url + &nocache=0 directly in browser is also an error.
I guess going over sockets (sockets write, read etc.) and away from JSON wouldn't help any, would it? It's not reading API then, but reading websites and understand their content, right?
After all, it's just 1 line of information that I want to read out. And sofar, we can read it, but only once.
The only other alternative is to go over visual basic script, where the constand reading of the url does work (I've tested it there first). The big problem with VBS is data exchange between it and mirc. That would only work over .txt files, which I wanted to avoid useing a direct methode from mirc.
