How the range request actually work?

[xingya1822]

https://github.com/mozilla/pdf.js/wiki/Frequently-Asked-Questions#range

according to this FAQ,
Why the very first HTTP request made to server seems loaded the whole file ?

And why all those following requests were made ? While I didn't begin to scroll down the pdf ?

[SamWiIIiams]

Notice that the response code is 200 for that first call. All of the other responses should be 206. The first call to the server requests the entire document and returns the metadata (contentLength) needed to issue the subsequent range requests. By default - in my experience at least - it will request the entire document and the chunks necessary to display the pages that are on screen. The initial request will continue streaming the entire document in the backend regardless of how many chunks you have requested. If you add disableStream to the request, it will no longer do this and instead only load the document in chunks.

[SamWiIIiams]

Additionally, if you don't want it to continue to load chunks you can use disableAutoFetch