Free GETs aren't going to fly

Autonomi Network Token (incl (e)MAID)
41

What if the cost for getting public files was not free but pretty small, like insignificantly small, a tiny fraction of the PUT cost? You could stream all day and it would only cost a few pennies worth of Safecoins. Wouldn’t that fix the problem? I know I woudn’t mind, my vault would bring me enough to pay for my needs.

I know they said that browsing would be free but let’s just pretend for a moment they didn’t. Of course it would means you need an account to browse and it would slow adoption but how bad would it be in 10 years when the network is ubiquitous?

I wonder how much of a show stopper it is. Just pretend you never heard of the concept of getting files for free.

1 Like
42

I don’t think it is going to be that cheap. See: Vault routing burden

1 Like
43

for the first 10 years the farming of new coins will be in full force . so this isn’t a problem for the first 10 years, which in the next 10 years a hell of a lot of stuff is likely to happen

Any ideas that are thought of now are not needed for the next 10 years , because there isn’t any problem for the first 10 years, and 10 years will dramatically change perceptions and needs, especially being able to see the network function and see exactly where the problem areas might be. On a network that is intended to be constantly updated and improved as time goes by depending on the information it gets it seems like a long enough time to come up with solutions , once there is solid information regarding the subject

There might be the need for GETs to charge but in 10 years time most people who have joined the network will already have coins and it might be possible to set micro amounts of coins for GET’s and it could after how ever many years be changed to that system, or adapt one of hundreds of possible solutions in the years to come

but charging for GETs on the network to begin with, only for a problem that isn’t even proven a problem that might occur in 10 years time would completely restrict adoption of the network, a network that with growth becomes more powerful, secure and lowers entry level requirements according to the post you have just posted.

As mentioned before we have :

private data that will earn farmers zero . if we say that 10% of the network as an estimate is private data, which i think for a network offering the most secure option for data ever to come out is a pretty low end estimate.

considering private data is going to cost more than public data, if it costs 2 times more that public , you are going to have 20% of coins spent every day not farmed at all.

you also have vast amounts of information on the network that will earn farmers of that information less than the cost to put that information there. How many views of the piece of data until it has effectively broken equal? no one quite knows yet but it could be in the many 1000’s possibly . The more popular the network the more requests for data , so over a period of time, it could be possible that if 100,000 people are on the network a certain amount is paid, but when that doubles the farming reward is halved , twice as many people accessing data so earnings would remain roughly the same.

Not a clue how much of a percentage of data that will be because just like this discussion its highly speculative and not even launched yet so impossible to get any realistic numbers but a low end estimate of 5%

we would have a 25% of coins spent every day are coins that will likely contribute to the payment of farmers. (this doesn’t include break even data)

Kryders law will have somewhat of an effect on the data as well, don’t know right now if positive or bad that the rising average node size will do to the mechanics, pretty sure you could argue for both sides currently .

Popular data is cached , so there will be a maximum amount of rewards per day for any piece of content no matter how many views it gets, preventing any app or piece of data earning many multitudes of safecoin more than data that is far behind it in views.

this puts a huge cap on popular data earnings on a daily basis. How much? we don’t know because the network isn’t running yet.

There has been talks about planning the option in the future to delete data , which might incur a cost , over a long time period how much % of data a day would be paid for to be deleted? again unable to speculate accurately but could contribute a nice % of daily data contributing to the rewards paid to farmers

This currently is a 10 year problem which cannot be verified or detailed as an actual problem until the network is running for at least a while really . until the network is functional as it should be and analysis of the data can be done.

1 Like
44

I would just like to see a model that is spelled out…

I think people just want to do what they want done but I don’t know that anybody has done any math to figure out what is going to work or what will not.

Bandwidth is the culprit. It is not following Moore’s law… Hard disk space is not a big deal… It gets cheaper and cheaper all of the time.

But It will be VERY hard to change – So if you start it up with insolvent numbers 10 years down the road the network may just cease to work while you have a Bitcoin - BitcoinXT war go on.

I would suggest a fiscally conservative approach… You can always become more liberal later - but when you run out of money you are out of money and your ability to buy solvency is nil…

1 Like
45

Bandwidth is not following moore’s law but is increasing 50% year on year apparently . (this is for high end users to be frankly honest thou)

But the cost of bandwidth for a user of the safenetwork would become lower and lower the more people join , as already suggested in the discussion you posted.

If you start by charging , drastically reducing the numbers of people joining the network, then think “oh it looks fine, it works, we can make it free” only to find out later down the road that now it isn’t working because of unknown factors that happened because of the no longer charging for GETs and a huge influx of peoples then you gonna end up right back at square 1

If the network grows and people pay for data on a day by day basis , then solvency is not much of a problem , it means working out on a daily average how much the earning potential is, so the earning total of the day , divided by the farmers to make it brake even on a day to day basis , working out a payout amount that is equal to the amount paid for in data.

I would also be interested in seeing a model but at this point in time i think it is impossible for any realistic numbers or stats. it would be hugely guess work and unknown factors galore .

1 Like
46

Microsoft just yanked their free storage…

It is always the “abusers” that ruin it for everybody. n

I am not in favor of “charging for gets” per se – Just that there needs to be some mechanism in place to limit abuse…

If SAFE displaces netflix and bittorrent – which probably is one of it’s killer apps. I suspect that will kill its economic model. I don’t know so — But neither does anyone else.

