Topic: [SKY] Skycoin Launch Announcement - page 122. (Read 381579 times)

If anyone is not happy then just pack up and go
simple and easy

Distribute more for less or wait until the platform is fully finished. Surely the devs can see what a catastrophy selling only 2% at this stage would be.

Mildly interesting that the dev here seems to be top notch but the IPO parameters seem pretty crazy.  I really want to see the meshnet situation though.

98% of the coins under devs control that will be a main argument against this coin by many. I've seen cure coin bite the dust because of the same approach and it was also revolutionary and fighting for worthy cause...

+1

I am very disappointed with the IPO, only 2 per cent is ridiculous, I was looking forward to participate on this one, sorry I am out!

From what I have read so far, this coin can become the one to show crypto world how to not do an IPO. For most non tech people its too difficult to participate in it .

Everyone remember this:

This is not as much of an investment as it is more of helping a visionary and revolutionary project come to reality.

From what I have read so far, this coin can become the one to revolutionize crypto world and maybe even the monetary system.

So don't think too much as an invesment (although that is important in its own way), this is more of an idea that we can help become reality.

I'm looking forward to this project, I have yet to see another one like it.

Exactly what I was thinking. I would propose for at least 40% of coins for the IPO. 2% is so tiny, with a huge risk.

I mean every investor will care for the price. Will investors be able to decide when the other 98% will be given and in what amounts? Will there be a voting procedure or they will be watchers?

Please make an ann here in advance, before you officialy start the distribution process.

Well I decided to read some older pages on the thread.

I saw you decided to make some others ipo, or giveways during the next 5 years with the others coins not sold in the first IPO.

So, I think it's not a bad idea, by this way you can control inflation and make a sell only when the coin can afford it. But with a start of only 2% during the IPO, this leads us to a 200 000$ market cap for the total amount in circulation.

In fact, I would suggest you to sell in the first IPO at least 4% for this 200k$ market cap. That would give a price of 5 cents for a coin. 4% is not far of your idea and that would be more fair for us, early investors. Why ?

For 2 reasons :

- If you decide to make a giveway or an "IPO" of 0.1% of all coins, after the first IPO, with 2% firstly distribued, that would make a 5% inflation. That could be dangerous, and that's only 0.1%. Now imagine this 0.1% with first 4% or 8% ipo (or more), the inflation would be really less dangerous. In fact, if you decide to put on the market within the five years 50% of the total amount of coins, with a 2% start, that give us an inflation rate of 500% per year, and this is REALLY too much, you see ? Not even Bitcoin with mining is at this rate. The best start for me, for the IPO, would be at least 25% of all coins, for this reason (but still with the 200k$ market cap). I meant earlier 4% because this was not far of your first idea.

- The other reason is more simple. We are the early investors. This is us who will take all the risks when investing in your coin. Nobody knows skycoin, buying now is the most dangerous period for investing in. The market is not made, everything is only speculation. That's like buying some Bitcoins in 2010. Who would buy in 2010 a bitcoin for 200$? For 20$ ? Investing a coin before it's launch is the most dangerous thing an investor can do on this place. And... this should be greathly rewarded.

Anyway, thanks for reading (sorry for the maybe weird english btw )
I hope you'll answer to me. 

Is it a Pow or a Pos?
Can we mine the 98% of the coins left?

Yeah it seems expensive, the most important thing to know before we invest is where and how will be used the 98m coins. If the coins are not moved and not put on the market, we will have a really high market cap 

But if the coins are not used, why keep them on the blockchain ?

Hmm.

10 cents per 1 Skycoin it's ~ 45 000 BTC marketcap for project, will be #10 at coinmarketcap.

Bitshairs marketcap is ~ 100 000 BTC and Litecoin's is 300 000.

That means that the maximum Return on Investment is ~ 2x, may be 3x. Risky. To expansive.

Sadly, it seems the dev is running a monologue in this thread.

You're saying about a blockchain reset, so in our clients, we will have to delete some files to download the new blockchain ?

