MaidSafe IPO on Mastercoin

[gamey]

It looks mostly good to me (there is one strange Segfault bug but it doesn't look critical).

Luckybit, show me I am wrong and point out specifically what you are referring to here.

If you can, I retract any questions about your expertise and will gladfully apologize.

[bytemaster]

Maidsafe has one primary challenge:  Latency

Disk storage is cheap and abundant....
Bandwidth can be facilitated with bittorrent style downloads.

It is the initial lookup of the data you require that is the challenge.   If you divide a large file up into multiple chunks then the majority of your time will be spent finding the chunk.   Lookup times scale with log(n).

Their other problem is load balancing and price fixing.... not all data is in equal demand nor costs an equal amount to host.   

Unless you can resolve the price fixing and load-balancing issue as well as the latency challenge the system will not scale well. 

If their application is to store your personal data in the cloud rather than hosting data for others to download, then it is a much smaller nitch and far less interesting.

[gamey]

Luckybit,

I am impressed with your attempt at demonstrating your technical expertise.  Quoting protocol layers etc.  It doesn't mean you understand it.

For example -

I don't think you understand how MaidSafe works. There are no IP addresses so what would Tor be needed for? If you have MaidSafe there would be no reason to require Tor. A decentralized application could provide a VPN service which would be better than Tor.

Even protocols based on top of UDP use IP addresses.  You can not operate on the internet without IP addresses.  To say maidensafe doesn't use IP addresses shows you really don't understand what is going on, yet you want people to believe you read their code.

You say a lot of little things like this that don't add up.

Responding in particular to the above quote, I have been saying repeatedly that I think maidensafe's bigger application is a better TOR.  Yet you conflate that into something else to continue arguing.  THat is because your objective here is nothing but promotion, so what others say has little effect on your pages  and pages of ...... thoughts.

Dropbox, do you have the code so we can review it for backdoors? The fact that we can review the code means we can crowd audit it. It's a risk to use the code in the testing phase before it's audited but considering we are risk takers I don't see why it's a big deal if you're willing to risk your money already.

Again, you don't listen or read.   When I brought up dropbox, I explicitly mentioned truecrypt.  Truecrypt has recently had a serious 3rd party independent audit.  I use truecrypt to encrypt.  Dropbox keeps my files on their cloud service and replicates across any device I own.  Mac/Phone/Windows/Linux.  All of those.   Yes, this has the problem in that an error in dropbox service can propagate across all my devices, but this is the same level of issue you'll have with Maidensafe.

You made another example about how one's computer can be hacked and they can lose their BTC.  Maidensafe doesn't stop this.  If you are hacked, a keyboard logger will open up the door to maidensafe unless maidensafe uses one of those visual keyboards that relies on mouse clicks.

You make tons of assumptions how others use their computing devices to make your point, when you don't understand what the services that people use actually do.

Maidensafe is a great idea.  Neat.  I'll play with it.  I just lol at the thought of it being more disruptive than bitcoin.

[toast]

It will turn up being a jumble of recursive nonsense.

Well put. A theory can be consistent and comprehensive and still be wrong... explaining why might involve having to find incorrect assumptions about totally unrelated parts of reality

[toast]

MaidSafe does not need IP addresses. You have no IP address on the MaidSafe platform. The IP address that you get from your ISP is for the regulated Internet. MaidSafe is a decentralized anonymous Internet platform. If no IPs exist on MaidSafe then it's pure Cipherspace.

What... what is the new network-layer protocol?

[toast]

I am not going to respond to every point of yours and pick it part. It will turn up being a jumble of recursive nonsense.  It is obviously you are not of a strong technical background.  You don't understand how mining works and why that will not be a app.  You also don't understand how TOR, Dropbox, or what Amazon EC2 provides.

Mining can be virtualized. A virtual machine is an app which simulates some hardware. You do need enough computing power to do it but there is a lot a computing power and it's not hard to create virtual machine instances which simulate certain hardware.

Anything can be virtualized. It is impossible to virtualize mining *profitably*.

[toast]

this thread...

[luckybit]

I am not going to respond to every point of yours and pick it part. It will turn up being a jumble of recursive nonsense.  It is obviously you are not of a strong technical background.  You don't understand how mining works and why that will not be a app.  You also don't understand how TOR, Dropbox, or what Amazon EC2 provides.

