If it need to send a transaction, it's hard to avoid the blockchain layer Especially if they have to be signed in the browser, which is important for security. And keeping private data private. Aka private key never have to leave the UI in clear form.


But in the idea Yes it's script to program this sort of application logic.

But I think with a bit of intelligence most of the complexity of smart contract can be avoided with simple code with a good scripting engine to handle this kind of problematics.

Ideally The blockchain state transition are also related to the delivery of the content, and there is a séquence of events involved between different nodes in triangle between music producer (content owner sign reward to distribution, fix distribution condition), distribution  ( content server, check distribution condition, send content, get reward from producer), and customer ( html5 app sign tx to content owner, get the content from distributer).

But the music server in the idea he is still the bitch of the system, he has no decision in the blockchain, it just has to check the client is ok with the content owner and deliver the content for probably cheap reward. It doesnt need to hold private keys or anything, specially if it store crypted files that can be de crypted by the html5 app from interaction with content owner.

But any case it has to work with a triangle.

So it needs some form of accounting and higher level définition of transaction based on the accounting layer or something like this to completely abstract it away from the blockchain layer.

The music server need to be able to check chain state, either it's by monitoring Event or polling the state from a node when it needs it

I guess you can call this a smart contract, but the way it's done it's more like ready only function in the music server, it doesnt have to sign anything or emit new blocks. Just check the access and earn a reward. The reward is signed by content owner from another node when the stream is verfied by the player app.

For the moment as i said the design is still partially centralized.

It does because the UI has to manipulate addresses, and sign transaction, and check balances etc

It's not necessarily for smart contract, but script language to program the blockchain node itself, the block validation logic, Event handling of network message, blockchain request, tx signatures, and the distributed applications itself, mostly the UI with html5 / js template, with a js modules for in browser crypto, and some module call via http/rpc/json to do the heavy work related to blockchain request or heavy computation.

Server for music content are not completely independent from blockchain, they need to check permission access to the media , and also involved with reward for content storing/distribution.

And how do you plan to go for the blockchain developpement ?

You want to start from an existing code base, or code a whole new blockchain node/wallet/block explorer from scratch ? That is monthes of steady daily work, at best

Especially if you want to have high level scripting engine, on top of Blockchain protocol, how many time to get all this from scratch ?

Because I have already large part of this already done, not sure after how you see the priority between the blockchain itself, the script language, and the application layer , and if you plan to start from scratch how many time before an application can be programmed with it ?

If you are ok with the model of script I explained before ( https://bitcointalk.org/index.php?topic=1739268.msg18375980#msg18375980 ), I can start working on it and have something working soon, but it would probably not be directly suitable for javascript transpiling as such.

What you have in mind is starting on node.js code directly, or develop the transpiler first, and then develop the blockchain and application with the language directly and transpile it to node.js ?

Or starting from a code base in c++ or other ?

Normally my framework in C can do the node.js job for many things and I still have the prétention to say it's much better on all aspect, less memory use, more stability, direct access to kernel api for i/o & threads, compiled to binary code, can put in line sse or avx in the code direct, integrate ffmpeg with ogg, deal with binary data easily, and have efficient crypto, and it has the reference counter and dynamic typing, so I dont see how you can have better in shorter time

If you see big problem with it or how it can be improved nothing is really fixed

My criteria is i want cpu power and parallelisation, opencl/opengl, modularity and dynamic typing.

With node.js I dont have cpu power and efficient parallelisation, and not sure about opencl/opengl support. Dynamic typing and modularity  ok, but not at the price of giving up on all the rest

For quick prototyping ok, but then need something else at some point too than node.js modules.

The things I like with node.js is the module system to manage large software stack in js, but the node.js runtime in itself not too much,  it doesnt look all that stable etc. To export node.js app to browser bundles with html5 app ok, but not to run full nodes with the node.js runtime.

