Where do we go from here?
December 10, 2022 - Reading time: 5 minutes

After a lot of rambling, the basic point in my last post was that we generate a lot of data that is used by others for obscene profit, and all we get out of that is some convenience and the illusion of autonomy.

I also brushed on the point that a lot of the software that tracks, measures, and manipulates us is based on open source software, or uses standards that are open and well-understood.  So why do we consent to the invasive tracking and manipulation?  Why not just use open source software for everything?

Well.  As it turns out - open source software actually terrible in a lot of ways.  Yes, you can self-host various things, run your own servers or pay for hosting, and feel very smug that you've outwitted Google and Facebook and the rest.  But aside from the fact that you probably didn't outwit anyone, most normal humans can't self-host anything, with or without the aid of paid hosting providers.

The fact is that open source software is technically great, but suffers from multiple flaws:

  • Open source solutions tend to require competence that most people don't have.  The average user isn't able to download source code and compile it; for those who could do it - who actually wants to?  Not most.  Linux is not ready for the desktop, and won't be in the foreseeable future.  It's hard enough to explain to the average person what a Mastodon instance is; how do you explain the difference between Fedora and Ubuntu?  Nextcloud and Owncloud?  Docker and Kubernetes?  Rpm?  Deb?  Pacman?  Flatpak?  Snap?  AppImage?  Using open source alternatives generally requires the ability to work through arcane, poorly-documented incompatibilities.
  • Open source solutions provide poor technical support.  There's no 800 number to call or helpdesk to access; there might be a post on a message board or chat system somewhere, but the average user won't know where to find these things, nor will they have the patience, competence, or time to pursue them.
  • Open source developers tend to define usability based on their personal preference, or worse: ease of implementation.  Say what you will about Steve Jobs, but he understood that form wins over function, and ease of use wins over capability.  The average person doesn't want the most feature-encrusted, technologically advanced gadget they can get their hands on - they want the gadget that is pretty and "just works."  They don't want the application that solves your problem - they want the application that solves their problem.  Form matters - a lot - but it is frequently an afterthought (or completely neglected) in open source projects, and there is a depressing lack of continuity from one application to the next.  If nothing else - large companies like Apple, Google, and Microsoft have a lot of money to spend on the science and subjectivity of what makes a widget comforting and natural; the average open source project has no money to spend on lunch, much less on squishy things like science and psychology.
  • Most individual contributors to open source projects are solving their own problems, not others'.  It's no fault of the contributors - open source software, by nature, is very libertarian - it relies on the notion that if everyone solves their own problems, and shares the solutions, everyone's problems will be solved - but this ignores the problems of those who don't have the expertise or resources to contribute.
  • At a lower level, many basic standards (e.g. for e-mail, file sharing, and so on) were "embraced" and "extended" in the 1990s and 2000s by various companies to add critical, useful features - but those elements were not rolled back into the standards.  The result is that - for example - Microsoft Outlook, which basically every company uses, has a lot of really helpful features that basic standards (like IMAP, SMTP, CalDAV, CardDAV, WebDAV, and so on) now lack.
  • The lack of UI and UX standards in open source software means that any collection of open source applications will be a confusing muddle of visual styles and themes, and basic mapping will vary widely from one application to the next.  It's no coincidence that Apple, Microsoft, and Google all have fairly detailed and consistent standards for the basic elements of a user experience - where buttons go, what they look like, the look and feel and operation of common dialogs, even color schemes and the radius of corners.  This standardization doesn't exist in open source software.

Open source software is our flawed hero - the basic skills and capabilities are there, just not the discipline, refinement, or resources. 

That seems fixable, right?  The basic problem could be solved by taking open source software, ironing out its inconsistencies, adding features, and filling in the gaps.  Then we can self-host everything!  Right?

As it turns out, the world is imperfect and unfair, and no big problem has an easy solution.  Fixing the sprawling mass of RFCs, standards, specifications, and software in the open source world is entirely possible, but it's not practical with volunteer effort alone - it requires a centralized effort with vision, process, goals, and accountability.

That's where OSDAP could help.  More on this next.