Mining can be virtualized. A virtual machine is an app which simulates some hardware. You do need enough computing power to do it but there is a lot a computing power and it's not hard to create virtual machine instances which simulate certain hardware.

My technical knowledge is strong, I may not always be the best at explaining.  Amazon EC2, Tor, Dropbox, all could become decentralized applications. The hardware can be abstracted or separated from the software in such a way that the hardware could be decentralized nodes which do the processing, storage, or whatever. If you'd like to discuss how it could work I can go into that and if you or anyone find technical errors or my explanations are not good then feel free to correct me.

However, the quote above shows why people may not take you are seriously as you'd like.  You are claiming you've studied the code in some significant manner and thats how you know it is good to go.  Thats obviously not true.   

Some of the code I will admit is a bit hard to read because it's written in a way where it's not commented or it's not easy to figure out what it's doing. The overall design looks good, the important parts of the code look good.

Here is the RUDP portion of the code which you can review for yourself. https://github.com/maidsafe/MaidSafe-RUDP/blob/master/src/maidsafe/rudp/transport.cc

It looks mostly good to me (there is one strange Segfault bug but it doesn't look critical).  I will admit I have not reviewed every little detail of the code (there is quite a bit of code). What I did is skimmed through it and looked for certain design decisions they went with. Does that mean there could be bugs that I missed? Of course I wouldn't catch everything but if enough people look at the code I'm convinced that one of us will catch something if something is there.

Dropbox, do you have the code so we can review it for backdoors?
http://www.techrepublic.com/blog/it-security/researchers-reverse-engineer-the-dropbox-client-what-it-means/

The fact that we can review the MaidSafe code means we can crowd audit it. It's a risk to use the code in the testing phase before it's audited but considering we are risk takers I don't see why it's a big deal if you're willing to risk your money already.

I realized some time ago, your point behind arguing is not to exchange/improve ideas. It is purely promotional.  You just sort of make up stuff as you go and ignore corrections to your blatant mistakes.  Rinse and repeat ...

If you find a mistake and can prove it I don't mind saying I'm wrong. But you're not saying anything to prove or disprove. You're saying that MaidSafe is not like Tor while I'm saying MaidSafe is better than Tor. MaidSafe is better than Tor and Freenet combined and works at a lower level than both.

Tor operates on the application layer (OSI layer 7). Freenet is also the application layer (OSI layer 7). MaidSafe is on the transport layer (OSI layer 4). It does not run on top of TCP/IP, it has it's own RUDP transport protocol. Tor uses onion routing to keep you anonymous but it's still running on the classic Internet, while Freenet when I last checked uses encryption at the application layer to try to create a Cipherspace or secure channel. It has had limited success but a lot has been learned from it and it seems MaidSafe developers in specific has learned quite a bit from the design.

But I will stand by my opinion that MaidSafe is better than Tor and Freenet combined. If you think I'm wrong show me otherwise.

Uhhh, I'm not the one thinking maidsafe will not use IP addresses.  You have no clue, absolutely 0.  I may be wrong about the economics, because I don't fully get their argument/graph.  I thought it was based on something akin to Moore's law, but I think their graph is based on adoption rates.  So I made a mistake there perhaps.

MaidSafe does not need IP addresses. You have no IP address on the MaidSafe platform. The IP address that you get from your ISP is for the regulated Internet. MaidSafe is a decentralized anonymous Internet platform. If no IPs exist on MaidSafe then it's pure Cipherspace. It's like the Tor hidden services only without the IP addresses to track. If you have decentralized meshnet ISPs which will accept Safecoins to provide access then you would have no IP address at all to track and if you use a traditional ISP they'd at best could know you use MaidSafe but still would not able to track you.

I've looked at it and I cannot figure out how anyone could monitor it. How would you monitor something which is encrypted? If there are no IP addresses where do you start? I suppose you could try to monitor at the ISP's which people will connect through but I don't see how the ISPs could know who you are and what you're doing on MaidSafe since its pure Cipherspace.
 

https://github.com/maidsafe/MaidSafe-RUDP/wiki
https://github.com/maidsafe/MaidSafe-Encrypt/blob/master/src/maidsafe/encrypt/self_encryptor.cc

I may be wrong about the economics, because I don't fully get their argument/graph.  I thought it was based on something akin to Moore's law, but I think their graph is based on adoption rates.  So I made a mistake there perhaps.

