|
smooth also added the very clever point that in the case Proof-of-Stake devolves to a computation contest for computing Nothing-at-Stake game theory, then this is perhaps a Proof-of-Work system in disguise (and I add but it might still also have some of the bad traits of Proof-of-Stake as well)
Pos is weighted proof of work.
|
|
|
|
|
The solution to scaling & decentralisation to me would go with making independent subnetwork with good characteristic to assure fair topography on the network,it would probably result a loss of fongibility, but all big world wide network works like this. I dont have detailed design like this, but I think i have most brick for it. The adverse pb of total fongibility is that ironically it cant detect and deal with large number of colluding nodes, and enforce fair network.
Well I would more go that way but maybe it's not good solution.
Wild Block size increase is not viable solution to me, and it doesnt solve decentralisation and scalability.
|
|
|
|
I have been reading lot of your post on rust forum & git, I will try to issues some of them, it's so funny how no one thought the same concept I use to have all this things of collections and traits  I can say in my whole framework there is almost zero compiler defined type. Only for very basics structure, but no high level object is ever defined in the compiler, and modules can still exchange with each other, and they dont share any kind of compile time type déclaration  Saying this to answer a post I saw you made on rust saying program need to share data type definition, with my system it works without,only members/key name need to be know, and the native type can be determined at run-time, and converted automtically to a requested type without having to declare anything at compile time or knowing the native type. Can allow to detect directly if a member is even present or not at run-time. Instead of having non initialized/undetermined variable. It's similar to js with this, when json data is received, there is no type declaration anywhere, and the program can still access the data based on key name, and the order of the keys doesnt matter, or if there any other key in any order, without having to have shared declaration anywhere. All this is made with still relatively simple C code It's funny reading those post I remember the different thing I tried to get this behavior in C/C++, either with pure virtual variant class, with some c++ template, funktors, tried everything but the ultimate solution in fact is much simpler, just need to forget about compile time order and verification, which are still often not that useful, and focus only on the runtime and runtime checking, which is imo much simpler and safer, and can allow for plenty of runtime check on everything, and can cop with dynamic typing and run time type definition. The more it goes, the more i think all this trend with programming paradigm based on strict compile time defined type create more problem than it solve all together, and compiler are never able to deal with the true complex cases anyway. If you sum up the kind of situation compiler can solve with compile time type checking, to me it's mostly simple mistake, and I do think there is a whole generation of programmers traumatised from assembly programming and cheap cast trick in C that they obnubilated to conforming to high level language synthaxe , having often not that much clue of the real problematic involved, and just stick to synthaxe definition and compile time check, which in fact only solve minor mistake or trouble coming from assembler programming. I already solved the shape collection issue with the raytracer it doesnt even need any compiler defined type Now I think either there is a massive pb I dont get with my system, either i feel people are really looking elat the problem coming from wrong paradigm Looking back at those threads, im remembering all the nightmare in compiled languages with dynamic typing, I really wonder why no one seem to have followed the same road than me lol It just works so well for everything from usb drivers to blockchain rpc server, with simple standalone C code, I really wonder why such system hasnt Been tried before. All a function need is a pointer as arg, and it can check all the type, and members, members of members, and test the presence of any without any prior declaration of the type for the compiler. The data can come from anywhere from hell to heaven, the function will never crash and be able to get the data from the member it need, and check the presence of the member, and return failure if the data is not what the function expect, in any case, there is very little chance of crash of access to wrong memory address, even if the function has no idea where the data come from and there is zero tacit coupling at all between the caller and callee. I guess it's not silver bullet for everything, but normally it take very dumb programming to make a crash or memory leak with it, and it's only pure C89 code. With a layer on top of it, it would probably be even better, but in the same time I think also most complex part of the app are going to js and html5, so not sure it's really worthy to have another high level scripting language, or for having 100% standalone app who doesnt require a js browser or node.js. |
