Sunsetting CRTNet

CRTNet was an experiment to create an IRC network by and for the greater fediverse community.

Unfortunately, the project hasn't worked out in a desirable way. So, the network will be sunset effective March 15.

The rest of this post will be an examination of reasons why the project failed.

Software

CRTNet used what was believed to be a reasonably stable combination of UnrealIRCd and Atheme services. While there were many personal reasons I chose to use atheme for the project (like having previously written them), the choice of UnrealIRCd was largely a poor one.

A feature of CRTNet was integration with a bot called viera, which allowed for linking IRC services accounts to fediverse profiles. This feature depended on functional WHOX support, which UnrealIRCd did not provide. So, I found a module which provided WHOX support. All seemed well until a few months later when I observed UnrealIRCd was using 13 GB of RAM.

This lead us to discuss switching to another software, InspIRCd. Unfortunately, we standardized on using SPKIFP fingerprints to authenticate servers in the network with each other. Switching to InspIRCd meant abandoning SPKIFP support, so this proposal fizzled out. Meanwhile, my modified UnrealIRCd continues to consume large amounts of RAM.

From a technical perspective, the final nail in the coffin however is not software-related, but instead the result of IPv4 exhaustion: I needed to move the primary hub, but cannot due to being unable to coordinate access to the secondary hub. The reasons for that are complicated and not very interesting to discuss, so we will just leave it explained as a communications failure.

Cultural problems

The vision behind the project was to create a network for fediverse communities, much in the same way as Snoonet was started for reddit communities.

Unfortunately, what we have discovered is that creating such a network results in fediverse 'meta' drama and gossip being the primary topic of discussion. With this being the primary discussion topic, it provides no value to the userbase, so we were unable to gain traction regarding users.

Finally, structuring the network in an adhoc way instead of the way a traditional IRC network is structured (CRTNet had no shared responsibility for routing, etc) lead to the final set of technical problems.

Accordingly we are left with a network that has little value and little usage, and so I am sunsetting the project by terminating the primary hub on March 15.

To my knowledge, the main channel still on CRTNet is moving to their own server, irc.catgirl.network. I suggest giving that network a try instead.