... And same question about de 98m others coins, to preserve our investments, what will happen to them ?

Thanks

Quoting cuz I'm really interested

when all 100m coins are created in the genesis block - what will happen with the ~98% that will not be released in the ipo?

Could please define "too quickly"?
What's the pool size exactly?

And thanks for the update.

2 million out of 100 for the IPO? Isn't this a tiny amount?

How should we expect that our investment will stay >10cents/skycoin through time?


It seems more like a donation (or fund raising) than an investing opportunity, correct me if I am wrong.

IPO Update:

I am glad people have been able to get the client working!

Two things left, the IPO
- finishing up refactor. wallet state is now in /src/gui and visor/daemon are independent of wallet now. more work on this, but after IPO
- The tox bot is in python, it needs to be in golang to use some functions.

How IPO will work:

There will be a bot on TOX. Install Tox, make sure it works.
- you will message it with Skycoin address
- it will message you with Bitcoin address
- You send bitcoin, it will look on address for outputs and send skycoin

IPO Price

The IPO price will be 10 cents per Skycoin. At whatever the Bitcoin price is. Adjusted daily or hourly.

2 million (of 100 million Skycoin) will be sold in the IPO.
- if the coins sells out too quickly, then up to a total of 10 million coins will be released (so that everyone can get a chance)
- we want to keep the coins scarce

A pool of coins will be set aside for developers and people helping project. They will be able to buy in at the IPO price later, regardless of what the price is. This should encourage work on fixing bugs, build scripts and cleaning up codebase.

Distribution Over Time

- 40% to 60% of the coins will be distributed in the next five years.
- The undistributed coins will be locked in 100 blocks of 1 million coins and publicly verifiable.

The rational for the Skycoin Distribution is as follows:
- A slower distribution over time, will reduce volatility compared to distributing them all at once
- The distribution will be N coins per day. The rate will taper and decline over time, like Bitcoin
- If the price gets bubbly or goes up too fast, we will increase the distribution rate
- If the price goes down too far, the distribution rate will decrease to take downward pressure off the price.
- If the price falls down too far, we may buy coins back to stabilize and put in a price floor

Distributing coins over time, with a taper, (like Bitcoin) still encourages early adapters who come in after the IPO. It keeps the price low and allows a stable and gradual appreciation over time.

Instead of raising 30 million dollars at once, it is better if coins are sold off to pay on-going expenses. The expenses will increase as we hire more developers and hit milestones.

A large fraction of the coins will be distributed as bounties for development, to key contributors and subsidies for hardware.

Negative Things

I feel horrible writing this, but its true.

Mistakes we are trying to avoid:
- In NXT the IPO was closed early. We have no idea if two people own 95% of the coins and could dumping at any time.
- In Mastercoin, all the coins were created in the IPO and there was no incentive for people to come in after the IPO. So the valuation was high but the liquidity was very poor.
- In Etherum. I dont know what I bought. I do not believe there was a valuation on the IPO. It was open ended.
- In Ripple, the user base ended up with 0.1% of the total coins. There were 100 billion coins. 100,000,000,000 and users received 20,000 coins. The distribution was so unfair, that multiple insiders each have their fingers on the keys to a market suicide machine. People received coins at below market valuation in back room deals. We are not doing this. There are mandatory investor lockup periods to prevent investors from dumping and crashing the market. We are not doing this.
- The MtGox theft resulted in a loss to users of 744,408 Bitcoin. The scammers are cashing out these coins and it is putting downward pressure on Bitcoins price. This theft was so large, that it could depress Bitcoin's price below the trendline for the next two or three years, regardless of exponential growth in adaption. We need to be vigilant.
- I think Stellar did a good job on the IPO. The ownership became more diversified than Ripple. They gave out about 3% of the coin. Then the valuation is at 15 million.

IMPORTANT

There will be a blockchain reset soon. There are some minor last minute changes to block header and I dont want to delay IPO further to fix this.