|
|
|
Will destroying Bitcoin by fixing it make us insanely wealthy?@IadixDev, I appreciate very much your and other app devs' interest. App devs should be able to use which ever programming languages and frameworks they prefer. If you have some preexisting code which can be readily integrated, there is a golden rule of software development to never entirely rewrite existing source code from a blank slate. Perhaps a music app can be popular and earn money in OpenShare's funding model, even without requiring the content providers to signup. Need to be clever as I pointed out upthread. Personally I love music so I would use it if it is done well. No need to repeat here the debate about some of the tradeoffs of C. In addition to all that recent discussion at the Zenscript Github issues tracker, @keean and I had been discussing Rust in 2016. W.r.t. to those technical insights/choices/preferences, I think over time programmers migrate to what is proven to work best. So we'll find that out as we iterate. I think this OpenShare project should have its own slack or forum. Slack is better for interactive discussions, but I don't want to be on interactive discussions all day else I won't get much coding done. Forums are better for permanent records of information. So I suppose what we need is a forum and a slack which I visit on a schedule, perhaps 2 - 3X a week. I think that would be sufficient for now and we don't need an official discussion thread on BCT yet because we have still too much coding work to do. Did you check a bit the code on the git ? https://github.com/iadix/purenodeWrote a few article on the website too on those topics of memory management in C , and binary compatibility, but it's not complete yet. http://iadix.com/ico/blogs/enI think i found a good way to deal with the C downside with internal allocator with reference counting, and dynamic tree graph, I think most issue of C with memory & stdio glitches are in check. Normally there is not a single call to malloc or the stdlib at all, it use only kernel level functions , and internal allocator with free zone stack, and reference count, supposed to be quite fast. It's completely GC like, except memory is made available once reference count reach zero, with intel lock atomic instructions for reference inc/dec, instead of being garbage collected periodically. All variable that have no reference are freed once the ref count reach zero. https://github.com/iadix/purenode/blob/master/libcon/base/mem_base.c#L1138It dereference recursively all children , and free them when ref count is decreased to zero , but otherwise it's the same than than GC with manual handling of local variable scope. The only Goody of cpp would be custom operators and automatic destructor/reference dec for stack variables when they get out of scope. With C the scope of stack/local variable need to be managed manually with manual ref dec when the reference get out of scope. The system with the tree layer source is not on the public git, there is only the binary module and the import defïnition, but I can give access to the git with the full source, or publish them, but id rather avoid prematurely And it already get rid of native C runtime, to have compiler/system independant binary modules. With dynamic tree graph with runtime data type definition, secure log system, and handling of http/json/rpc based on module export. And normally it completely cop with C downside with pointer ownership, with no memory leak, dynamic list of dynamic type with reference counting etc. All like GC languages. And with the system , even list/array can be typed, eg to differentiate between a list of block hash and a list of tx hashes, based on json interpreter with optional explicit typing. |
|
|
|
Any way, that is why a strongly typed language that transpiles to JavaScript is something I am considering. C can also be compiled to JavaScript with Emscripten (as I think you've alluded to above).
Server-side we probably want C/C++ code, but perhaps Node.js is okay for the proof-of-concept stage. Remember we are trying to get to market asap. Then there will be funds to throw into development to refine everything.
Let's talk this out on a slack soon.
