Why is downloading a file of unknown size so rare these days?

Does anyone remember around 10-12 years ago, downloading files (game demos etc) from web servers often used to be very frustrating because Internet Explorer (I was young and naive) sometimes didnt tell you the size of the file you were downloading. This could lead to frustrating waits because sometimes the web page didn't tell you the file size either! The download just ticked a long, it used to say "1.2meg download of unknown" or similar, and eventually it would magically complete! (If you modem didn't cut off before then)

Why does this not happen anymore? What 开发者_如何转开发has changed and how?

---

HTTP servers typically supply a "Content-Length" response header when serving files, which didn't always used to be the case. Even when it's not a static file being served, developers are gradually becoming more aware of the importance of setting HTTP headers correctly to work with HTTP proxies etc.

It also used to be that a major reason for using FTP instead of HTTP was to be able to restart downloads-- the Range/Content-Range headers have eliminated that, too.