Gemini’s standard is missing several important means of semantic tagging that screenreaders need to properly deal with content. I feel that if we are going to replace HTML, we still need to ensure that the disabled are first-class citizens from day one of whatever new standard we use.
Yes, I agree accessibility is a vital concern that needs to be fully addressed before Gemini is finalized. Here are some posts I found using this Gemini search engine: gemini://geminispace.info/
HTML already isn't meaningfully accessible. Gemini will still be an upgrade, because there's much less to spam screenreaders with, by virtue of being effectively just a hyperlinked document layer.
I think accessibility is critically-important, too, but acting like it's something required for success (however you define success) is denying the existence of the platform Gemini's already trying to supplant.
"As of 2012, 4.2 million Americans aged 40 years and older suffer from uncorrectable vision impairment, out of which 1.02 million who are blind; this number is predicted to more than double by 2050 to 8.96 million due to the increasing epidemics of diabetes and other chronic diseases and our rapidly aging U.S. population."
"Approximately 6.8% of children younger than 18 years in the United States have a diagnosed eye and vision condition. Nearly 3 percent of children younger than 18 years are blind or visually impaired, defined as having trouble seeing even when wearing glasses or contact lenses."
Any definition of success that intentionally excludes tens of millions of users (even people who aren't completely blind may need the aid of a screen-reader) is a very questionable definition.
The Web already does that. So does...all of computing! Hacker News is an English site. Billions of people do not speak English! Hacker News, therefore, has not succeeded in its goal to...be a place for sparking intellectual curiosity in English-speaking populations?
Anything is discriminatory somehow, and "We have basically a plain-text format with links" is millions of miles more accessible than HTML, computers in general, and pretty much everything in the world.
Plain HTML has a million attributes, isn't actually what any website is made of (this one even has spacer gifs), and isn't actually used in a semantic fashion by anyone (find me a page using <summary> and I'll show you a page that was sued into using it).
Gemini has two elements (link, text), and is much more accessible.
I love HTML, and I write my own sites to be accessible. I've used a screenreader, though, and can definitively say: every bit of computing sucks for the blind.
Because of the highly nested structure of HTML, it is possible to crash (from experience, with all the big names) your screenreader if there are a large number of elements, especially if there are a large number of sub-trees, like you often find in technical documentation.
HTML is accessible in theory... But in the real world, it can still be a terrible experience, unless the reader is highly optimised for streaming the HTML (not always possible), or the document is simple.
I think this is an interesting point. Now that nearly everyone uses the Internet, the bar for accessibility and the accompanying workload for implementing it is considerably higher than what it was in the 90's.
There probably weren't as many accessibility advocates back then. But accessibility is still crucial to have from a modern standpoint.
Actually, back then html was just a document format. Something to read. Alt tags were the biggest issue. Flash and js brought hidden controls and unannounced behavior trying to reproduce the native desktop controls that had accessibility build in.
For example, Gemini lacks language markup so that you can tell the screenreader to pronounce a quoted foreign word(s) in the correct language instead giving it a mangled mispronunciation in the document's default language. In HTML this is done by using span or div tags with lang="XX".
