Outernet: Humanity's Public Library | Hacker News Reader

Outernet: Humanity's Public Library

outernet.is

96 pointsdsr1226 comments

Love the idea; don't like that they ultimately own the satellite.

I would like to see a satellite similar to this, storing and broadcasting a queue of information, but that has a crypto-democractic method of submitting data to broadcast. Requests for data to broadcast submitted to the satellite must be accompanied by a Proof-of-Work. The required level of POW can be dynamically adjusted by the satellite itself based on recent demand.

In other words, the satellite is fully autonomous, with no central control. Anyone can submit data to be broadcast. They just have to "pay" for it.

This solves the problem of who owns the satellite (everyone does), but does still mean money can buy broadcast power. Though people can build systems for helping to "crowdsourcing" POW solutions for desired broadcasts.

NOTE: I haven't read up much on cubesat, and just satellite technology in general, so I don't know how possible this is (in particular, how to manage transmission to the satellite and whether cheap kickstarter sats/cubesats can broadcast data on their own, or are linked to some other satellite).

EDIT: Also, this has the same problem similar systems suffer from: potential to broadcast illegal data. No great solution to that problem other than revising the laws.

I'll note that your system still allows for the actual owner of the satellite to denial-of-service anything they want to.

You are correct, which is why we eventually want our own constellation.

Remember that much of wikipedia (and the internet in general) are considered 'illegal data' in some countries.

In this sense -- I believe that the idea here is to indeed broadcast illegal data. I find this project very interesting because it exploits the fact that we still have national laws in an increasingly international society. Because let's face it, whatever data they might broadcast it will likely be illegal somewhere.

It's effectively pirate radio for the internet age - broadcast outside the domain of law and thus avoid censorship.

That's exactly how I pitched it to our investor.

philosophically and technically very exciting, however I'm worried about the business model:

"Do you have content to share with the whole world? Skip the Queue, pay a small premium, and flag your content as Priority."

https://www.outernet.is/en/broadcast

They've "baked in" a corrupting factor to the content: money. IMHO this is what soils services like Facebook, Twitter, Angies list, etc. Places like archive.org and wikipedia are flat donations (I believe).

I'm open to other ideas for revenue. We have a hard stop on the amount of capacity that we'll sell--no more than 25%. We have several pilots starting next year where the sponsored content is educational material, courseware and such.

Information is inherently a bit of an odd duck economically speaking. The fact that you're targeting the less-developed 6/7 of the world, generally, means that you're also targeting an other-than-lucrative advertising market. There's also the question of just who it is that you might attract as content promoters -- EDUs is one thing, but if it turns out that you're mostly of interest to those who've got unrest to sell -- whether it's OECD foreign / intelligence offices or third-world mercenaries, then there's a bit of an issue.

Who you accept money from matters in some ways more than what's being sold -- I'd suggest your team sit down and draw up a few short lists of groups or sources whose money you don't consider green.

A few ideas do come to mind:

1. Classifieds. You could do Craigslist to the World. Even a small fee might prove lucrative. All the better if you could target the ads to the regions where they're of most interest (geographic segregation).

2. An infrastructure for other organizations to tap into. It could be that it's the content-distribution network itself that's of value. If you've got a one-way comms channel that could be used by organizations for their own use -- encrypted transmissions (if desired) that only their own users could pick up, this could be a valuable service. Of course, there's still the chance that it's intelligence and/or terror organizations which are interested in your services. The challenge here is that if this is your paying work, it might consume your interest over the public-good components. I suspect there'a a fairly wide class of NGOs and possibly commercial interests who might be interested in such capabilities.

3. Partnering with regional governments. If it turns out that you're the best way to get information through, then (for reasonably benign governments), carrying some of their content might work.

You've got a number of partnerships with national broadcasters (I saw DW on your partners list). Carrying content from such organizations for payment might well work.

The other factor is what your own overhead is. If sufficiently low, you don't need a lot of monetization to be successful. Just enough to cover expenses.

All that said: I really like this idea, it's similar in regards to some I had years back but never really pursued.

A couple of ideas. On revenue: I think dredmorbius really covered it well. And I want to emphasize the infrastructure part. What about selling hardware that focused on the receiving endpoint? The target demographics probably have literacy/language issues with the content, so providing dedicated hardware to read text (instead of broadcasting actual audio, saving bandwidth), translate it, and do text-to-speech could be an example of some additional products to sell. Sell an app version too for middle-wealth countries. Selling infrastructure items doesn't create that conflict-of-interest in my mind that paying for content does.

Jumping off of the classifieds idea. What if you charged a flat fee, the same flat fee, for every submission regardless of what it was? If submitters truly have altruistic intentions, then this small flat fee is a small hurdle for the little guys, and probably a dead end for the very wealthy to try and overpopulate the system with whatever they want. So that the pressure is not on you to arbitrarily regulate the income sources. Decouple money from content is my main theme here.

