Triple Entry Blogging
tomcritchlow.comI like the tone. But what the author describes as potentially world changing enablers is not triple entry blogging, but open standards, open APIs and (very) easy to use distributed nodes in some Form of mesh that uses these open standards to build cool stuff on top of this networked open standards based mesh.
I just sadly doubt that most people outside a very specific bubble would want to have their own node in said network. Even without the need to care for it, think about hosting and so on.
Most people are fine with using "rented real estate" on a centralized platform. It doesn't cost them time, nor money (advertising and data collection as revenue streams). They connect with people (social networks) or use a service (Goodreads) to achieve some form of goal.
Why would say add any friction for themselves? Any complexity above that?
A much as I would love to have these decentralized open networks, I will miss them in the future. Sadly it seems that central platforms have won.
I play Blood Bowl, both in virtually and as the little miniature game, and since every Danish person in the world of Denmark is basically on Facebook and nowhere else, it’s been the primary platform for coordinating leagues and tournaments for the past many years. The platform has gotten so bad at it though, that there are serious thoughts about moving somewhere else, the issue being that everyone is on Facebook and nowhere else.
Maybe you’re going to be correct, but I think that these closed platforms are all going to do what Facebook has and slowly become places that a lot of their users no longer like. In that environment I think someone is going to come a long with a decentralised form of social media the way e-mail or a phone works that is easy enough to connect through that everyone in Denmark is going to be on that and not everywhere else.
I think the only reason it doesn’t exists yet is because you sort or needed a world where GitHub pages, and Dropbox, and so on figured out a way to sell their platform IRON through the use of open APIs. I think it’s coming though.
> Maybe you’re going to be correct, but I think that these closed platforms are all going to do what Facebook has and slowly become places that a lot of their users no longer like.
I agree with platforms turning into a state that is not liked by their users. While I still have my FB account more than one year ago I removed everything as an experiment and didn't want to return after a month.
IG is becoming more and more the same for me, while I still use it for a secondary account for business. I still don't like it.
I would love for some Form of decentralized thing that I could host as well as could join a network that monetizes somehow and offers space for people who don't want to host. All embedded in a Form of mesh that enables shared groups across nodes, discovery, followship while enabling me to decide what specific parts/apps/tools to host and use and how much I am willing to share.
Edit and btw: Greetings from near Hamburg only "slightly south of the border" ;)
So double entry accounting records each entry as a debit and credit which enables all sorts of new possibilities in accounting. Triple entry blogging is just having two backups for everything with completely different technology from those backups enabling some other things? I don’t get it. I don’t see having two backups as being any sort of secret sauce the way double entry was.
I think this blog post is the result of coming up with a headline that sounded cool, but doesn't really have, as followthrough, the kind of "punch" you're expecting to be behind it. He ends up shifting gears midway through and talking about something perhaps completely unexpected to the reader if the reader were to begin forming their own conclusions about what the piece is truly about based on what's written at the start.
To be fair, I recognize that this is a rough (but enthusiastic) sketch of a lot of not-fully-formed ideas.
I also recognize what he's talking about. I spent a lot of time during COVID lockdowns refining similar thoughts. It's weird, though, the continual references to platforms like Replit. The real point of "Triple Entry Blogging", if I understand the decision to lead with that, is that it operates on inert data. Just plain, dead media—like RSS/Atom (or JSON blobs). It even uses as an example the same use case that I point to here—Mastodon. What he really wants is something less demanding of compute on the backend. Again, though, nothing that relies on Replit- or Glitch-level powers makes for a good example of this principle put into practice.
Moar/previously: <https://news.ycombinator.com/item?id=30862781>
I could see that if triple entry blogging included saving and exposing a raw text version of the blog. But it doesn't.
I handwrite all my blog posts in a journal then use Paper Website [1] to publish them online. This is great because I have a physical paper copy, and a digital one.
This "double entry" system has been pretty useful - I accidentally deleted a page once, and recovered it back by simply taking a picture of the physical page again - was pretty neat!
The comparison with double entry bookkeeping seems a little bit far fetched, in my opinion. But I like the enthusiastic tone of the article and I understand the comparison as a starting point for brainstorming new ideas!
I am now reading his post about RSS (https://tomcritchlow.com/2022/04/21/new-rss/). Good stuff!
How to implement what OP wants to have:
- their content stored on a PouchDB/CouchDB server. blog posts, comments, shared memes, listened songs... anything can be represented as an ActivityStream document
- files/attachments can go to an IPFS store
- the ability for authenticated users to pull this data, and remix into their own database.
That is it. You can then have all kinds of tools to author the content. You can have your static site generator pushing your content there. You can have a audio scrobbler and get last.fm. Any commit to github can become a entry in the journal. You can send pictures from your phone. Saved bookmarks. Shared cooking recipes...
So you will have your different sources of truth (git for the text content, your media library for your pictures), you have the database and you have your peers databases as the "remixed" copy of your data.
I think Tim Berner-Lee's SOLID project intends to - at least conceptually - provide the needs that you cited: https://solidproject.org/about
Maybe the implementation layer details might differ, but it does seem similar to me...which i should add that i like this approach.
One of the problems with Solid is that it's definitely guilty of the sort of criticism that surfaced (brought up by a reader) in related to some of the things discussed in an earlier post about "Library JSON" <https://tomcritchlow.com/2020/04/15/library-json/>:
> That said, the thing that kills the indieweb is too much enthusiasm for specs, too few[...] implementations, and, ironically, too much focus on the ‘indie’ (building complicated self-hosted everything-machines) and not enough on the ‘web’ (noticing if anyone’s using any of the things you built). So if there’s a killer implementation and good content at the start, then momentum would potentially just carry you through.
One thing I'm fond of is a turn of phrase earlier coined by Noel De Martin, a contributor to Solid: "Autonomous Data". The intention to put resources behind "Autonomous Data" as a brand/movement was set aside to focus on the Solid effort instead. I think, though, that the term is well-suited for the rough ideas presented in this post. (Better than "triple entry blogging", in any case. Better than "Solid" or "Zero Data"[1], too, in my opinion.)
Agreed with your comment on autonomous data; and thanks for sharing that link, info!
Interesting. I have taken a look at SOLID when it was first announced, but there wasn't much there, there. Now, seems like now they've progressed quite a lot! Multiple server implementations, listing of service providers... for all the talk about how committees can not deliver, it seems they are miles ahead of the Bluesky people in realizing their vision.
Imagine something like bittorrent or ipfs (?) would be part of web browsers: ad-hoc publishing without the headache for "servers".
Check out Beaker Browser! It’s pretty much exactly what you’ve described: https://beakerbrowser.com/
Beaker is no longer in development IIRC, you might want to check AgregoreWeb which supports Hypercore, IPFS and more https://github.com/AgregoreWeb/agregore-browser
Beaker creator here, that's correct, but Agregore is still going
I love the idea of physical libraries establishing federated social media offerings, leveraging protocols like ActivityPub (to ensure interop)! To be sure, there would be challenges, and the model would not be ideal for all types of social media users...but I really like the opportunities that this approach can provide.
So.. a CI/CD pipeline?
You have the source (your local text files) which is deployed somewhere (GitHub pages) which other entities can access (spiders crawling it).