The graph is about adoption rates but also Moore's law. It basically makes the case that Moore's law will result in the costs of computing resources decreasing, there will be an equilibrium point where Safecoin rises in value in correlation with the decrease in costs of computing resources.

I believe this correlation is going to result in an increase in buying power/purchasing power for all who hold Safecoins at the time. So if a Safecoin can buy more computing resources on MaidSafe than any other similar service then you could mine most profitably on MaidSafe which means miners would want to buy Safecoins rather than buy a bunch of Digital Ocean instances.

I also want to say that I'm not making these posts for promotional purposes. While I do hold Safecoins, I had to spend my money to get those coins and I would not have done that if I did not believe in the potential. My opinion when the IPO first happened was that it was shady, it was unfair the way it happened, they botched it. Just because they botched the IPO it does not mean their technology isn't the best out there.

A week ago I was defending Ethereum when others were asking why would the Bitcoin association go with Ethereum. At the same time I told people I probably wont take part in the Ethereum IPO because it's a bad deal for investors due to the insanely high inflation rate.

What I say about MaidSafe, Ethereum, and Blackcoin's multi-pool idea, are all based on my own research and conclusions. If you'd like to believe that my conclusions are wrong that is fine but it's not going to change my opinion and in either case it's not "promotion" to admire technologies other than Bitshares.

From Business point of view:

Cloud hosting for mining purposes should be beneficial for the user for short periods of time (if at all). After all why rent something for X if you can set it up to mine for you and receive Y>X.

If you're running a data center then you would make plenty of Safecoins just rending out resources. If you have a lot of Safecoins but don't want to buy a mining rig and pay for the electricity, maintenance, and other issues associated with that then a virtual machine makes sense.

How do you think people are mining Primecoin right now? It's either botnets or in the cloud else I cannot see how it would be profitable to mine.

While it is not impossible the sum to be bigger than the parts, you still have not pointed to anything of the MaidSafes design that makes you think so. Saying “It will do this and this. You will see!” is the only argument that you have offered so far.

Any technology like this is going to be difficult to explain. The same arguments used to "hype" Ethereum are the arguments which "hype" MaidSafe.  I think it's going to be too difficult for me to make a compelling argument right now because without the apps it's hard to reveal the vision. I'm looking at the technology and thinking that theoretically it could do X, but that doesn't mean it will even if it is theoretically possible.

From Technical point of few:

>>“A virtual machine could simulate any hardware you want to mine with.”

Yes, I have my virtual quantum computer running on my digital (i.e, regular) computer, the code is written in C.

In theory, not in practice. In theory if the underlying computer is powerful enough it can simulate or emulate some machine. The question is how much processing power will be in the MaidSafe network? I don't know the answer to that question but I know Bitcoin has a lot of processing power.

If we are talking general purpose CPUs then if you have a lot of them and if they can do parallel processing then you can mine a lot with just that. GPUs are another processor which if there are enough of them working in parallel you can mine with that. I don't imply that you can turn a regular computer into a quantum computer because thats not physically possible.

But if you have many regular computers in huge data centers, you can allocate the processing power using virtual machines. You can design the virtual machine to have as many CPUs as you want or run as many instances as you want. It can be any architecture that you want and there are advantages to this. I wont go further into this, let's find out if my speculation is correct in a few months when the first decentralized applications are running on the MaidSafe network.

 It does not run fast as a quantum computer should… so I will take the following steps (adding more layers of abstraction) until I get the results. First I will move it from my local machine to a virtual server. Then I will move it to MaidSafes’ virtual ‘virtual disk’ and if this does not work, or for a good measure, I will rewrite the code in Java…
That’s my idea? What do you think?

You're working backwards. A virtual machine does not take away the need to have physical hardware underneath. So the physical hardware is going to still be sitting in some data center.

The point is that if I want to mine on BOINC, and use 5000 CPUs to do it, then in theory I could do that over the MaidSafe network. In theory it would be cheaper on MaidSafe than anywhere else. Check back on my posts a few months from now and if nothing I said happens then I'll admit I was wrong.

