Problems with "access the internet"

So... I believe if you don't specify an XPath you can load a webpage just fine (albeit only smallish ones).

But I think I've had this error too - because HTML != XML. If someone wanted to contribute some changes that would ensure we used a parser in Gadgetbridge that handled HTML too that'd be great - but as I remember I didn't find an easy solution so I decided that some XPath support was better than none.