How do I test my application with malformed messages?

Unfortunately, this is not yet supported out of the box.

Sorry about that, we are trying to find the resources to make it possible. (Have some? Get in touch!)

What we need to do is create a special-purpose App that can be run as at a Node, and that can be instructed to do various terrible things, such as emitting malformed messages. We tentatively call this App the “Hungarian Babelbox” (think of it as Douglas Adams' Babelfish using Monty Python’s Hungarian Phrasebook.)

It would essentially:

• Be a full implementation of the entire Fediverse protocols stack including commonly implemented variations. Such as, for example, being able to deal with Nodes (like Threads) that require signed messages for almost everything.
• Whose entire behavior is controllable and observable through an API on a very fine-grained level, such as individual data elements in JSON files. By way of comparison, an API such as the Mastodon API is very coarse. Even the ActivityPub C2S API is not sufficient.
• Can emit the same data in various variations as permitted by JSON-LD.
• Can behave invalidly (such as emitting HTTP errors when there were no errors).
• Can emit data that is malformed in various ways, from missing fields to invalid fields, invalid encodings, broken streams etc.

The still open issues for tests that required this are tagged with tag “protocol-level”.

It’s a bit of a project, so we can’t just quickly hack it together. We have the beginnings of its with the Imp that’s part of FediTest, but it can only perform client-side, such as for WebFinger tests, and it does not naturally extend to also act as a server because of the requirements for DNS hostnames and TLS certs for servers.