The sampling methodology described here illustrates one of the most frustrating things about everybody's attempts to understand #Mastodon. People focus on bigger instances with more active users.
I wouldn't be at all surprised if there were instances filtered out for which:
But they don't matter, because they're not big.
https://mastodon.social/@AltTextHealthCheck/113580107673787668
=> More informations about this toot | More toots from felyashono@disabled.social
@AltTextHealthCheck
For example, I just checked the local timeline here at disabled.social. Over the last roughly 48 hours, 50 out of 54 images posted had alt text. That's 93%, which is higher than beige.party's average. But we're not on the list.
=> More informations about this toot | More toots from felyashono@disabled.social This content has been proxied by September (ba2dc).Proxy Information
text/gemini