abhisuri97

Screen readers easily allow people whoa re visually impaired to cycle through html elements and accordingly perform normal mouse click actions on them.

While the idea of showing extent of engagement is great, I can't help but notice that articles which would only have 10-12 recommends have 200-300 claps mostly bc those claps come from friends of the author trying to promote the story. It makes it incredibly difficult as a reader to objectively ascertain what is actually more recommended by people. Additionally, as a reader, I have no idea what amount of claps is appropriate (eg is 10 claps excessive for a good article or is it not enough?). The recommend model was excellent, don't fix it if it aint broke.

For the threshold value, how is 0.25 determined to be the most optimal? is this value just specific to the datasets YOLO was trained on or is it something that is universal.

Also, this is really awesome!

ELI5 anyone? My basic understanding is if you use React and decide to sue FB for patent infringement down the line, FB can countersue(?) bc they revoke your grant to use the patent upon you suing them. But what _exactly_ constitutes using React? Of course using the library is one thing, but what about using some other library that installs React as some dependency? Does using JSX also count as using "React"? I just feel that what exactly constitutes using React is very vague esp considering there are many non-FB tools that use "React" (and associated technologies, eg the idea of Virtual DOM) in some form or another.

I actually made a chrome ext that uses im2txt to help blind/partially sighted individuals get scene descriptions for images. Perhaps that can make up for the fact that a lot of devs don't take the time to write alt tags for their images.

http://abhinavsuri.com/aat

A chrome extension to help visually impaired individuals see images http://abhinavsuri.com/aat

Does anyone know if this issue appears with React?

So, one thing I didn't quite address in the blog article was how tried to publicize this multiple times. (you can take a look back at our code and see our README was rather dry). While doing this probably won't work for everyone (hence: how we got...), I think all of the things I mention should be elements of code that is open sourced.