I already can develop app quickly with the C framework Already have html5 player & ogg p2p streamer, it can get to the market quickly, quicker than starting from scratch with node.js I can have a prototype running in the week, no shit. And it works on android browser. The 3d, crypto, and audio stream. And the idea is there is almost zero double code between the "server"/node side in C, and the app side in html5/js. Because each part have different purpose. And the goal of using C is having binary machine code, with sse, system calls, opengl/opencl, and direct full kernel api for sockets and i/o, transpiling this to js is not very useful. And I already have working system of cross platform binary module, with the http/json/rpc server, that bind module export as http/json/rpc interface, with the source and binaries on the git. It already works, it's already developped, and can already be used to make html5 application with html5/js with rpc calls to the node modules. It just miss some kind of high level définition for interfaces and modules and application code to have something really clean that can generate html5 UI and define applications components , with Event handling the json like data type, inter modules interface etc Maybe sometime it can make sense to have double code as fall back for when javascript engine in the browser is not enough. |
|
|
|
Maybe I can appear strong for people to think im weak  Some have less subtle approach  https://youtu.be/URybdpu_NhINot sure who is winning, everyone is cheating anyway :p Let's do it! I just think maybe "we" will choose other apps as a priority instead of the music app. But I am also okay with doing a music app first too. I love music and would surely use the app myself! We'll brainstorm about it soon... I will be very enthused about apps I will myself use, because I will have many ideas of how to innovate them. My million user successes in commercial software were the ones I created because I wanted to use them (and saw the market was lacking the specific features/capabilities that I wanted). For a music app, I want it to be my music player and also keep track of all my music so I never have to hassle with backing up my music, transcoding formats, etc... And I don't want it to tie me into any walled gardens, no adware, no bullshit, etc.. Yeah we'll be "cheating" also but copying for example the way others are already cheating: http://www.listentoyoutube.com/ (get my idea yet?) Have you ever noticed you can overlay and obscure the SoundCloud HTML player buttons? Instant library of songs without the 10,000 plays per day app limit. I have some (clever or out-of-the-box thinking?) ideas we can discuss. Remember our apps need a social component. There needs to be sharing (likes), commenting, etc.. But the more important point is how you the app creator will get paid. And how much money you will be able to earn creating apps. All you need to do is make apps that people will like to tell their Facebook friends to use. I think from there, you can start to deduce what I have in mind, but I hold off on the details while I try to finish up getting the preliminaries of the scalable, decentralized blockchain for OpenShare into code. Then we will start to talk in earnest about collaborating, launching, and making a lot of money while shocking this community with our STICKY, VERIFIED adoption rate (into the millions I expect). I am excited because now I see that real capable app developers are contacting me. So I only need to convince you that my blockchain technology and onboarding strategy is viable, then we can go change the world. I am on the 2 drug treatment now. Let's hope my energy is ready roll now. I need go finish up on the proposed changes for the (optional proposed app) programming language and see if it realistic for me to write the transpiler in a matter of weeks. I am going to try to go wrap that up now after doing my forum communication now. Are you interested to help on the transpiler? Should I start a Github project for that? Which name is best Async, Copute, Lucid, or Next (Nxt)? All app programmers please keep in touch. I want to make you wealthy. Let's have fun also. I'll make an official thread and Slack once I get through the preliminaries and am confident that my production (in code!) is back up to normal. Yeah I think the music buisness need to do the switch with uber and p2p, with all the thousand of starving artists who will never get signed anywhere, and even the one who are signed not sure they are very happy with it, but they are locked and cant go to the toilet without asking their producer, the people to convince are more their producer for the one who are produced, and the economy is not really Blooming in this sector for a moment I can say with fair amount of confidence many are looking for new solution markets and income. And the music industry is not that good with internet since the beginning. And no producer for an artist it's the dream Need to think also about good back office to have stats & méta infos, and good way to remunerate participants, node owner, artist, developpers. For the language, I have mostly the low level part on the git, with the module system, the dynamic tree with reference pointer, http/rpc server, and most C runtime, and the vector lib for the raytracing. Then the modules method can be called from a js application with http/json/rpc. For the high level part to make application, i dont have much theory I would go for something very basics who can call module function and handle Event with json like data type, and in the spirit akin to BASIC with simple one line statement easy to evaluate, charm is really the kind of things im looking to have good OOP encapsulation in module with an interface. Js as far as i know is weak for defining good object sur typing and interface, but if I understand well, your idea is to have a source language with good sur typing and interface définition, and transpiling to js to have équivalent code. But for me i would still rather stick to a kernel in pure C, that can expose module interface to js/html5 app, and having the node/server side in C rather than node.js. Much better for performance, memory use, and portability. The only thing it really miss now is a good html templating system to generate html5/js page based on input data. For this im not sure if the best is to have browser side generated html like angular js, or something that can generate preformated html from the node, or using xslt which can be done by both server & browser. Or something entierely different to define UI and Event handling and rpc call in html5/js. That would do in sort to have part of application in C modules with the framework, and part of application on js/html5 who can call those modules. But having another source language to transpile this part of the app with the UI and modules interface, why not. But all the part with binary data crypto & transcoding in js  |
|
|
|
Maybe I can appear strong for people to think im weak Some have less subtle approach https://youtu.be/URybdpu_NhINot sure who is winning, everyone is cheating anyway :p |
|
|
|
But... Michael Jackson said Sony kills the music.The popular musician agreed to be in the music industry's system.They are kidnapped as Kids in the getho and brainwashed by lawyer, it's not exactly the same But yeah I guess it's not completly good synergy for the moment, but the fruit might just need a little kick to fall down too even if some people seem to think they are entitled to the tree.
