2016/Longweb
Building for the Long Web was a session at IndieWeb Summit 2016.
Notes archived from: http://etherpad.indiewebcamp.com/2016longweb
Participants:
- emmak
- tantek.com
- isgeek.net
- yourmom.io
- cassie.wtf
- ...
Background:
- Longevity
- https://adactio.com/articles/1522/ Author:Jeremy Keith
- https://www.w3.org/Provider/Style/URI.html - TBL
Build the web
Iron Gaul Ink; primative technology but very durable. Magna Carta, Constitution of US
Adding to that: They would scratch back the ink to reuse the velum, but the traces are still there for archeologists to recover the image of the previous text.
Emma: Unix account, "as an alumn, you'll have this account forever". A couple of years later, they deleted it and I was so made. Applies more broadly than personal blogs
Comment (Lloyd): Interested in finding meaning in the long web. Just end up with a pile of data, no schema.
yourmom.io: As someone with a historian past, the amount of data on the web that is stored in a fragile, ephemeral way is shocking. Less worried about the unstructured aspect, historians are good at dealing with that.
JJ: Wish it was not a problem that only nerds have (awareness). Make this a normal concern
Open Question: I use plugin "amber" for WP which sends to Internet Archive, wayback automatically.
Look at what happened with Geocities shutdown. People involved who weren't archive.org
Salt: What happens when some maintainer say bzip
Tantek: Like what happened with code.google.com?
Will: The projects that weren't migrated by their maintainers to github went into cold storage, and the urls will persist.
yourmom.io: but data has always been lost (in the past). We make best faith effort, some of our data that the future will want will be lost.
JJ: Incremental noise. New data: I now have my sleep patterns. My kids will have a copy of their DNA on their terrabyte keychain.
Will independent sources exist?
Salt: If you own your data, can you opt out of the long web.
Sandro: not just my stuff, but the things I like, and, popularity increases the number of copies (Sub resource integrity - https://www.w3.org/TR/SRI/)
edrex: merkle DAGs allow verification of data, including dependencies.
examples:
- git
- IPFS
- camlistore
Solutions:
- emmak: I'm using static content for my site
a number of attributes useful for long web
- aspects: inspectability - can I look at the data and understand it
- durability - how resistant to corruption is it
- portability - my site is on amazon web services, who knows where it will be in five years
eventually I'll have to migrate it if it's a chore to import it all, that's going to jeopardize preserving that data
- redundancy. I'm storing my data on both s3 and git. the copy on github has a timeline, if I screw something up, I can go back.
edrex: "git repo is like the zip file of our generation"
JJ: Github getting away from just code, more to helping people collaborate
Tantek: I do static content too, still compile but source is microformats HTML
Common: mysql doesn't get archived, and you get an error (which is then archived) https://www.google.com/search?q=%22Database+Error%22+%22Error+establishing+a+database+connection%22+-fix+-wordpress+-website+-login+-help+-account+-join+-atg+-scary+-problems+-mining
Another aspect: EU Right to be Forgotten
yourmom.io: Google has been taken to court about not being able to comply with it
Google in contempt of court in france
JJ: Technology is moving faster than Moral framework
yourmom.io: and legal framework
We're culturally almost used to generational data loss (VHS, tapes, etc)
edrex: DRM'd Walled gardens (esp younger generations) with user generated content
Sandro: Archiving my kids' Minecraft worlds
Salt: how to save just the good stuff?
William: I loved digital photography when i had sight. I didn't always know what was good. My friends and relatives would find gems later. I appreciated the large memory card.
JJ: Traveling, I take lots of pictures. Looking back reminds me of places, experiences
edrex: "We are creating these empathetic entities that are non-human"
Tantek: How are people preserving their private data for the longweb?
JJ: URLs done the right way as an important contribution to the long web (Tim Berners-Lee)