On content: I've never liked one-dimensional voting systems that are simply popularity contests. It's not that hard for a organized group to up-vote their propaganda (one group of many) while the altruistic content submitters are all left by the wayside (many groups of one). What I'd like to see is voting be not for submission, but for classification of content. It could be instructional, news, public service, propaganda, entertainment, etc. Then let your paid submitters be the ones to vote separately on ratios of these categories of content that finally get broadcasted. And only content that has been reviewed and voted for past a certain threshold gets proposed for submission. Right now, looking at the stream, I am already biased AGAINST stuff that is paid. Disclosure of that fact is great obviously, but what if a private org really is producing great content? I'll probably never check it out. A community filtering/classification of the content would put my mind at ease as the same standards of review would apply for paid and free content.

This article has some information about how it works: http://www.latimes.com/business/technology/la-fi-g-outernet-...

It answered some of my questions but doesn't say about how are they planning to select the initial content, which needs to be a lot of data to get people interested. It just says "you can send ideas to our facebook wall... or send us a paper letter".

It would be interesting if Outernet were to transmit the Bitcoin blockchain. The maximum data transfer rate of the blockchain is artificially capped at something like 13kbps right now (1x1MB block per 600 seconds), so it would probably be feasible to transmit it. It might allow for some novel economic activity in very poor areas without internet. It would obviously be receive-only, but that could still be useful. It might provide an equivalent to pre-internet banking in areas too poor to be served by banks. ("Pre-internet banking" means you have to take a trip to the bank to send money, but you can call up and ask your balance to make sure you got paid.)

An SMS based service for checking the balance of a Bitcoin address seems much more useful, since people in very poor places are much more likely to have cheap cellphones than devices capable of processing the whole blockchain (and an SDR or so to read the signal).

SMS services aren't secure. Lightweight SPV servers (well within the resource constraints of ultra-low-end smartphones, which will see increased adoption in very poor areas over the next few years) are. And the whole point of outernet is to reach areas that might not even have SMS capability.

Actually, that IS planned.

The general idea sounds cool. For the most part those people who are in the dark and don't have internet access, don't have a wifi enabled device and the luxury of having/maintaining a satellite receiver thingi to receive someones' offline sub-internet + "sponsored" content. I hope google's balon project takes off sooner.

I've worked in many places. There are plenty of off-brand dumbphones in India with wifi and bluetooth. In many places I've been, the poorest people still have two things: a cellphone and access to satellite tv. A satellite receiver in Ghana costs $50. And that's also how much I paid for one in the rural Morocco.

Google Loon is not going to be a free service; it's paid internet--which is important, but $25 a month is still out of reach for many.

I am really excited about this idea, especially the disaster information aspect. I hope they build a decent system for prioritising what information should be distributed. The connection between an sms request and receiving the most relevant information is going to be tricky, perhaps something like google's i'm feeling lucky might work.

Interesting idea, presumably you could just download the broadcast as it came across and then offer it up as a locally accessible store of data. Even better if the outernet folks provide a means to auto-insert the download into a structure that allows folks to search, index, and correlate it.

Your input would be invaluable. Feel free to start a thread at https://discuss.outernet.is. You can see a very early alpha of the user experience here: http://librarian.outernet.is/en

Can someone try to explain the metaphor of a public library in this case for me? It seems more like a radio broadcast; I feel like I'm missing something.

What I understand of the project is that the Lanterns(the data receivers, which you can make yourself apparently http://outernet-project.github.io/orx-install/) store data received from the satellites on a daily basis. The information, which is voted by global citizens, is stored on it wherever you are on Earth, so you can access to sites like Wikipedia even if you are lost in the wood, for instance. But the goal of the project is not to make money but to give the opportunity to population of countries like China and North Korea (and many others) to access information otherwise censored in their country. So yeah, it's similar to a global radio broadcast.

> But the goal of the project is not to make money but to give the opportunity to population of countries like China and North Korea (and many others) to access information otherwise censored in their country.

This is a laudable goal, but if this became a threat then wouldn't their governments just jam the signal?

For example, China already does this with shortwave radio: http://en.wikipedia.org/wiki/Radio_jamming_in_China

As this could completely undermine the project in large regions of the world, it seems like a huge operational risk.

So how would the requests for information be routed up to the satellite? It's not like it can continuously beam down ALL the information all the time.

We could do 10 GB per hour, easily. And that's just one transponder. A twin tuner for a set top box costs $8. To a slightly larger dish and a bit less error correction, a receiver could pull down 1 TB a day.

Also see https://projectmeshnet.org/

Reply on news.ycombinator.com