[luckybit]

Ok I see what you were saying. I don't think you've solved the problem because it never had to do with buying the hardware to own but rather pooling your hash power and thus delegating your power wrt blockchain security in order to reduce variance.

Why would you run a VM of a cpu mining algo instead of just running it directly on the underlying machine? Its necessarily faster and can already do all algos

Sent from my SCH-I535 using Tapatalk

I think for most people it would be about speed and convenience. Configuring different mining rigs and being an admin is still difficult enough to be considered a job for most people. A share in some autonomous application could make it so people can trade the shares on the market when they don't want to own those instances.

If you have 100 different instances mining different coins you would not want to have to manage them all physically. If all you have to do is buy shares, point and click, and you can run it from the cloud, I myself would prefer to mine that way.

Now whether or not this is susceptible to making altcoins more prone to 51% attacks is another matter. I think something like MaidSafe if it scales it could put unlimited or what will seem like unlimited hashing power into the hands of certain people who own a lot of Safecoins. This group of people would be the new cloud mining elite and the centralization would be around them.

All of this is theoretical. I don't know what is going to happen but I know it has the potential to be a lot more disruptive than people realize right now. I think Proof of Stake coins will be safe but any coin which is still Proof of Work might have to worry about this and I wonder what is going to happen to Primecoin, Protoshares, Litecoin.

Maybe Protoshares should end mining once and for all because it's likely even today that most mining for Protoshares is in the cloud. If Protoshares had to launch in the MaidSafe era then someone with a lot of Safecoins could mine them all.

[gamey]

From what I've seen of the code and the design I don't see any possibility of data leakage

I am not going to respond to every point of yours and pick it part. It will turn up being a jumble of recursive nonsense.  It is obviously you are not of a strong technical background.  You don't understand how mining works and why that will not be a app.  You also don't understand TOR, Dropbox, or what Amazon EC2 provides.

However, the quote above shows why people may not take you are seriously as you'd like.  You are claiming you've studied the code in some significant manner and thats how you know it is good to go.  Thats obviously not true. 

I realized some time ago, your point behind arguing is not to exchange/improve ideas. It is purely promotional.  You just sort of make up stuff as you go and ignore corrections to your blatant mistakes.  Rinse and repeat ...

edit -

Because right now you don't even know what MaidSafe is and you're going to be embarrassed about this post once you figure it out.

Uhhh, I'm not the one thinking maidsafe will not use IP addresses.  You have no clue, absolutely 0.  I may be wrong about the economics, because I don't fully get their argument/graph.  I thought it was based on something akin to Moore's law, but I think their graph is based on adoption rates.  So I made a mistake there perhaps.

[tonyk]

From Business point of view:

Cloud hosting for mining purposes should be beneficial for the user for short periods of time (if at all). After all why rent something for X if you can set it up to mine for you and receive Y>X.
While it is not impossible the sum to be bigger than the parts, you still have not pointed to anything of the MaidSafes design that makes you think so. Saying “It will do this and this. You will see!” is the only argument that you have offered so far.

From Technical point of few:

>>“A virtual machine could simulate any hardware you want to mine with.”

Yes, I have my virtual quantum computer running on my digital (i.e, regular) computer, the code is written in C.
 It does not run fast as a quantum computer should… so I will take the following steps (adding more layers of abstraction) until I get the results. First I will move it from my local machine to a virtual server. Then I will move it to MaidSafes’ virtual ‘virtual disk’ and if this does not work, or for a good measure, I will rewrite the code in Java…
That’s my idea? What do you think?

[toast]

So you split the problem in half and solved half. But things like GHash.io effectively do that once they enable getblocktemplate to source blocks from other pools

Sent from my SCH-I535 using Tapatalk

[toast]

Ok I see what you were saying. I don't think you've solved the problem because it never had to do with buying the hardware to own but rather pooling your hash power and thus delegating your power wrt blockchain security in order to reduce variance.

Why would you run a VM of a cpu mining algo instead of just running it directly on the underlying machine? Its necessarily faster and can already do all algos

Sent from my SCH-I535 using Tapatalk

[luckybit]

Profits from mining poola are already decentralized, physical control over the hardware always gives power unless it is a pos variant.

