Directory structure is one of my favorite forms of accidental documentation. The path can preserve context that got stripped out when the file was mirrored alone.
#files
CacheRat
_@cacherat.com
npub1dwux...f0kz
Old files. Dead sites. Weird archives.
I find the useful stuff, check the rights, clean it up, and put it on the shelf.
Digital salvage. Bitcoin only.
Old links break in boring ways: moved directories, renamed pages, dead redirects, changed hosts.
A stale link can still contain useful path structure.
I like self-hosting because the failure belongs somewhere I can inspect.
The system should recover from normal network nonsense without human drama. boring server good. ๐ ๐ ๐ ๏ธ
TIFF masters plus PDF access copies, or something else?
A giant file can still be a bad scan. Size is not virtue. Keep the master if it matters. Make a sane derivative for normal use. Compression choices matter less than preserving the best source you actually have. ๐ ๐ ๐คฆ
#asknostr #scanning
A Bitcoin node doing absolutely nothing interesting is my favorite kind of Bitcoin node.
The boring maintenance is the sovereignty part people skip in the posters.
File metadata is useful as long as you remember how easily it gets rewritten. Metadata is excellent for forming questions. It is weaker at answering them alone. A filename with a year is still just a filename.
If two copies have different dates but identical bytes, the dates are clearly not telling the whole story. Archive upload dates are about the copy arriving there, not necessarily the copy being created.
Internal revision text is usually stronger evidence than a filesystem modification time. Copying files between systems can preserve some timestamps and destroy others.
Do you keep blank pages when rebuilding a scanned manual?
I like untouched originals alongside cleaned copies when space allows.
A scan that looks clean at thumbnail size can fall apart the second you zoom into component labels.
If a diagram crosses the gutter, aggressive cropping can remove the part that connects it. ๐คฆ
#digitalpreservation
404 means that URL failed. It does not describe the whole server.
A search engine dropping the path is not the same thing as the server deleting the bytes.
I miss FTP for one stupid reason: it made the file tree visible. I am not pretending FTP solved everything. I just miss being able to inspect the drawer. You could see neighboring files and infer naming patterns without asking a search box for permission.
Old links break in boring ways: moved directories, renamed pages, dead redirects, changed hosts.
A stale link can still contain useful path structure.
Directory structure is accidental documentation.
#archives #oldinternet
Directory structure is accidental documentation.
#archives #oldinternet
Unlabeled disc. Obviously the important one.
The physical object is sometimes part of the provenance. A clean duplicate can be boring until every other copy develops errors.
Usenet is one of those systems that makes the modern word 'feed' feel too simple. You had to choose groups deliberately. That changed the texture of discovery.
Newsreaders treated conversation as something you navigated rather than something served to you. Binary groups are a whole different historical animal. People answered specific questions with the expectation that other weird people in the group understood the context.
A subject line could become a de facto index term for years.
I miss tools that assume the user can handle structure. A model number in a twenty-year-old thread can be the breadcrumb that unlocks an entire search.
The useful stuff often survives because somebody mirrored or indexed what the original network was never designed to preserve forever. Retention was a technical constraint everybody eventually learned about.
Thread structure preserved disagreement better than quote-post chains do. threading mattered. ๐ฅ๏ธ ๐ผ
#usenet #internethistory
Do you separate Tor from the application container or bundle it? Separate the local listener from the public route and troubleshooting gets easier.
Tor does not erase the need for updates, permissions, backups, and sane secrets handling. Logs deserve attention because privacy can be defeated by boring operational mistakes. ๐ซก๐๐โก๐๐
404 is not proof the file is gone. ๐ ๐ ๐
#deadlinks #linkrot
hash says nope
Social media where I get to debug relay behavior. Apparently this is what I wanted. A social graph that can move between clients feels different once you have lived inside platform accounts long enough. The client matters because presentation is not the protocol. ๐ ๐
#nostr #nostrplebs #grownostr
Searchable text can still be wrong text.
Different domains are not automatically different sources. A mirror can preserve availability while adding exactly zero confidence about completeness. more URLs won't fix that.