This Time Self-Hosted
dark mode light mode Search

So You Want A New Web?

Obvious preamble: what I’m about to write about is my own personal viewpoint, and does not reflect that of my current, past, or future employers. I do work for a company that develops and provides services that are affected by the kind of movements that I’m going to be mentioning herein, but I do not work on any relevant product, and my views on this topic have been widely shared well before this current employment started.

I’m convinced that I’m a purveyor of mild takes, either written before events unfold or that are written some time after but set to publish weeks if not months later. I like this concept because, while not always being topical, and rarely being flashy enough for it to matter in the grand scheme of things, I don’t particularly enjoy the current “hot takes” society. Many things are, at the end of the day, nuanced, and in need for analysis and discussion, not immediate response.

So in this spirit, I’m writing this post in December, while the amount of discussion (if we can call it that way) around so-called “web3” has been going on for weeks, to the point that I decided to mute the keywords because even the random dunk on that crowd was getting boring to read. This post is not about web3. I think that Jürgen has covered that very well, and I don’t feel I have much to add to this beside my posts about Diaspora.

But, in a similar spirit to what I said about IoT, I thought I would at least attempt to make an argument on how to salvage the discussion of “a new web”, fully conscious that my last and current employers are both part of the incumbent that those wanting a new web would like to see shut down, or at least significantly resized.

First of all, the new web being decentralized is something that comes up over and over and over again. It’s not new, and the current web3 mania appears to mostly align with the need for investors from making money off the trend. We had a number of decentralised services of the years, and some still exist. I noted above about Diaspora, which you may remember as being touted as the decentralised Facebook. We had StatusNet before Mastodon, that was the distributed Twitter. We had (and some still have) Jabber/XMPP for instant messaging.

Why did none of these catch on? Well I can imagine there’s a lot of different smaller reasons, but in general, the core of it is that it’s much more convenient for users of a network if the services are actually centralised. This is true both for FLOSS solutions (with the added problem of self-hosting I noted previously), but also for proprietary solutions! Think about how many memes have been made about Netflix, Disney+, Prime Video, and the other streaming platforms splitting the supply of material, and bringing in a new golden age for piracy, after Netflix nearly destroyed it, by being just so convenient.

And to make it clear that the problem here is not just about cost (it is a factor, but not the only factor!) let’s take a different example: video games stores. Even without going through the topic of segmentation caused by consoles, we can notice how PC gaming became interestingly complicated once Steam received even the minimal of competition from Epic with their Epic Game Store. Steam was not even the only store before that, and indeed I already had at least three (if not four) game stores installed in my computer for many years: Steam, GOG, EA’s Origin, and at times Uplay (for The Settlers VII, but I disliked that game so much I gave up.)

The games didn’t change their price range, and indeed if anything the competition from having multiple stores to provide the audience to developers might have helped lowering fees, and prices in general. But gamers actually complained about having to separate their libraries across different services! And to be honest, there’s a few minor downsides, namely when using Big Picture mode in Steam, not being able to see the Epic games does annoy me a bit, but I’m sure we can fix that if we want… though it’s actually a good example of where decentralizing does cause friction.

There is an argument to be made that I’m looking at semi-centralized platforms, though. After all you cannot go and start your own game store or launcher, and share the data across all of those. But honestly I don’t think we will ever get to the point where every user of a platform will be running their own node or whatever you want to call it. That would be in my opinion nightmareish, as most people are not interested in becoming system administrators, so throwing a lot of different nodes onto the Internet would be just allowing more DDoS to be performed with even lower costs.

And together with DDoS will come spam, as (and I repeat myself) I noted back in the Diaspora discussions. Spam is not a solved problem for closed platforms and systems, but open systems just tend to make its management next to impossible. That’s the main reason why I stopped using back in the days: the amount of replies that were spam was increasingly drowning the actual useful replies, and at that point it made no sense for me to be on a service that couldn’t keep their users safe.

So instead, what would be my hope to get a better web? Data portability. And that’s not an easy thing to do, and not a thing that would be welcome by many of the people discussing this out there already. Data portability means the ability to take all of your data from one service and upload it to a different service, and keep most of the data accessible. This can only be done with the cooperation of both the source service and the destination service. The best option for data portability would also include the ability to keep all of the previous information valid and linked, and I honestly doubt that’s going to be happening ever, as most services that I know I used in the past would barely let you keep the validity of the URLs for a short amount of time.

But even before that, it would be awesome if we had more tools available for creators first, so that the consumers can rely on these. As I wrote before both WordPress and Hugo, despite being started as blog engines, ended up straying from that path and are nowadays used for a lot of other reasons, most of them commercial in nature. Nothing wrong with it per se, it’s a great way for development team to be funded, but it does make for a sad situation when you’re trying to look for something that is designed with blogging in mind — and it’s hard to argue that there’s no more blogging, when there are no more blogging engines!

Instead of chasing digital scarcity with bad receipts, I would love to be able to support creators in a more direct way. I mean we already have Patreon, but that has its own limitations (besides the already discussed ones.) For instance there’s no way for you to receive updates from your supported creators in an application like NewsBlur — you can try to use the newsletter-to-feed support, but I found it fairly unreliable for Patreon messages. What I would love instead would be for NewsBlur to be able to authenticate to Patreon (or vice-versa) and have a “private feed” available for me. I’d be willing to pay extra to support this kind of complication in NewsBlur!

Most importantly, I think that if you expect creators to go and run web applications themselves, those apps should be able to run in semi-managed environments too. Gandi Simple Hosting definitely made it a lot less miserable for me to run WordPress, and I think that having applications that can be packaged and self-updated would definitely help reducing the cost of self-hosting them. Unfortunately, I have a feeling that the crowd currently working on it (even excluding the outright scammers) will have more feelings than experience with Cloud, and will end up complicating their choice even further.

Anyway, we shall see if anyone, anywhere, any time soon will start actually creating the kind of New Web that users, consumers and creators alike, want — rather than just some geek wet dream of complexity and investor idea of milking more people for more money.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.