On a weather website, archived is a label, not a folder name. It means the page or the observation is frozen: the bytes are not being updated, and they must not be read as now. At /archived the historical TNET Weather station site kept that word in the URL. This article defines the word for station publishing—frozen pages, dated observations, and how to mark the difference from current products. It is not a dump of old Mesa tables, and it is not a registrar-lapse essay. Lapse is what happens when the hostname dies. Archive is what you do to a living hostname so yesterday’s file cannot pose as today’s weather.
The original archive index, if any, is not rehosted. There is no live cabinet of recovered HTML reports here.
Historical context
Station software produced files that looked like pages: daily NOAA-style summaries, weekly reports, “last year,” monthly climate tables, copies of NWS forecasts. Operators left them in the web root because FTP was easier than a CMS. Visitors arrived from search and saw a temperature. Nothing on the page said the file had not been written since 2005.
“Archived” in the navigation was supposed to be the honesty layer. It collected yesterday’s products so the front door could stay current. In practice the word was used for three different objects:
- a frozen page (the template is still served; the numbers no longer change);
- a dated observation (a file that was always a past interval: “April 2005 monthly”);
- a retired tool (a script or webcam that is no longer run, but the URL still returns 200).
Those objects need different labels. Calling all of them “archive” without a date is how a frozen dashboard becomes a fake live station.
TNET Weather later restored this hostname so mapped URLs remain useful. Restoration is not a claim that the current team wrote the old files, and it is not a claim that recovered snapshots are current Mesa weather. The About article states that continuity is of subject, not of team. This page states how to label the frozen part of that subject.
Frozen pages versus dated observations
A frozen page is a URL that used to update and stopped. A “current conditions” template still titled Current, still showing 72°F, still showing a clock that does not move, is frozen. The honest header is: archived, last written at an explicit datetime, not current. If you cannot state the last-write time, you cannot publish the number.
A dated observation was never “current.” A monthly summary for April 2005 is a climate table for that month. It does not become false when April 2006 arrives. It becomes false when someone strips the month off the heading. The honest header is the interval, the place, the units, and the producer (this PWS, not NCEI).
A retired tool is neither. A parser demo or a lightning page that no longer receives a feed is a documentation node. Label it as historical explanation. Do not leave a blank table titled “Live.”
| Object | What to write on the page | What not to write | |---|---|---| | Frozen dashboard | Archived; last update datetime; not current | “Live,” auto-refresh, undated degrees | | Dated interval file | Interval, station ID, units, source | A clock, “today,” a missing month | | Retired tool | Historical; original function; not rehosted | A fake run button |
The Internet Archive stores snapshots of whatever a URL returned. TNET is not the Internet Archive. A legacy article at a historical path is a living explanation. It is still not a live instrument. If this /archived page quotes a rule, the rule is current editorial practice. If it mentioned a 2005 temperature, that temperature would be historical and would need a date. Specific logger values are omitted here so they cannot be reread as a live report.
How to label archived versus current
Use words that survive a screenshot.
- Put the interval in the H1 or the first sentence. “Monthly summary for April 2005” beats “Climate.”
- Put the word archived on frozen live-slots. If the template family is “current conditions,” and the feed is dead, the first line must say archived. CSS that greys the numbers is not enough; crawlers and accessibility trees miss color.
- Keep current on a different URL when you can.
/weather.phpas live (if you still have a live station) and/archivedas the index of frozen products. Reusing the live URL for a frozen file trains bookmarks to lie. - Do not 301 archived files onto the homepage. That is how a monthly table becomes “the site.” The lapse article covers the related failure when the hostname dies. Here the hostname is up; the file is old. Keep the path.
- State producer and place. “Hobby station, Mesa area, not NWS, not NCEI.” Official archives live at NOAA NCEI. A PWS archive is a publisher’s cabinet.
- Date the label itself. “Archived 3 March 2014 when the logger was retired” is better than a bare badge. The operations log is where that retirement should already have been written.
A current product is one whose responsible publisher still updates it on a stated cadence, with a last-updated field that moves. If the field does not move, it is archived whether anyone said so or not. The label’s job is to catch up with the fact.
What this URL is not
It is not a Wayback search box.
It is not a download cabinet for recovered PHP, ZIPs, or images with unclear rights.
It is not a parking-page warning. This hostname is an active TNET Weather site. A historical snapshot of another path captured a lapse-era message; that story belongs on /main. /archived is about labeling frozen content, not about a domain that failed to resolve.
It is not a lander. Arrival and routing after revival are a different job on a different host.
It is not current weather, a warning service, or a substitute for weather.gov.
It is not an identity page. Who publishes the station website is About. Who used to read mail is contact. Those pages can themselves be current articles about historical functions. “Archived” refers to observation products and retired tools, not to every legacy URL.
Practical rules for operators and citers
If you still publish a station:
- When you retire a product, change the title and the first paragraph the same day you stop the cron job.
- Keep the old URL. Add the archive label. Do not rely on a directory named
old. - Export a copy off-host. A label on a page you later delete is not an archive.
- If you move, write a station-history note; do not silently start a new series on the old ID.
If you cite a historical weather URL:
- Record the access date and whether you retrieved a live packet, a frozen page, or an explanatory article.
- Never treat an undated temperature on an “archived” page as climatology without the interval.
- Prefer NCEI or NWS when the claim needs an official series.
Modern relevance
Dashboards, CDNs, and “last 24 hours” widgets still serve stale JSON with a friendly clock. The archive problem is now faster: a file can be frozen for ten minutes and still look live. The remedy is the same as 2005. Name the interval. Name the last write. If there is no last write, do not show the number.
Research systems that mix observations, models, and derived scores have to label those families in public. Frozen inputs must not be formatted like a nowcast. TNET’s method pages exist to keep that distinction visible. They are not a warehouse of old Mesa HTML.
TNET research bridge
The organized entry to restored station and software pages is the legacy hub. That hub is the primary next step if you arrived at /archived looking for the cabinet rather than the definition. The site-history hub groups identity, operations, and continuity articles.
How the service works states how the modern product distinguishes observed, modelled, and derived information. Data sources, quality controls, and methodology is the public note on source identity and freshness—the same discipline as refusing to call a frozen table current.