My main objective with blockchain is to develop distributed applications, because I think it open new paradigm with public sharing of information etc, the fact that there are well valued coins only add to the attractivity of the ecosystem, cause real value can be traded "in app" which is completely cool

After im not here to invent the perfect blockchain or ideal currency, im only in to program distributed application, for the economic aspect I trust the high valued coins, and i make a module to use them in the applications

I believe more in user controlled private network based on modular application layer to optimize security for the application / user group purpose, rather than having to rely on complex whole chain conscensus for everything, hoping that everything will always sort itself out the best for everyone by itself

The way i would see it is application producer have key pair to authentify affiliated nodes who run their modules, to have a degree of verticalisation in private network, and nodes not involved with this application just skip the block data, and just keep the proof of existence hash in the block headers, if they just keep track of coin transactions. And the node running the modules can validate the blocks based on higher level protocol rule implemented in the modules. People who are affiliated and run the application modules can get some form of reward from application supplier if their module is run, or if transactions happen through it ( nodes never have to know private key , all signature happen in the browser with in browser crypto). Need to be careful it doesnt turn into some pyramidal scheme too much, but with only one level of affiliation normally it's ok. But I bet there are always some people who can turn it into a pyramidal scheme

Only the nodes involved with the application affiliation can validate the application blocks, and the higher level of the protocol is handled in the affiliated group. It could have its own pow difficulty targeting , or pos based on app tokens.

I also plan to develop tradional web market, to integrate app in cms or prestashop, symphony etc for integration of web app in traditional web platform.

Either it's blog content, audio, app data for social media ( caregories / threads / post / users ), or shop with article item, it can be handled in application layer modules.




It's more the way i view it globally

Yes I wanted to have your opinion on this nushare system , it's a solution I looked over to solve this issue to avoid all the power to be driven by the whales, but maybe it's not a good solution

Seemed to go in this direction to have token for decision power separated from the main blockchain mining / emission

Like this can accumulate decision power independantly of economic interest and wealth.

Did you see the system they did with nubits/nushare ? Seemed to get in that direction to separate monetary unit from décision power with 2 tokens.

https://www.nubits.com/nushares/introduction

NuShares are units held by individuals who wish to help support and maintain the Nu network. Owning NuShares is not required to use NuBits. Instead, NuShares are intended to be a source of network equity for developers, entrepreneurs, and speculators. NuShareholders can receive network revenues in the form of Peercoin dividends paid out by a custodian.



What can NuShareholders vote on?

There are several actions NuShareholders can vote on. The actions to vote on are Custodian votes, Park Rate votes, Motion votes and Transaction Fees. NuShareholders will be able to vote on these different network actions when a block is minted. By setting their vote in the client it will be recorded into the blockchain each time they mint a block. The purposes of these categories of votes are to select custodians to maintain the network, choose interest rates for parking, make decisions on the development of the network, and determine how much using NuBits should cost a user.

The buzz word for this is collaborative economy

The thing is in my idea, im doing a framework to deal with blockchain protocol, or build new blockchain easily, to extend easily the protocol, more than doing a monolithic code to handle one blockchain with hardcoded values. The same code/exe can handle different blockchain like multichain. With different config or module for each.

There is not even really any compile time definition of any type at all. All the protocole message template could be put in a json file and loaded at run-time.


https://github.com/iadix/purenode/blob/master/export/iadix.conf

All the code for block & tx validation is modularized, the code to handle key pair and signature, as well as the modules behind the http interface like block explorer and rpc server for web apps.

Different blockchain can be made with only very localized change to the code, and preferably via config file for easier integration with above layer.


If you want to build a block chain engine from scratch , with the feature you want, I dont think you can find better starting point, because ive looked in the same issues than you and didn't find anything that had all the good requirement.

But im being objective here, not specially trying to " sell my stuff because it's my stuff"  

I already have the whole type class like things working with an already fairly debugged framework, with implementation of flexible/modular blockchain engine.

After custom blockchain can be made with other protocol with different modules. And keep common code as much as possible  ( not whole git fork to change 3l ).

All code is very localized with low coupling. Even my gran ma can compile it.





