Everything is labeled.
Also related to the Lemmy software: good support of the Markdown language means everyone can add alt-text to images (which wasn’t possible on Reddit, Reddit was by design not blind friendly).
Absolutely. While alt text for image posts is still necessary, you can add it in images in text posts.
For transcriptions, do you think it would be a best practice for users to add them to the alt text, the post body, or post comments? I’m guessing alt text would be most salient for screen reader ergonomics, but not as widely noticeable for errors, bias, or omissions, like with titles. Body text would be more commonly viewed, and thus held to more scrutiny and correction. Comment text would be easiest to track corrections or revisions on transcriptions, but not as discoverable if buried in the comments.
Given the facilities we have here, I’d go with short alt text and a longer description in the body. That way, screen reader users know what’s in the image, but everybody gets the explanation and context.