Skycoin is designed so that we can
- export the transactions in JSON from the current blockchain
- reimport the transactions in the new blockchain

This means that balances will be preserved. Bitcoin had a lot of bugs that could not be fixed because they would change hashes and break the blockchain going forward. In Skycoin we plan to fix these errors and then roll over the transactions. The balances should always be the same at the end and there should be tools to verify this.

Note:
- The emergency alert system is not implemented yet. When its implemented it will give exchanges and users warnings about resets and required upgrades. Exchanges will get a field returned in JSON to signal human attention. Users will get an HTML banner in the client upon startup or the receipt of the message.

There are several scheduled changed that will change binary format and require blockchain resets:
- we are adding merkle hashes of the outputs spent and the outputs created by all transactions in a block. This will guarantee bit for bit agreement between all clients about operations applied to the blockchain. If a cosmic ray hits your RAM and flips a bit, this will detect that before it causes problems. If two different CPUs, compiler versions or optimization flags produce different values, this will detect that. This change will happen soon.
- We have a new merkle construction that allows proof that a block header exists in a series of N blocks given the genesis header and header for the blockchain head and log(N) blockchain headers. This needs to be worked out before it is implemented
- We want a merkle-tree of the unspent output set in the headers. However, removing N0 outputs and adding N1 outputs to a tree with N2 outputs would require too much hashing, if we simply took the whole set and reconstructed the merkle tree each block. If there are 1 million unspent output, that is 500,000 hashes minimum, every block for verification. A modern CPU does 50,000 hashes/second. If we have 1 second blocks, we cannot have block verifications that take longer than the block times, or a client downloadings the blocks will never catch up. One solution is to have a list and only allow appending to the list and zeroing elements in list. Then store this run-length encoded. Then the merkle tree can be updated, with N0 new outputs created and N1 outputs zeroed, with (N0+N1)*log2(N2) hashes, where N2 is the number of outputs spent to date. This means the depth increases forever in number of outputs that have ever been created. Another method zeros the spend outputs and keeps a sorted list of the zeros and adds new outputs hashes to the zeroed slots. It is unknown, when we will have this worked out and tested.
- We want to upgrade the hash function from SHA256 to something better. The new "sponge" functions look very good. They are faster and offer better security, but are not ready yet. A sponge function has an internal state of 1024 bits, and you push 256 bits in and the state is updated. Then you squeeze 256 bits out. SHA-3 Keccak is slightly faster than SHA-256 and in hardware or FPGA the speed very good. We are waiting on this, until there is time to evaluate it.
- We may add a turing complete scripting language, which is severely restricted. There will be a deterministic LLVM type virtual machine with restricted type system and minimum number of operands. A piece of code (a function) will hash to H. A transaction using H will prefix the transaction with 256 bit hash H and then the parameters will follow. This would allow new transaction types to be added (multi-sig transactions) and old transaction types to be deprecated, without breaking earlier blocks. Blockchain parsing becomes independent of the golang code or compiler. Skycoin transaction types would be severely restricted to a minimum set, but personal blockchains would have access to arbitrarily powerful contracts. This is deferred indefinitely.

Meshnet Update:

We have a new group working on meshnet hardware.

Most hardware today is OEM and off the self. Apple gets sames from dozens of companies, tests them and then chooses best one for the price. Then they put the hardware into a nice looking case. We are now looking for off the shelf modules or chips for 5 Ghz and 24 Ghz that give us control over the beam direction or other options. This is being handled by separate people, so wont slow down software development anymore.

We are talking to contracting firms who can source chips, do prototypes or keep us informed of new hardware. There are two groups in San Francisco with same hardware requirements as us, working on a similar project, so we work with them.

5 Ghz is 6 cm band. It is line of sight. It is blocked by leaves, but bandwidth is 150 Mb/s to over 1 Gb/s. There are single chip radios implementing the 5 Ghz band 802.11n/802.11ac protocol with beam forming or a raw 5 Ghz radio with 40 Mhz channels and 4 antennas outputs. We can attach an FPGA to this and a trace antenna or commercial antenna and test it. This is line of sight, so it might have to be placed a pole, on a roof or out the window. However, the antenna direction can be re-positioned electronically and potentially connect to multiple others.

