IMO the HTTP Range header (the thing that lets you resume downloads) was one of the best innovations back then. It let the client tell the server where to start the download from, and how much of the file to download. This means it also let you speed up downloads by downloading multiple pieces in parallel. I used to use a program called GetRight and loved it.
We still split up downloads like this today. These days, internet connections are often fast enough that you can’t reach full speed with a single connection, so speed tests and things like Steam will open multiple concurrent connections for their downloads to maximize download speed.
Fuck a bunch of 50000 pixel tall images
back in my day that took 3 hours to download
And someone would pick up the phone 30seconds from the end, and there was no resume!
IMO the HTTP
Rangeheader (the thing that lets you resume downloads) was one of the best innovations back then. It let the client tell the server where to start the download from, and how much of the file to download. This means it also let you speed up downloads by downloading multiple pieces in parallel. I used to use a program called GetRight and loved it.We still split up downloads like this today. These days, internet connections are often fast enough that you can’t reach full speed with a single connection, so speed tests and things like Steam will open multiple concurrent connections for their downloads to maximize download speed.
In my day it looks like shit