Most transfers of data between computers involve a client (the thing that requests the transfer) and a server (the thing that responds to the request.) A download is a transfer from the server to the client, while an upload is a transfer from the client to the server.
Of course, there are grey areas where some kind of client/server relationship has been established, and yet the server requests a transfer from the client. You might still call that an upload if the overall client vs. server status is clear. And then there are more equal peer-to-peer relationships where the client/server roles switch all the time.
It depends on what ‘device you’re controlling’ — your words. If the device you’re controlling is a cloud server and you’re copying a file from a remote device (again, your words, ‘remote device’) and that remote device is your friend’s office laptop, then you’re uploading data to the cloud server.
An example of the hierarchy answer: Lego Mindstorms robots have a “smart brick” that’s part of the robot, and which stores and runs the program. You typically create the program on a computer, and then transfer it onto the brick via a USB cable. Even though the user is sitting at the keyboard, and initiates the transfer using the computer rather than using the brick, this transfer process is still called “downloading”.
Not quite. Space > cloud. Any transfer of data to a satellite or space probe, even if it’s from a cloud source, is still an “upload”. This isn’t entirely consistent with the reasons for the other hierarchical rankings (a spacecraft is usually a “small or passive device”), and is probably due to contamination from the common usage of “up” and “down”.
So the three industrial control powerhouses that have industrial controllers that together control the manufacturing machines of just about everything have been wrong for the past four decades because some guy on the internet says so. All three have software I have used and from the computer I am at, I upload from the controller when I want a copy of its current program and I download any changes to it. The controller will download the data to any of its remote devices such as motor controls and input bases.
ETA: In my experience download is send data and upload is receive. This being among peer units. I will grant the cloud hierarchy explained earlier.
Indeed. I’m somewhat bemused by all these alternative interpretations being offered.
That said, I’ve worked in tech support and I learned to resist the urge to correct people who say things like “So can I download these pictures to my grandson?” because… what’s the point, really?
I use it from the perspective of the initiator, whether it be a human or an automated process. So if a satellite had a daily task on its computers to initiate a file transfer from another system I would call it a download whereas if the ground system initiated it I would call it an upload. If you buy MP3s and then actively transfer the files to your PC you are downloading. If the cloud server somehow had a process that automatically touched your PC and re-sent your previously purchased files it would be uploading them. When you send video files to YouTube you are uploading them.
I guess it could get theoretically tricky if there are multiple layers involved, such as if a person initiated a process to send several files to a remote server, but the files are put in a queue which is actively managed by the remote server, in which case the entire process is an “upload” but I wouldn’t quibble if each file action were called a “download” since the transfer would be initiated at the destination.
I can’t think of a single website that distributes files that has a button or link to do so that says “Upload”.
They all say either “Download” or something else such as the file name or “get file”.
On the other hand, sites that want you to send a file, such as job sites, say things like “Upload Resume”.
Read the manual, or help files. Seriously.
In the industrial world, when communication to devices, it is about a 50/50 split who defines a transfer TO a device as a download, and who defines the same transfer TO a device as in upload.
When dealing with different software packages, it can become a rather spectacular disaster if you get the direction wrong.
