Over the last decade or so, we’ve witnessed a continued ramp up of the information age. Social media has become the “social” place. These days, YouTube publishes 3.7 million videos a day, that’s 271’000 hours every single day. We’re bombarded with information from all directions day in and day out, so much so, that we often don’t even care or stop to think about, whether the provided information is true or actually relevant to us.
While information overload and misinformation are rampant and a real problem, I want focus instead more on the fleeting, ephemeral nature of how we’re communicating and specifically within FOSS (Free and Open Source Software) communities.
The Situation
A lot of online discussions have moved from mailing lists and self-hosted web forums to platforms such as Reddit, GitHub, or even Facebook groups. Additionally, the need for real-time chat has increased and many picked simple and modern solutions such as Slack or Discord.
The advantages are quite obvious:
- Everyone has already an account most of the time
- There’s no hosting fee
- There’s no maintenance cost
- It scales for any size of an user base
- It’s just so simple to use and setup
For SFML we do have a subreddit, a Discord server and all of our code and issues are on GitHub. However, we’re still hosting the website on our own, run our own “old school” web forum and provide a bridge between IRC and Discord. Unfortunately, it often feels a bit like a battle to retain these “arcane” things, as people “suggest” to just move everything to one or the other platform instead, but I think there are important factors, that are often overlooked in favor of low maintenance cost or ease of use.
The Factors
Having your own information sharing platform, gives you the power to ensure the retention of said information, i.e. that it remains available, for as long as the product exists. It can observe many “exciting” services launching and their slow deaths, or it can easily outlive the original author. Nobody, is going to just delete “old information”, because the new system doesn’t support it or there was a need for more space. When retention is not prioritized, you can end up in the situation like the Microsoft documentation, which will return 404 pages for a lot of existing links, because those sites were moved or removed, leaving people with no access to the valuable information. When information vanishes, it might be, that it never returns.
The SFML forum hosts discussions dating back to 2007, so you can find discussions on old decisions, track down problems with old systems or hardware, if that’s your thing, but also just look at the latest posts – old data doesn’t hurt new data.
I think the hardest points for people to understand is information governance, especially since it doesn’t necessarily affect them as an user and especially not right this moment. Being able to own your data is an essential step in ensuring your independence and guaranteeing information retention. You’re not at the will of some platform, be it that they suddenly require a retro-active enforcement on some political correctness of the day or outright ban you with no reason or recourse. You’re in the driver seat and you own the information, allowing you to move it around as needed.
For SFML, this is a constant uphill battle, as people see forums as obsolete (there’s Reddit!) and old (there’s GitHub Discussions!), or when the SFML Website goes down for some reason, everyone wants to move somewhere else (there’s GitHub Pages!).
All that information isn’t exactly useful, when it remains hidden. Searchability is crucial in getting access to this wealth of information. This is unfortunately, where the use of real-time chats and specifically their implementation, leaves a big hole. Even with IRC, you didn’t have a default way of archiving the history or make it searchable. Mods and bots were created to make this possible. Yet, we see the same thing happening with Discord or Slack. Whatever you’re shouting into these chat applications, will never leave their platform and remain unindexable. You can help a thousand people with the same issue on Discord, but since no (web) search engine is returning any of your past discussions, people will keep coming.
The SFML website and forum are getting around 750’000 impressions and around 97’000 clicks per month from Google searches alone. This isn’t a small number, but I’m convinced if all our discussions were on a platform shared with millions of other topics, the search engines would have a harder time to point you in the right direction and chances are, that not all the old posts would be getting indexed.
The last point, with which I don’t have much experience with, is accessibility of information. A lot of “modern” solutions make it harder or impossible to consume your content in alternative ways. Which platforms work fine without JavaScript running? How easy is the website to consume with a screen reader? But most importantly, what can you do about it, when you’re just a customer of a platform?
The Solutions?
While some believe the main solution is to avoid anything that isn’t FOSS itself, I see things much more as a balancing act. Be at the places where people expect you to be (e.g. GitHub), but if you do care about information retention or governance, have solutions in place to ensure those factors.
I’m a big fan of old school forums, so I might be biased when telling you to make use of them again, instead of fully switching to some platform. My biggest issue with forums, is that there aren’t many good, “modern” forum software implementations. SFML uses Simple Machines Forum, which haunts me, whenever I take a peek at their code or think about the plugin system, that is just applying diff patches to the core PHP files. If you know any modern forum software (which isn’t Discourse), please let me know!
My recommendation is to keep your most important information on your own “platform”. Don’t make your “landing page” just a Discord link or a GitHub repository. You’ll face a massive issue, if you ever chose to move somewhere else or the platform “uninvites” you.
The Conclusion
Despite the Internet Archive trying to prevent a lot of information loss, I’m predicting that in a decade or so, we’ll be noticing a huge gap in the retained information, as so much of it was lost on inaccessible platforms. You can help alleviate this by not moving all your projects to information black holes.
Additional discussions:
One thought on “Black Holes of Information”