Caching reduces bandwidth costs significantly but there still is a lot of traffic maintaining caches…

What I do know is that asymmetric markets rarely regulate. If one user can run up costs and you compensate for it by increasing the bill of somebody else – there is no incentive for the behavior that is running up the costs to stop, and there is more and more incentive for the behavior that pays the bills to happen less and less… In our case Downloaders may run up the cost and the system will pass the cost on to uploads, which does little to fix the problem.

You can keep raising the price of stamps, but that never seems to stem the tide of people not mailing stuff as much as they used to.

2 Likes
47

Microsoft yanked unlimited uploading (PUTs). These commercial cloud services let you pay for a particular storage limit, there are no limits on downloading (GETs) in their plans.

48

The one does nip the other in the bud though… You cannot download your DVD collection if they don’t provide you the space to store it…

They claimed that it a minortity of users that where hogging massive amounts of terrabytes…

With an automated autonomous system you need to plan the rules in advance – Because it is going to be very hard to change later…

I would just like to see a mathematical model before anybody gets too attached to anything. May be too late for that though…

49

Yes, and SAFE won’t let you store your DVD collection for free either. I still don’t get how you want to use this example to discredit free GETs.

50

Did the routing burden issue get resolved or not? That was my main issue.

51

I’m not sure if it’s going to be an issue, it depends on many variables. One might argue that here the network can balance itself. If bandwidth consumption for a vault is very high, few people can farm, which leads to storage capacity scarcity, which leads to very high PUT costs, which leads to less uploading of huge files like HD videos, which leads to less bandwidth consumption.

My analysis in that thread suggested that it will become exponentially less of an issue as the network grows in size (number of vaults). Maybe in the early days the network won’t be attractive for video uploading and such due to high upload costs. I think there are plenty of use cases for SAFE that don’t involve huge files (secure messaging, regular browsing, etc), so I don’t think it’d be that big a deal.

52

This is an assumption, and it’s clear that David Irvine doesn’t see it as hard to change features
later, so I don’t share your view on this. He sees the network as “evolving” and doesn’t see it as vital we get things right first time. Good to try, but not vital.

1 Like
53

I like the smaller uses better – SAFE cannot be beaten for security — I do think the most obvious app is the whole bittorent thing though. Hopefully PUT costs will be high enough to disuade that – Time shall tell.

If there is not high confidence in the economic model long term, folks will have trouble trusting SAFE with their data… I prod just because I think it is very important to think out with more than "maybe"s …

If you build in dynamic mathematics from the beginning it ought not be too hard… The network is in trouble if MaidSAFE can control too much though - It needs to be decentralized to prevent corruption…

Politically consensus is always hard to get, particularly when somebody is getting paid, and others think they ought to get paid less…

2 Likes
54

I think safenet is going to need torrenters being able to download big files to help differentiate itself and get early adoption. There’s too many free options at the moment that work just fine for securing small files, especially since most people don’t feel the need to go about hiding their data from NSA level threats, at least to the point of actually doing something about it.

55

@dyamanaka seems to think so too:

1 Like
56

Okay, I kind of dropped out of this debate (not just this topic, but all related to rewards) because none of them will reach a conclusion, but it’s a slow day so:

  • What measure is most useful is relative. It depends on what the network is supposed to do.
  • To me, SAFE seems positioned for read-only downloads (upload once, download many). Some other network may be targeting other workloads, but not this one.
  • MAID has been released, many parameters are already given and at least 50% of the code was written. Not much can be changed.
  • What to do: create service classes (at least two: free and paid) for both GETs and PUTs. Create a smart contract that can punish farmers who go down. Introduce deletion (maybe in v2) so that payments can be made for periods of N months or years.

Non-technical fanboys (especially “the age of abundance” crowd) only know what they want and are ignorant about what’s possible, given known constraints on time, money, etc. That’s why these discussions will never have a good outcome.

5 Likes
57

Technical thinking outside of the box is haaaard.

1 Like
58

Finally got a source, but it’s 50%, not 10%:

That had been bugging me for awhile.

59

Are you guys forgetting that coins are getting recycled all the time? It seemed like @Anders was very worried about what happens once all the coins are farmed. But that doesn’t often occur because they’re always getting recycled back to the network

2 Likes
60

I tend to agree with the OP.

Someone mentioned private bittorrent trackers as a remedy to the free-rider problem. The way they work is that a downloader has to upload as well as download, in some ratio determined by the tracker site’s owner.(1) At the beginning, a new downloader is allowed some small amount of download “for free” but after that he earns downloading volume according to how much uploading he has accumulated. This is all tracked by the fact that each torrent file (the small *.torrent table of IP’s of full and partial seeders of the content) is encrypted and is decrypted by the client only with a key negotiated at the time the torrent file is downloaded from the tracker site. Being a private site means that each client has a persistent identity (i.e., pseudonymity, although weak in the sense that it is linked to your IP, of course). This is all handled by the tracker site.

So the bittorrent clients of private tracking sites are indeed paying their way, and then some, by reseeding the files that they download, and the system works well in the examples I’ve seen, except when the authorities shut the tracker site down, or someone traces IPs and threatens to sue seeders (and thereby extorts some cash out of them).

No-one pays money to the trackers (their business model is paid advertising) but they pay by reseeding (hosting content) which is no burden to people who already have a computer and internet connection.

I’m sure there’s something for SAFEists to learn from that incentive structure, and it solves the problem described by the OP.

(1) Note that tracker sites don’t host content (i.e., files that downloaders want), but just the tiny torrent files.

2 Likes