Automatic ranged queries when server allows

[akshaysharma-tilda]

We are loading document from google cloud and I have following conditions checked:

• PDF is linearized
• Accept-Ranges is being sent as bytes
• Accept-Ranges header is exposed, JS can access it.

According to Gemini, GPT and Sonnet, if the conditions meet, React-PDF should automatically (since it uses pdf.js and that does it apparently) be making range requests and loading doc efficiently with first page being shown quickly, although, it is making a single request for entire 10 mb file.

[wojtekmaj]

Hmmm, hard to say what might be the reason here; the logic is entirely on PDF.js's side, we're doing no changes to it. It sometimes works, and sometimes does not, and I can't pinpoint the exact conditions for it to work.