Ah, good point
Objects have an id
which is an IRI, and the object needs to be fetchable through that. It’s possible you may need some redirects, but those should already be in the code (see https://akkoma.dev/AkkomaGang/akkoma/src/branch/develop/lib/pleroma/web/router.ex#L738). If not, you can still see if you can “translate” from one to the other at the level of your reverse proxy.
There are other IRI’s as well like inbox and follower collection, but at first glance, the format is the same on akkoma and mastodon, so shouldn’t give problems.
To give an example of what I mean; This is a simplified example of an object in Akkoma.
{
"@context": "https://www.w3.org/ns/activitystreams",
"id": "https://ilja.space/objects/885a03c3-db2c-4d58-b1cf-6b74a1582907",
"actor": "https://ilja.space/users/ilja",
"to": "https://www.w3.org/ns/activitystreams#Public",
"content": "Blablabla",
"type": "Note"
}
You can fetch the actual object using it’s id with curl -L 'https://ilja.space/objects/885a03c3-db2c-4d58-b1cf-6b74a1582907' -H 'Accept: application/ld+json; profile="https://www.w3.org/ns/activitystreams"'
(Note that there may be extra authentication requirements, but generally you should be able to fetch public posts and actors and stuff like this).
It also works with redirects. For example, if you would check this id in the browser, it will redirect to a different url. But if you try to fetch through that url, then that will also work because you follow redirects.
Basically, try it on a copy of your instance, see if you can still fetch everything that looks like an url correctly through that curl request, and if that works, it should be fine (but again, no promises, do it at your own risk, and all that
).