Good night, sleep well 

I know you are not fond of ico, and I can understand why, but we have been thinking the pb in many ways, ICO are good model of funding, we can have some contact in the public funding things , or private investor, but it's clear to fund good independant project with open source based on blockchain it's still the best.

But I know what you mean too with the craze and the number of scam and all

Well can always try, if it works and can get good funding, it can already solve a lot of pb to quickly kick start commercial projects.

But the ico blockchain will not be the only focus of developpement, just maintaining it and integrating new functions in it when it make sense for this coin.

All what im talking about is already made and working and available on the git, with some bit of documentation on the site.

In the plan we will probably be dealing with many different blockchain, either the one of existing crypto coins, our own app blockchain, or other blockchain created with our framework, or integrated in the framework with specific modules.

We are not really a company for the moment, my friend is musician

I dont think we have hard line with cooperation, idk what you have in mind ?

I already have most of the protocole implemented with the block explorer and html5 raytracing demo on the site

The only thing it really miss it the memory pool and good debuging / testing on the p2p network, with profiling , and why not adding Zlib comp, and also having all op code of bitcore script engine implemented, but I have already the basics, most of the op code dont seem to long to implement.

Hello

For the moment the team is preparing the ICO, and it's not really in my hands for the moment, so i have bit of free time

For the development part, anyway for the moment i'm more focused on development distributed applications, with the modular system it can handle any blockchain protocol and connect to any blockchain, so i'm not so so concerned about one coin in particular, i think if bitcoin can gain good reputation it can be better for the world of crypto currency as a whole, so i can only wish all coins gain maximum of trust and value

The next thing i might be getting into though could be the script system, but it would be something very simple in my mind more something like a BASIC with mostly built functions that are call to module methods, the only thing it would really need is a way to declare json variable, and call to module function with these variable, with fixed function prototype like with javascript event handler with a single json object tree as parameter.

For the moment i have a system to add sur typing to json object like this


you can see the definition of the built in type there

https://github.com/iadix/purenode/blob/master/libbase/include/tree.h

so it look like

{ ("key_name", TYPE) : value }

instead of just

{ "key_name" : value }

To add a surtyping to the key, it is useful for serialization, but it can also be used for OO like programming if certain module or function need to react to a specific type of object.

The script language doesn't even need a module definition, as they are module export, the framework can already load up the function address from the module and function name. And it's already cross compiler with a resolution of function name based on the declared decoration specific to the compiler used to compile it. Like this a module compiled with visual studio can import symbol from a module compîled with gcc , or vice versa, with unified function export name.

So the runtime can detect if the method exist or not, or could make script that can parse module exported function list, so for the moment the script language can remain very simple i think. Most function of looping and list processing can be handled by the framework in modules.

The only thing it would really need some good thinking on, is a system to easily generate html5/js template for most application, to make the development of distributed html5 application easy, and i see there are some discussion about this on your git, if you have idea that can be easily implemented, i'm all hear

As for collaboration, normally there is no special licensing on the the code or anything for the moment, it doesn't use any GPL code, with the modular design, it's made in sort that several person can develop their own module for their own app and only share the code in common.

I made it this the very beginning to have very low debugging time, as it's a system i developed for micro kernel operating system, it has to have zero exception/seg fault in the code, that was first very important requirement, even dealing with interupts, preemptive multi tasking, and high level application, with zero memory leak, the code had to be made in sort to avoid any low level fault, because they are very hard to debug in this kind of environment.

With this system of dynamic typing, all the information can be available at runtime, for debugging, even in very low layer of the code, it can still acess all the data structure of drivers, library and application, even from a bare metal exception, so it helps a lot with debugging as well with the modular design it's more easy to know where the code goes wrong and having debugging output.

So it's all made to make safe application quickly normally, and already can handle lot of the http/rpc/json protocol as well for audio streaming and html5 applications, for the moment the audio system is still based on centralization for certain things, because for cthe moment i don't see easy way to design perfectly decentralized design that can keep everything secure for all parts involved in music distribution, not that it's impossible, but to come up with something really useable in large scale, it cannot be done without funding.

