Gab.com
Gab.com started out as a simple, home built, twitter clone in 2017. I jumped on it as soon as it began as @bol (my 3 letter username should indicate how early I was) but never really worked hard on it.
Last year in July they dumped their own proprietary code and switched to the open source Mastadon code. This allows anyone (who has a strong background in running servers and keeping complex software working) to run their own social network loosely along the lines of Twitter.
Each instance of Mastadon is able to "federate" to push and pull content to and from other instances of Mastadon.1.
Gab's founder, Andrew Torba, has always been scathing about using blockchain tech as the back end to a social network. I'd show you some of his posts, but Gab doesn't have anything like a functioning search engine so you'll have to take my word for it.
It would appear Gab has also ended its Federation experiment. I've heard the same thing from other administrators of Mastadon sites, federation is complicated and resource heavy. The only other instance of Mastadon I've used with any regularity is No Agenda Social and it is from there that I'm encouraging people to check out the No Agenda Producers Community here on Hive. The only reason I found that Gab had killed federation, was because I noticed none of the Gab posts I follow in No Agenda Social were showing up any more.
Leaving Star Fleet?
A few days ago Rob Colbert, Gab CTO, posted this poll:
Clearly, not long after, he cut the federation with the rest of the Mastadon world and gave this explanation:
The routes were deliberately disabled by decision and our push and pull workers are being killed.
Gab is 99.3% insular as measured by independent and unbiased tools. Literally 0.7% of our community federates. So, just like nothing was ever done to circumvent any blocks or cause undue stress to any operators, we're just quietly detaching and very much going our own way in support of our community.
The ActivityPub model represents an amount of load and cost that is disproportionate to the Gab community's use and realized benefits of it. In terms of value for value for the Gab community, there is no clear way to justify further participation in the ActivityPub/Fediverse experiment. Additionally, our community is not pressing us to invest in it at a loss because they love it's vision and want to build in that direction.
Gab is, therefore, spending too many dollars from sources who don't use Federation services at all or care about them in any way. That is not fair to the people who spent those dollars hoping for a better or more stable system that they want or need to use.
The Fediverse, for Gabbers, is provably not part of the value equation. They just don't care. Gabbers are tolerant, but they will get angry at me for wasting their dollars on things that they don't use or care about. Because that is just common sense so the question would be, "Dude, wtf?" and I would understand.
The ActivityPub Fediverse didn't make the cut from a business perspective. So, Gab has chosen to end it's investment in further ActivityPub/Fediverse development. It isn't that we couldn't monetize the Fediverse, exploit it, or squeeze the dollars that don't exist out of it. We'd be happy to put resources on things Gabbers use.
It's literally just that no Gabber cared.
So Webfinger got stuffed into an outbox and was tossed into the great .well-known by people who were basically over all of that.
Gab is just another Web 2.0 site
Gab is just another Web 2.0 project trying to get its users to pony up some cash with subscriptions for Gab Pro while trying to compete with the Goliaths who give their wares for free whilst selling their users to any all who'll pay for the data.
Other than running on their own fork of an open source project, Gab is nothing more than a repository of their users posts and data and the connection they have to their users.
I wish Gab all the best, and perhaps if JPBLiberty succeeds in damaging Twitter with our #CryptoClassAction, they'll benefit, but going head to head with Twitter using exactly the same centralised model and without the benefit of selling their users out to the highest bidder, doesn't seem a great path.
Footnotes
1: I was involved in a startup here in Israel which I've now parted with. It is still working, quietly, on another method for inter-server communication so the same app running on multiple servers can draw from multiple external instances.
