[JimmXinu]

I don't recall hearing of archive.transformativeworks.org before.

The current adapter code allows several domains based on what was allowed by their ssl cert at the time, but always normalizes to archiveofourown.org:

    @classmethod
    def getAcceptDomains(cls):
        return ['archiveofourown.org',
                'archiveofourown.com',
                'archiveofourown.net',
                'download.archiveofourown.org',
                'download.archiveofourown.com',
                'download.archiveofourown.net',
                'ao3.org',
                ]

But you appear to be correct that it is an official mirror judging by the FAQ What URLs can I use to access the Archive?

The question is whether those different domains point to different systems, or are just aliases to the same system. DNS does give different IPs for archiveofourown.org vs archive.transformativeworks.org, but given today's cloud systems, how much does that mean?

Another problem: I was not able to login on archive.transformativeworks.org -- it still tried to log me in on the usual archiveofourown.org domain. So even if it works better, you can't use it for stories/features requiring login.

Yet another issue: the list of most recently updated stories I get back from the two domains don't match. archive.transformativeworks.org appears to be several stories out of date. That argues against it just being an alias, but how out of date?

However, this is worth trying out.

(Very) Test Version

Attached is a test version that adds archive.transformativeworks.org as if it were a new, unrelated site.

I do not expect it to work with logging in, and it does not use your existing AO3 settings.

This is purely so people can test and report if it's less error prone for them.

If it is, I will look into how using archive.transformativeworks.org might be integrated into the AO3 adapter.