"Building Mastodon to be frozen"
https://parkerhiggins.net/2017/04/building-mastodon-frozen/
I'm thinking about how we can build archiving and preservation tools right into Mastodon — where the Fediverse presents some risk — and what other opportunities that might unlock.
@riking yeah technically I'm not really sure what I'm asking for. But it seems really important that a "frozen" view should FEEL native, even though it'd have to be a different client
@xor Ah, right. Tools like IA can archive mastodon posts today with the individual view.
Discourse has the different problem of "public views use the webapp" so google cache serves a blank white page.
@xor It's also true that current federation things tend to involve tons of mutation... might be possible to do less of that, which would be better for archival: https://dustycloud.org/blog/an-even-more-distributed-activitypub/
@xor pls no. i don't want to lose opportunities because of some shit i said on the internet 10 years ago.
@kaniini it doesn't seem like the solution to that problem is to not keep a record of anything people say, but your point is noted
@xor allow people to opt out if they don't want to be involved :)
@xor This reminds me of what #ipfs is supposed to do. An out-there idea might be a mastodon-like service that uses IPFS as the backend. In the eyes of IPFS, a timeline and even an instance -- a collection of timelines -- is just another immutable directed acyclical graph.
I know the IPFS folks want a *completely* decentralized microblog, but there is value in the trade-off towards federated instances, I think. (Namely, identity management and moderation/spam control.)
@xor but some kind of indexing tool is needed. Ideally something that pulls in feeds from multiple servers and builds an index. It should be very simple to build such a tool, but is there some good reason not to?
@xor awesome post, about an important topic. I think it would be really useful to allow users to download an "archive" similar to what Twitter or Vine provided. But I wonder if it might also be possible/useful for that archive to be imported into another mastodon instance?
@inkdroid @xor Created an archive of this toot with Webrecorder: https://webrecorder.io/despens/mastodon/20170501083943/https://mastodon.xyz/web/statuses/1920693
@inkdroid thanks, and I've wondered the same. And also, like, could the Twitter archive download possibly back-populate into a Mastodon account?
> Archiveteam heroics only go so far — designing for the long-term preservation of our spaces should be a priority.
YES.
I've also been thinking that a site that continually rendered a static version of itself, committed that to a repo and then redeployed itself at some reasonable rate, pruning the database as data was pre-rendered, could have great scalability and trivial post-mortem hostability.
@xor I've thought about the difficulty of archiving discourse.org sites as well. Tried to build a better crawler view using AMP, but that was a mistake - should have just gone with plain HTML. The code isn't merged.