How I Internet

[2019-09-23 Mon] on Yann Esposito's blog
How I protect myself against attention grabbers and many social media anti-patterns.

TL;DR: Self-hosting is not only about controlling your data, it also influences how you consume Internet for the best.

  • I POSSE1; Publish Own Site, Syndicate Elsewhere:
    • I own my data, I self-host my blog, notes, my repositories, my bookmarks, etc…
    • I use espial2 for self-hosted bookmarks and notes
    • I use node-red3 to publish my blogs, bookmarks and notes to twitter, sync bookmarks with pinboard4
  • Enhance not only publishing experience but consumption of Internet
    • read Digital Minimalism5
    • Control notifications
    • Control real-time interactions

The way we use Internet as changed drastically in a few years. A popular meme about it states:

  • toilet ⇒ 5min
  • toilet + smarphone ⇒ 55min

The book Digital Minimalism6 is really insightful on the subject. In particular the apparition of attention grabbing features that target our "social brain".

One solution to protect ourselves from the problem generated by these platforms would be to get rid of them. But those platforms are useful.

My current7 personal solution is to keep the useful features of these platforms while minimizing my exposition to most anti-features. Everything starts by how I produce content. It then affects how I consume Internet.

Producing

I self-host many services. I control my data, and then I broadcast those info to different platforms. This is called POSSE8.

Articles: self-hosted blog

First thing the classical blog.

It is more and more common now, to simply use micro-blogging, or commenting. I think blog article format is important. It is a longer form than a comment or a tweet. But in the same time, it is not necessary to work on it as hard as for a journal article.

I wrote an article that explain the technical details behind my blog. It also describes how I try to make it respectful.

To self-host anything, you should buy a domain name, and configure you DNS correctly. That is certainly the biggest blocker for non technical people.

Code: Git Broadcast

I self-host the code of my open-source projects. As Github is the de-facto developer social network, it is easier to find contributor on Github than on your self-hosted repository.

Thus I sync my repositories between my self-hosted instance and GitHub. If something goes wrong with Github, I could easily switch to my self hosted repositories only.

This is how I configured my git repos to push to multiple URLs:

git remote set-url origin --push --add <remote-url>
git remote set-url origin --push --add <another-remote-url>

Bookmarks: Espial

I also wanted a tool to keep track of web pages I like and might want to keep track of. For that, I self-host espial9; an open-source, web-based bookmarking server.

It is a very easy to install, this is a single binary. Your bookmark are kept in a single sqlite file.

This is perfect if you want to keep a lot of bookmarks some private some public. But as well as I use espial I plan (I haven't done it yet) to synchronize my bookmark from espial10 to pinboard11.

Notes: Espial

Another feature provided by espial12 is the ability to save notes.

You can generate public or private notes. I intend to use those notes for my "micro-blogging" needs. Useful, for just making some short remark without investing in a full blog post.

RSS for articles, bookmarks, notes

It is important for me to provide RSS feeds. People should know when I update my content.

So my blog, bookmarks and note generate RSS feeds13.

Syndicate Elsewhere: node-red

With those RSS, it is then quite natural to syndicate elsewhere. For that I use node-red14.

This is a web-based tool that make it easy to write flows. Think about it like a super IFTTT.

To give you an example, each time I save a new public bookmark, a new blog post, a new note, I tweet it.

Consuming

Since I generate my content using my own, self-made environment, it also influenced me for the best the way I consume and interact on Internet.

Before I used to read a lot of news directly from my smartphone. Most of the time using many apps dedicated to some social networks.

The natural presentation is an infinite scroll of content, with buttons to engage in the social network with likes/upvotes/comments etc… Most of the time, notifications where enabled by fear of missing a comment or any kind of interaction.

Before explaining how I consume Internet news, I like to make a short digression:

By writing this article I realized that, I mostly consume Internet content via news. More than that, now, Internet is almost synonymous to news on the Web. Which is only a very small part of the Internet.

Consuming news via a social networks makes you a lot more passive. I can remember being a lot more active on the Internet just a few years ago.

This is something to keep in mind I think. I will certainly write an article about that in the future.

Here is how I consume Internet content now.

News

My entry point to news consumption are:

I plan on generating RSS from those different sources with "smart filters". Typically number of upvote filters for lobste.rs, laarc.io, sub-reddits, but also number of bookmarks in popular pinboard, etc…

My preferred Internet consumption environment is elfeed15 inside Spacemacs. I really enjoy staying inside emacs as much as I can. This is a clean, dense, text-oriented environment.

I also use elfeed-org16 to organize my feeds and I also take care to remove feeds with too much volume. Generally we shouldn't read more than a few articles a day.

Mail

Most of my notifications go through my email. Social network notifications are moved inside a dedicated folder and are not directly present in my inbox. I check my social notifications once in a while. So if you are waiting for an answer, sorry for the late reply, it might take a while.

Github

I still get notifications on Github because I use it a lot for my work. But only via email and the web interface. So even for Github, I can take a few days to react.

Conclusion

I described how I control my usage of social networks. I own my data. I am a lot less exposed to attention grabbing techniques.

For now I'm quite happy with the system I made, and I'll certainly improve it in the future by synchronizing more and more services between a self-hosted one and a social-network one.

I really advice anyone with sufficient tehcnical skills to do the same. This is really worth your time.

For other people, I know some platform intented to be self-hosted and here to provide a bunch of services for you. But having a self made environment also enhance greatly the experience. And really, self-hosting is still reserved to few people.

I think we could be inspired by espial17 to create a simple small platform to provide those feature to most people.

Federation

I think I can say a few words about federated networks like Mastodon. Somehow, Mastodon replicates the anti-features promoted by Twitter or Facebook.

Furthermore I don't really like some details about the federation foundations (ActivityPub).

As an example, I wrote a commenting system that I could easily self-host. I first intended to use it for my blog. But after a second though, I'm not sure comments are that positive. I prefer to edit my articles with comments that people would send to me via other communication channel, typically, my mail.

Anti-features

A last note about anti-features. I call anti-feature a feature that provides very few or no benefit for the user but provides a lot more benefits to the platform. Generally it is a feature just here to make you stay on the platform and many quite talented specialists work on optimizing those.

Most anti-features share the same pattern; they use spaced random reward:

Spaced Random Reward

Typically the few first random gifts in a new downloaded game. The main way used to hack your brain, is by giving it something he likes at a random time. Then you start to give reward with lower and lower probability. Your brain will then be in a search mode where he will hope to get another reward by staying a bit longer in the system.


  1. https://indieweb.org/POSSE↩︎

  2. https://github.com/jonschoning/espial↩︎

  3. https://nodered.org↩︎

  4. https://pinboard.in↩︎

  5. http://www.calnewport.com/books/digital-minimalism/↩︎

  6. http://www.calnewport.com/books/digital-minimalism/↩︎

  7. Writing this article helped me to improve a lot my workflow, and I will continue to make my Internet usage evolve for a long time.↩︎

  8. https://indieweb.org/POSSE↩︎

  9. https://github.com/jonschoning/espial↩︎

  10. https://github.com/jonschoning/espial↩︎

  11. https://pinboard.in↩︎

  12. https://github.com/jonschoning/espial↩︎

  13. espial is written in Haskell, and I made a few pull requests to add RSS feeds of my public bookmarks as well as an RSS feed for my public notes.↩︎

  14. https://nodered.org↩︎

  15. https://github.com/skeeto/elfeed↩︎

  16. https://github.com/remyhonig/elfeed-org↩︎

  17. https://github.com/jonschoning/espial↩︎