For point-to-point and backhaul, Ubiquity has 24 Ghz hardware. "AirFiber". It has two antennas, one with horizontal and one with vertical polarization. The latency per hop is 1 ms. It is full duplex, so the latency does not increase if other side is transmitting. The cost is $150 per unit. The speed is 150 Mb/s to over 1 Gb/s. However, it requires installation on a pole. The antenna requires alignment to within eight degrees and it is point-to-point only. Wind can knock the antenna out of alignment. However, it also runs linux.

There is another version used for WISP (wireless internet service providers). It can connect four to twelve users at 150 Mb/s. A directional panel beam, with beam forming.

The meshnet is viable now. The technology is here for small scale meshnets, over some geographies. The equipment available is advancing rapidly. Our initial assessment was significantly more pessimistic than reality.

- WiMAX failed because of licensing fees and spectrum costs. Hardware costs never went down to wifi levels. All technologies going forward will be open access (unlicensed spectrum, wifi style licensing).
- 700 Mhz hardware is becoming available. 20 Mb/s radios are available now. This has mile range and penetrates concrete. Will work in urban areas. Latency is poor. Will work for text messages and basic internet.
- 24 Ghz and 5 Ghz are now usable for unlicensed operation. These are high bandwidth and line of sight.
- the 50 Mhz to 700 Mhz television whitespace bands are opening up (802.11af). This will fail miserably because of channel width and restrictions, but could improve in future.
- freespace optics is becoming viable. DIY and professional hardware exists commercially.
- cell phone providers are allowing customers to run femto cells on premise. If customers are running stations on premise then we may have an unlicensed LTE band. We may be able to run meshnet over LTE.

The hardware for meshnet mass adaption will be ready commercially within five years. It will only improve from there.
- point-to-point and backhaul is ready now. Wireless Internet Service Providers (WISPs) are proliferating
- meshnets will not be easy for the first decade. It will not be like plugging in a wifi device. You will have to climb a roof or get someone to do it for you. You may have to install panels on side of house at an elevation. You will need line-of-sight. The wind will blow the antenna out of alignment until hardware improves (beam steering or mechanical gimble, second generation). This is more dangerous than Bitcoin mining.
- The first areas to get deployments will be rural areas or areas where existing internet is slow, not available or expensive.  We have the hardware for this right now,
- Dense urban areas are edge-case. Need SDR, custom equipment.

We looked at the athens meshnet. We determined that the barrier is currently software, not the hardware side. Commercial Wireless Internet Service Providers (WISPs) are essentially commercial meshnets. There is a growing market for hardware aimed at WISPs and the first meshnets will be using this hardware.

Their networks are structured hierarchically. The depth max is three. There is a large microwave uplink to a location they have roof or tower access. Then they beam point-to-point 1.5 Gb/s connection to another roof. then they have directional antenna covering a 60 degree to 120 degree arc that services four to sixteen customers at up to 150 Mb/s each. The network fans in, to fatter and fatter pipes. These are small companies, with usually less than 1000 premises (50k/month revenue).

What Skywire provides WISPs is two things
- non-hierarchical routing (in a non-hierarchical namespace). This means load balancing and potentially redundancy.
- inter-operability between WISPs. This means ability to bridge network across customer premises, third parties and other WISPs.

For testing:
- We need to get Skywire prototyped and working as darknet as internet overlay (in progress)
- we need option to disable link encryption so it can run at line-speed on the ubiquity hardware used by the WISPs. (new requirement)
- the routing needs to work (designed but not implemented)
- load balancing may need to work (more difficult, possible but open problem)
- we need to find people who operate a WISP and determine what their requirements are.

After the prototype for Skywire is done, the meshnet part and the internet part will split and need to be staffed by separate development teams.