About this archive

Typophile was a discussion board about typography and type design. It went offline in 2015, returned in late 2016, and went offline for the final time in 2019. What you are reading here was rebuilt from copies that web archives happened to keep.

Where it comes from

Nothing here was scraped from the live site — there is no live site. Every page was recovered from one of three independent web archives:

• Internet Archive Wayback Machine (56,947 pages)
• Arquivo.pt, the Portuguese Web Archive (no pages)
• Common Crawl (9,683 pages)

For each address, the newest valid capture before the final outage is used. Known maintenance, offline and server-error pages are skipped, and every file is checked against the checksum its archive recorded, so what you see is the bytes that were captured rather than a reconstruction. Each thread names the archive and the exact moment its copy was made, and links to that capture where the archive publishes one. Common Crawl does not: it releases its crawls as bulk data with no page to open, so those threads can only give the address the copy was taken from.

It is incomplete

The archives did not capture everything, and some of what they captured is partial. Long discussions ran across several pages and not all of them survived; where that happens the thread says so rather than quietly showing a fraction. Images and files mostly point at addresses that no longer answer.

66,630 threads and 412,434 replies recovered so far, across 34 forums.

Changes to the original content

Posts are shown as they were captured. Where that is not true, it is listed here. This list will grow as more is done to the material.

• Links point here where they can. A link inside a post that pointed at a thread or a member we recovered is repointed at our copy of it. Only the address changes — the wording of the link and everything around it are untouched. A link to something we do not have keeps its original address and stays dead, rather than pretending to lead somewhere.
• Every link says where it leads. An arrow marks one that leaves this site, a green dot one whose page or file we hold, a red dot one that is gone.
• Mangled addresses are mended. The old site turned a few hundred pasted links into broken wiki addresses. Those are put back the way they were written.
• Anything that runs is removed. Scripts, styles, embedded frames and objects, forms, inline event handlers and javascript: addresses are stripped before a page is rendered. The words and the structure of a post are left alone.
• Embedded images and attachments point here where we have them. A picture in a post, or a file somebody attached, is copied out of the archive and the post is pointed at our copy. Where the archives never captured the file, the tag is left as it was and the image stays broken rather than pointing somewhere that looks right and is not.
• Member pages show less than the profiles did. Where somebody lived, when they joined, and details they filled in about themselves — gender, social accounts, occupation, education — are not shown. Only a name and a home page remain.

The originals are not ours to alter: every thread names the archive its copy came from and the moment it was taken, and links to that exact capture wherever the archive can show it.

Who made it

Put together by Stefan Huber. The code that fetches, verifies, parses and renders all of this is on GitHub — including the parts that record what is missing.