For the nash thing, yeah as you say anything it's something that is just supposed to emerge by itself from the collective, the only reason why it's not supposed to be already there is because there are some anti ideal force at play or some lack of development in certain area of financial infrastructure, which is think blockchain can partially solve, i guess blockchain as world wide distributed decentralized system can be a good medium to test nash theories, but there is not anything special to do in the blockchain to make it happen normally =)


For blockchain themselve, i would be really in favor of adding some kind of identity metadata associated with addresses group, to have more possibility for trusted peer exchange, or even some kind of private cache of mempool shared between trusted entities, and some possibility to black list certain address on the network level, or to give possibly certain restriction that could be added by the protocol, or some warning about certain address or node that the user could bypass or not, i'm not sure i'm ideally all for 100% annonimity/ fungibility and i think the network also can have room for also private network, possible with an application layer above the blockchain protocol with other type of data than transactions.


https://www.youtube.com/watch?v=AHzeP0WeKIA

Well it's where the concept of selfness can become confusing with empathy. With the simple definition, people who act with empathy are not selfish.

I dont think torturing people, even "enmies" really bring much pleasure to anyone. Only need to see what remain of vets after few years like this. They dont look very happy.

Haaaa the great pleasure to torture your enmy in the name of freedom



Sight  

This is entierely false. Level of life and energy available /  head grow exponentially with global population growth, steady. Place with high density of population always have more manufactured ressource at there disposal.

If you are better on your own without anyone else,  you dont need currency or economy.

I think this quite apply to the crypto currency problem

There are two arguments for continuing jobless benefits, a moral argument and an economic argument. The moral argument is easy to make, and can be stated as a Moses-like 11th Commandment.

Thou shalt not fuck other humans up the ass against their wishes, even when they are strangers to you.

I note only in passing that the Earth would be a far more pleasant place to live if all humans everywhere adhered to this new Commandment.


The point of course is that bad things happen when you fuck a stranger in the ass (FSITA). Yet humans fuck strangers in the ass all the time. They are strangers because they are not in some person's in-groups.

In the years preceding the financial meltdown, a considerable number of people in the housing industry—the realtors, the mortgage lenders, the finance people who "securitized" those mortgages into bundles which they sold to "investors", etc.—fucked a considerable number of strangers in the ass, including millions of American homebuyers who were trying to make a buck off rising house prices. Those naive people were told that home prices would rise forever, Those people believed that nonsense!  As a result, many bad things happened, and there was a lot of collateral damage, as with The Dude's car.

When an a multinational corporation moves manufacturing or services offshore, leaving the workers of some American town or city high and dry, that's the owners of that corporation fucking a bunch of strangers in the ass. In fact, FSITA is ubiquitous in large, liberalized market economies, and especially in the United States, a large, diverse "melting pot" whose  economy is based on "free" markets. Lots of strangers there!

I am talking about this rarely-reflected-upon phenomemon in crude terms in the slim hope that you might become sensitized to what goes on every day here in the United States. FSITA is so common that all but a few people have become desensitized to it, especially when it a happens to strangers. If you're the one taking it up the tail pipe, of course, it's a different story.

Currently it seem bitcoin is stuck in this situation where no good decision can be made because the way that power on the network works with pow doesnt encourage this.

And there is still need to make good decision for it to keep going forward. Or it will keep stuck with greedy idiots wanting to cheat each other. And hash based pow is unable to solve this.

The example of how dao solved the hack show this.

Consciousness of consequences of bad and good action on other is necessary to have any viable society/economy.

The argument with pleasure seeking can be that there is no higher pleasure that seeing someone else perfecting himself

It's what predatory economy fail to integrate in their schema

Even amazon ceo made good article on how  100% selfish motivation never improve economy. Because having more poor underdeveloped people never bring economic growth. Having more educated and competent persons does. And free market often tend to skip this. And you can say he is not a socialist.