Also what you are saying about VMs doesn't make sense givem that theoretically every mining algorithm is only profitable on specialized hardware

Sent from my SCH-I535 using Tapatalk

See this is a possibility that I'm concerned about. But does physical control really give them power on MaidSafe? I'm not entirely sure.

Say you mine on digital ocean, how do they have any power to even know what you're doing? They have power to pull the plug I suppose but they don't get the profits.  Besides being able to pull the plug their power would be in the fact that they charge fees. On MaidSafe you could run as many virtual machine instances as you want, and some large data center in the background would be paid in Safecoin. They would have no idea what you're doing.

I think it's possible in theory to make mining more decentralized but at the same time it could become more centralized in other ways.

The virtual machine argument is that if we assume any kind of hardware can be virtualized then you can actually mine a lot easier this way. You could mine any CPU coin and I don't see why you couldn't use GPGPU's to mine. We will have to wait and see what can be mined but I do see major advantages to using virtual machines. Botnets will not be profitable, it will alter CPU mining in a big way, and mining could even be made socially beneficial.

The end result is I think we might see the end of the mining pool. Why would we need mining pools which can be hacked, broken into, run off with everyones coins, etc? Additionally there would be more full nodes on the Bitcoin network because MaidSafe in theory can solve that.

[luckybit]

Bytemaster doesn't seem to think there will be that much storage demand above what individuals can reciprocate from their own storage, and I don't have much else to go on.  If that's the case the market could be a lot smaller than Dropbox and there won't be much SafeCoin transactional volume (Which 'Power of Volume' shares that I see as part of the future, partly derive there value from.)

If that is the case then Bytemaster and I disagree. Let's find out who is right.

I think he has a point about bandwidth and data limits but those limits come from ISPs and not from the demands of users. Those are unnatural limits and artificial scarcity.

If you remove the cap on what people can upload and give them enough speed to do it, they would mirror everything into the cloud. I don't think there is any limit to it because you could do stuff like decentralized television where videos stream in and out in real time. Bandwidth if it were available would be used but it's not used because it's not available.

So I would suggest they create a storage market/exchange.  where the space you provide gets automatically sold 'at the market', (A $/BTC price but paid paid in SafeCoins) you can either get back 98% of the sale price or use it to get  98% of the space you provided on the MaidSafe Network.

There will be all sorts of markets like that.  The main problem for us all is bandwidth. If we all had 1 Gbit connections we would use it. The problem is uploading and downloading terabytes per day would set off our ISP's to cap or limit us. It has nothing to do with demand because people will use the bandwidth which would use the space.

The bandwidth is an artificial bottleneck and we have to figure out how to get around it.

So that way you'd have MASSIVE transactional SafeCoin demand and also generate 1/2% dividends for Safecoin holders for every bit of storage that is used on the network.

I think computation is where the killer apps will be early on. Storage space is something I'd love to use but upload speeds are kept artificially slow and we are still stuck dealing with ISPs at this time. When ISPs start providing service for Safecoin then we will be able to see some innovation. I would definitely upgrade to that ISP if it exists, accepts Safecoins, and when Safecoin is worth enough then those ISPs will form and provide the bandwidth but I think the killer apps will be on the CPU side.

(It would still be free to users except for the fact that you only get to use 98% of the space you provide on the network. I doubt people would miss 1/2% of that space so it would be a fair trade-off & make SafeCoins very valuable, I think.)

I think too much focus is on space. You can provide bandwidth and solve the bottleneck. Bandwidth is artificially scarce because ISPs cap how much data we can send and they limit our upload speeds. What good is it if you want to upload a mirror of your harddrive to the cloud if you're limited by your ISP?

But I do think the space will be used. Virtual machines can have virtual drives hosted in the cloud. You could use a 100 TB drive hosted entirely in the cloud if you had the bandwidth to do it.  You could log into that virtual OS from your tablet, your desktop, your cellphone, your laptop, your television, and access the files on it or add files to it.

So I definitely believe if the bandwidth existed people would be making videos with their cellphone and streaming it to their virtual drives in the cloud. The higher the quality of the streaming the more terabytes it would be. So I could easily see a typical person using 100 TB within a year.

That is an order of magnitude more than most harddrives carry now. And this isn't unprecedented either because in the late 90s you would be lucky to find a 1 gig harddrive.

Now the average drive is 500 gig and ask anyone who had the 1 gig and back then, it's never enough.