It's been 14 years since https://xkcd.com/949/ came out and we still don't have a convenient way to send large files to other people. Will it happen in the next 14 years?
Resolution will be subjective, as per my judgement of whether it feels like "the email of file transfers" has been achieved. Some particular features that I will be looking for:
Easy to set up. It can take a few minutes, but must be doable by the average internet user, not just the technically-inclined. (The core infrastructure can be more complicated, but it must be easy for the average user. Like how setting up a mailserver is hard but signing up for gmail is easy.)
Easy to use. I shouldn't need to mess around with config files or troubleshoot errors; I should be able to send the file to something akin to an email address, or provide the recipient with a private key that lets them download the file, or some other method that Just Works™.
Widely adopted. I should be able to send/receive files to/from almost anyone who might want to do so, without needing to convince them to use a particular method or waiting for them to set it up. (Or at least it's a well-known enough protocol that people will likely be open to setting it up if they don't already have it, like e.g. Paypal.)
Cheap or free. It shouldn't cost significantly more than any other usage of my internet bandwidth.
Robust. A disruption in internet connection or a power-off of the computer should not require any manual intervention to resume the transfer, and it must be able to continue from where it left off.
Public domain. It should be a general internet standard, not something owned by a particular company.
Efficient. It should be peer-to-peer, not an upload to a central server followed by a separate download.
No maximum file size.
Secure. It can't be open to the whole internet like a Pastebin link, the files must be accessible to only the person I want to send them to. And they should be automatically encrypted in transit, via TLS or similar.
Supports streaming. I should be able to view the first part of a file while the rest of it is still downloading, at least if it's a file format that supports that.
Cross-platform. Should work on every major OS, including mobile devices.
Doesn't require a static IP. I should be able to take my laptop through a bunch of airport Wifi networks and continue sending and receiving files without any manual intervention.
A protocol that satisfies this market does not need to have all of the above points (note how SMTP fails a few), but it should have most.
1,000People are also trading
Maybe BitTorrent should count? I didn't realize it was as reliable as it apparently is. Next time I need to transfer a big file to someone non-technical maybe I'll try using it and see how it goes.
Another related effort in this direction is https://github.com/9001/copyparty/
@sblaplace I guess this might fail on some of the "simple/easy to use" criteria. But i'd say this is the closest so far
I think email attachments is the most common way to do this. And I guess it works for relatively small files.
I guess there must not be that much demand to send files larger than, say, a hundred megabytes. I guess for cases like that, you're essentially talking about video and platforms like YouTube seem to be the economic equilibrium.
What would https://sendfiles.dev/ need before it qualified? Just wider use? Maybe it doesnt meet your robustness criteria (that'd be an easy 3-day project)
@autumn IME most browsers have terrible download managers, will cancel the whole download and make me start over if I drop my internet connection for 5 seconds.
This UI is also bad, I'm trying to receive a file and clicking the button just does nothing, no indication of why it's failing. But that's easy to fix. This does seem promising but it requires browsers to get their act together, and has a single point of failure at the sendfiles.dev domain name.
@AshDorsey This looks great, but anything that needs the command line is going to be a non-starter for public adoption.
@IsaacKing Oh yeah, but the technology is 100% there and it's solely an adoption issue. I think by 2040 that will be solved.
@IsaacKing I suspect people are missing that it has to be peer-to-peer.
If I only read the title and saw the reference to the xkcd, I would come away thinking that cloud-hosting is allowed.
The peer-to-peer requirement changes everything but it's buried as a clarification for the efficiency bullet point.
@Wott It doesn't have to be, that's just a point in favor. It's difficult for a system to be robust long-term if it relies on one organization to keep it running.
@MingCat ....All of the bullet points I set out above that it violates? It is not widely adopted (I hadn't heard of it before today), it costs $13 a month for 2TB of storage and more for higher amounts, it's a private company, it's not peer-to-peer, and there's no indication whether it can gracefully handle interruptions. This appears to just be another cloud hosting service, not significantly different from Google Drive. And a rather sketchy one at that, using nonsense marketing speak to describe basic encryption as though it's some revolutionary technology that competitors don't have.