I have web application completely in UTF-8. I define <meta charset='utf-8' /> and all files are saved in UTF-8 encoding. When I want to create a blob on github, the library automatically encodes the string to UTF8 representation. So it is encoded twice. In my case it breaks everything. For example it transforms ° to °.

Why is this doing? I am not expert in this area but I assume that if the HTML page defines charset in meta element, web browser automatically converts strings when it sends some HTTP requests and also set up properly the HTTP headers such as Content-Type.

I think the automatic encoding should be removed or there should be option to turn it off? Wdyt?