they will fight back with every dirty trick they can musterChicken :p |
|
|
|
We have a good contact within independent music producer, but I dont have the feeling they are very fond of internet, and have used & abused by many system already, they dont have high amount of trust on Internet to sell music. But the conditions are also no that good for many of them, so I guess it can be tempting There was the ODRL thing that was interesting, regarding the problematic of digital right, if a thing like this can be integrated as blockchain object, that could be convincing with labels & music producer https://www.w3.org/community/odrl/ The ODRL Policy Language provides a flexible and interoperable information model to support transparent and innovative use of digital assets in the publishing, distribution and consumption of content, applications, and services across all sectors and communities. The ODRL Policy model is targeted to support the business models of open, educational, government, and commercial communities through Profiles that enhance the model to align to their requirements whilst providing a common semantic layer for interoperability. Need to find the good mocking Jay to start the revolution or with anonymous blockchain for them to cheat their producer with a vo coder Maybe some would jump, like daft punk, or some other who understand technology and have good lawyer for internet right I think Kim dot kom was on this too, and then he got arrested  15 years ago we were going to see big producer to speak about internet music distribution, suddently their face turned like they are sit on a cactus. It looked very painful. I think they d rather deal with terrorists than with internet hackers |
|
|
|
Thank you for your input. I look forward to hearing your design.
I will get into wall of text mode to give a better picture lol The thing is originally my background is i got into programming browser component in the start of the 2000, with components based on ATL and NPAPI/XPCOM, with code that had to be portable over two browser interface, which include also the events related to the system windowing, the component model XPCOM/COM with their own interface definition language, and the system to link the IDL interface to javascript binding. Also related to the idea there is direct show, and it's when i started to understand how microsoft were totally genius with this system, because it allow people to develop new codec and filters, to be integrated into standard filtergraph and being used in any application. Result, microsoft they just do the skeleton for plugin system, and all application based on it can use all the codec and filters programmed for it. And same goes with activeX style programing, to extend browser functionality. The catch ? Well it needs completely hardcore type definition to fit with COM, IDL, dll, registry based, with the whole interface to define and instantiate those complex object. I got my brain totaly smoking on this one for a few year in the 2000's, with this whole mess of defining complex component to be instantiate by blind application, with massive surtyping and all this. But the real trigger is when i started to program with AS3, flex builder, and android sdk, i started to really see all this component based pattern everywhere, based on XML data definition like flex with eclipse, WSDL kind of programming with flex, it completely made the trigger to me. I post this to explain the background, and it fit with the github discussion of iamnotback on it, it's completly the problematics my brain have been crunching in the past years, and it's completely in the center of the problematic with distributed application where some component or module can be run on top of the block chain by node that are blind to it's data type. From there , there is two main path to get to real distributed application, there is the way how i explained a bit before, with some block being marked as being pure data block, replacing the merklle root by the hash of the data, and a system to mark those block like why not with a path system '/application/myapp/mydata' in the block data, and then some option to tell the client which path he want to download, and the other he just need to keep the block header to keep the chain coherent. After for the data itself, the thing i have in mind and turning around for a while, is to have those block also being able to contain structured data, that are related to module, why not with some kind of OLE like thing based on data type and component implementing interface to deal with them, and then you could have something like a coinbase like input to mark the type of the data, and eventually the parent block for hierarchized data, and then the members of the object as outputs of the transaction. Like this it would allow to define tree graph like thing on the block chain, with data hierarchy, and runtime defined object, that can be loaded in json like runtime type based on the transaction. Each member of the object can be changed with a transaction based on the output that define this member. And then having components module loaded by the node to implement the interface either via rpc or direct http request outputing html5 data, or xml/xslt. The general idea is this, but there are maybe still caveat, but i'm polishing the plan in my mind Add to this a little pinch of darpa browser experience science, and it's effectively the eth killer |
|
|
|
.
