Interstellar internet use cases, or "What if the internet was mail-order?"

Question

This question builds off Information Exchange In Space and How would interstellar internet work?, and specifically uses the communication system I developed here because of this.

In a summary of those links, imagine that FTL transportation is easier than FTL communication, and that there are no causality violations 'cause reasons. So, rather than really long wires running from star to star, you have analogs to the Pony Express and Carrier Pigeons. Voluntary runners (really, every ship that can spare the space) are automatically paid for having a box on their ship which handles all the storage, uploading, and downloading, and each planet has a "post office" system to handle routing and "layovers". Messages get from star to star reliably and cheaply, but slowly. Think days to weeks, like snail mail.

Note: this doesn't become the only network; the regular internet works just fine, but only on a planetary scale 'cause light speed.

What I'm looking for in this question is how people would use the internet if it worked like this. Personally, I think it would be a lot less like "browsing" and a lot more like "inter-library loan". Usenet, StackOverflow, and maybe modified versions of Facebook and Reddit would work, but IRC, Twitter, and real-time gaming would probably be limited to individual planets.

What do you think? What would the novel and/or utilitarian implications of the system be if this existed? What would its use look like to the end user? How would we go about doing all the myriad things we use the internet for today, but in space? How would this interact with each individual planetary internet?

Answer 1

I actually think the internet would work much like it does now. Each "major" site would have a presence in every system - and the messages being carried around would be used to synchronize the data in all of those sites.

So you wouldn't request a site on demand, instead every internet would be replicated to every other internet. Write a wikipedia update on mars and a few hours later it's showing on the earth wikipedia, a few days later most star systems have it. The less developed/connected your system is the longer the lag between updates.

The only noticeable delay would come if you wanted to directly send a message, get live updates, etc from events or people happening in another star system.

Data storage is cheap enough that everything would just be replicated. When you buy a hosting plan (30 credits, we host your website for a year! 50 credits we replicate to all class 3 and above star systems.) you would get replication options there.

Facebook updates from your pen pal in alpha centauri might come in a few days late, but for the most part you would not notice a difference.

Answer 2

In many ways, this is how the internet was once. We've become used to having enough bandwidth to stream video, but modem speed is MUCH slower.

It still worked, it's just a lot of the content was 'opt in' - your website would be low res thumbnails, and links to fetch content. This developed into 'download managers' where you'd queue up downloads.

In the current world though, you do still have bandwidth bottlenecks. We don't see it so much in the 'first world' but there are plenty of countries that have a limited number of fixed lines to the internet - and a storm or accident can knock them out.

One of the most effective ways of dealing with a limited 'external pipe' to a local networks is actually bit torrent - you make good use of the bandwidth, but then you 'sync' within the network.

I think this would be how your interstellar net would work. You'd have emails etc. but mostly use couriers which are basically functioning as bit torrent supernodes. It would act as a proxy cache and gateway. Your interstellar bandwidth might actually be quite good - as a wise man once said, don't underestimate the bandwidth of a lorry load of backup tapes.

So your couriers would act as 'torrent caches' where indexes were replicated on every run, and each pending request with no local mirror would also be retrieved. You'd 'queue up' your downloads, and just wait the round trip... but once someone had requested locally, then it'd be on the local net for anyone else to sync.

There'd be obvious things to fetch by default too - anything 'released' like albums, films, ebooks. I'd expect most 'publishers' would maintain local replication by default anyway.

You could also 'cache' the internet, a little like google cache does. Google does have a complete copy of 'the internet' - at least the top layer. http://www.websitemagazine.com/content/blogs/posts/archive/2014/07/22/do-you-know-how-big-the-internet-really-is-infographic.aspx

Even with the planetary internet, as it stands local caching of the 'top layer' can mean faster search results. I'd imagine the same would be done with your interstellar net, just with a longer indexing cycle.

Answer 3

I think that each planetary system work exactly like it does now, and there would be 'portals' or 'proxies' or 'gateways' to handle traffic from off system. This could include other planets in the same solar system and any passing ships that come through with 'news'.

Of course the Currier is going to need to have terabits of data capacity and there would need to be a Currier system to handle passing off data to ships going to different destinations with different stops etc.

There would have to be a lot of redundancy built in to handle the same 'packet' arriving days or weeks apart, or in different orders from actually sent.

Also the gateways would be there to protect the planets network from any extrasolar virus attacks. Cyber warfare could be very disruptive and useful if you plan any mischief. Could be as simple as messing up shipping orders for a planet to beat them to their own market.

Answer 4

For many uses, I agree with this answer that you would see local internets serving live content (as now) backed by interplanetary replication. The technology exists today to manage large data repositories over unreliable or highly-latent networks with eventual consistency; we'd see that applied to this problem.

But not all Internet use is web-site consumption with immediate results, and for those uses -- individual email, for example -- communication would slow down, just like it used to be here on Earth. Before (and then concurrent with) the Internet, which grew out of the ARPANet, was UUCP, the Unix-to-Unix Copy Protocol. This was used to move email and posts to Usenet newsgroups around the net. There was a backbone of nodes that sent and received updates to each other several times a day, and other nodes routed their messages to this backbone. If you were a couple hops away from the backbone it might take a couple days for your email to get delivered, and if you were farther out (digitally speaking) it could take close to a week. That's not ideal, but it was ok because people knew that and set their expectations accordingly. If you needed fast, synchronous communication with somebody, you picked up a phone. Instead of sending many short messages back and forth, you usually wrote longer messages covering several points -- think letters more than texts.

Behavior informs requirements for technology, but technological abilities also inform behavior. In your scenario people will have access to both modes -- fast but local (planet), and slow but universal (can go anywhere, eventually). We should expect to see local "pockets" of more-interactive communication that -- for sites like Stack Overflow and Reddit -- would eventually be replicated elsewhere. (Some modifications would need to be made, like how late-arriving answers interact with question closures, but those are matters for each site to work out.) We will probably also see a preference for the "in" sites; if Facebook is replicated across the universe and your blog isn't, you might move more of your participation to Facebook -- unless you specifically want to keep it local, which some do. If you have friends on other planets and want to stay in touch you'll still send email, adjusting your approach to the expected cycle time.

Answer 5

Look at the store-and-forward era of computer networking for inspiration: BBSs, Usenet, FidoNet, etc. You'll see email, Usenet-style message boards, file sharing, PEBM games, distributed reference works, and so on.