Ten searchable libraries for the facts you look up rather than calculate: HTTP status codes, network ports, media types, file extensions, open-source licences, CSS named colors, symbols and HTML entities, paper sizes, countries, and regular expressions. Each one is a filterable finder with a page per record, and several of them let you find an entry by pasting an example instead of knowing its name.
- 01
Debug a response you did not expect
Paste the log line, land on the status code, and read what the code actually obliges a client to do — including whether the response is cacheable and whether the request may be retried.
- 02
Harden a host
Paste `nmap` or `ss` output into the ports finder and it reorders by exposure risk, with the firewall command and the encryption note attached to each service you left listening.
- 03
Serve a file correctly
Identify the extension from a filename or URL, take the exact `Content-Type`, and check whether the format is worth compressing before you add it to the server config.
- 04
Pick a licence, or read one you inherited
Compare permissions, obligations, reciprocity, and patent grants across licences side by side, then copy the SPDX identifier the tooling expects.
Search by example, not by name
The regex library runs every pattern against a sample you paste and reorders by what matched; the MIME finder runs real RFC 9110 content negotiation on an `Accept` header; the extension finder takes a filename, a URL, or a media type. You can arrive without knowing what the thing is called.
A page per record
Every status code, port, media type, symbol, and country has its own URL with its own detail. They are real pages — linkable in a ticket, openable in a new tab, and readable without JavaScript.
Derived, not just listed
The value is in what the data does not say on its own: exposure verdicts and encryption notes for ports, inline-execution risk for media types, contrast pairs for colors, print-resolution pixel sizes for paper, seven encodings for every symbol.
One filter row, not a wall of chrome
Search, facets, and sort share a single row above the results, and the count sits with the list it describes — so the rail shows records rather than controls.
The search survives navigation
Opening a record keeps your query and filters intact, and the browser's back button behaves the way you expect, because the selection lives in the path rather than in component state.
Static and offline-friendly
The datasets ship with the page. There is no API call behind a lookup, so results appear as fast as you can type and keep working with the network disconnected.
What does HTTP 429 mean?
Too Many Requests: the client has been rate limited. The useful part is the `Retry-After` header, which tells the client how long to wait — the status page covers what to send, what a well-behaved client does with it, and how it differs from 503.
How do I find out which service uses a port?
Search the port number, or paste the output of `nmap` or `ss` and the library reorders itself around what you pasted. Each entry names the service, says whether traffic is encrypted, and gives the firewall command to close it if it should not be open.
Which Content-Type should I serve a file with?
Type the filename or the extension into the file-extension finder and it gives you the exact media type, whether the format is already compressed, and how browsers behave with it — including which types execute inline and therefore need care with user uploads.
What is the difference between MIT and Apache-2.0?
Both are permissive and both require the notice to be preserved. Apache-2.0 adds an explicit patent grant and a termination clause for patent litigation, and requires changed files to be marked. The licence comparison lays those obligations out row by row.
How many pixels is A4 at 300 dpi?
2480 × 3508. A4 is 210 × 297 mm, and the paper library converts to pixels at whatever resolution you name, alongside bleed sizes and the closest non-metric alternative — Letter, in that case.
Are these libraries kept up to date?
The reference libraries are maintained datasets rather than live mirrors: entries are derived from the relevant RFCs, standards, and registries, and reviewed as those change. Each record says what it is derived from, so you can follow the source when it matters.