — <#timehhmmss> / the realtime clock field, not the time the visitor’s browser rendered the HTML.
  3. Which values are derived — dewpoint, wind chill, heat index, “feels like.” Those are calculated. They are not extra thermometers.
  4. A stale rule — if the processed file is older than two upload intervals, the slot must say so. A green “LIVE” badge on a file from 11:40 is a false observation.

Missing values must not become zero. A logger dropout printed as 0.0 is a fake calm. Use an explicit missing token.

The slot may show today’s high and low, but those extremes carry event times that are not the packet time. Mixing “now 21.4°C” with “high 28°C” without the high’s clock is how visitors invent a trend that the page did not measure.

Trend thumbnails in the slot are figures of a series, not extra observations. If the PNG is generated from a homemade realtime log, it is only as complete as that log. The scientific-figure rules live on the JPGraph graphs article. The main template should caption the thumbnail (“24-hour temperature, station units”) or omit it.

Why /main is not / and not the other /main

Three URLs in this restoration look like “home” and are not interchangeable.

| URL | Host | Job | |---|---|---| | / | cumulus.tnetweather.com | The hostname as a software citation host | | /main | cumulus.tnetweather.com | The default template: chrome + conditions slot | | /main | tnetweather.com | Domain lapse and why historical URLs still matter | | /lander | cumulus.tnetweather.com | Entry routing after revival, without a mass homepage 301 |

A citation-host essay answers why the Wiki pointed at this hostname. A template essay answers what visitors read for “now.” A lapse essay answers what breaks when the domain dies. Putting any of those theses on the wrong path would be a second, semantic lapse.

The lander exists so a bookmark named /lander can send a human to the hub without claiming to be this template. /main stays the template.

Practical template rules

If you still publish a Cumulus (or MX) front page:

  • Bind the slot to one source file and one clock. Do not mix a web-tag temperature with a realtime.txt wind from a different upload.
  • Put units in the slot, not only in a CSS class name.
  • Keep chrome identity in HTML text. A background image of the town name is not a station ID.
  • Separate forecast includes from the slot. NWS text is a forecast with a valid time; it may live on the same URL as a second block, never as the headline temperature.
  • Do not autoplay a webcam in the slot. A still with interval and time belongs on the webcam hub. Motion is not a timestamped observation.
  • Mobile visitors who want a smaller chrome should get /m, not a squashed desktop table. The main template can link there. It should not pretend to be the mobile root.

The web-tag vocabulary is the binding list for names. This page does not extend that list.

Modern relevance

Default weather homepages still fail in the same places: a large number without a unit, a “updated just now” line driven by page generation rather than logger time, and a forecast icon sitting on top of a thermometer. Responsive CSS did not fix those errors. JSON APIs did not fix them either. The slot is still a contract about observed values at a stated clock.

Cumulus MX can push JSON and let a browser draw gauges. That is a different chrome and a different slot implementation. The scientific requirement is unchanged: label source, units, and time; distinguish derived indices; do not present a software forecast as an official warning.

TNET does not run this URL as a live dashboard. The page teaches the template rules that made a thermometer readable when one existed.

TNET research bridge

A conditions slot is a small observation record. How TNET distinguishes observed, modelled, and derived information in public research is how the service works. Source identity, units, and freshness for records used in connection research are on data sources, quality controls, and methodology. Those pages do not parse Cumulus tags.

The Cumulus hub lists related historical pages. Site history groups this template article with other continuity pages on this host.

Sources