All the other ideas can be done later, on other alts, or whatever.
I think that's what I'm trying to say. Aim low, and achieve it, before aiming high. It's harder than it looks. ;-)
And I wish you the best-- please don't take any of this as negative.
If you are writing something new, however, from scratch, with several key changes it will take you a long time. As an individual developer, at least 3-4 years.
Well personally, I looked into bitcore code, and it's far from being ideal to me. For plenty of reason, it's monolithic, not necessarily extremely well architectured, heavy memory foot print etc. I already have 80% of working block chain client with my framework, in 100% C with zero deps & using 30mo of ram. The good thing about bitcoin client is it's very well tested , mature etc, but still not that easy to quickly add new feature or encapsulate into a bigger application. And im perfectly in tune with iamnotback on the need of programming paradigm switch, and im already on this for 2/3 years So going back into multithread boost c++ like bitcore with the super high coupling everywhere, with all hard coded config, I would rather not |
|
|
|
Decentralized Social Media: Quietly I've developed a decentralized "message board" application on top of Blockchain technology. It does not have the issues with centralization that PoW mining leads to. I have the basic concept implemented but It's not ready to be released publicly. Unlike imnotback though, I will describe my specific approach . The basic idea is this: Instead of a "Single Global Blockchain" where there is contention to get your information into the chain through mining, every user has their own chain. They hash their previous header/new data then sign it from their new header (like traditional blockchain). Here's the difference, instead of mining instead the header includes "witness hashes" pointing to the most recent blocks from chains of their peers. The "witnesses" prove your data is more recent in topological time, as well as provide a vote to prevent a "double spend" attack in other chains. This structure obviously forms a DAG which can be sorted in linear time. In order to form "chain communities", people can subscribe to 1 or more "root" blocks. The software client looks subsequent blocks which have cryptographic traceability to these roots. The person who created the root block may publish branching messages on their chain pointing to other peers chains. This will inform software clients that they may also follow these other peer chains of they so choose. The peers chains can issues branching as well. If one of these chains starts spamming too much information/branches the client has configurable rules to discontinue following these sub-peers. Each person for themselves can determine how much of the network they can handle store, etc. The End Result: "chain communities" every single block/message is un-censorable and has full traceability back to the root (like in blockchain). My design has all the benefits of the chain, (integrity, non repudiation, byzantine fault tolerance, etc.) while being low-latency (< 1sec) by avoiding PoW mining. It should be resistant to Sybil attacks, spammers, double spends as far as I can tell. You could build crypto-currency applications, smart contracts on top of it if you so choose. Let me know what you guys think. I would go more for single chain with block data filtering than having a chain / user from the moment the data is not transaction, and doesnt relate to previous state, the whole block validation process become meaning less, as well as most of double spending related issues. And it remove lot of the issue to incencitive block emitters. From the moment it's content data, any block is valid. And the data can be skipped if the person dont need it and dont want to "host" it. There is not much rule to validate content data itself, if it doesnt contain transaction. I will probably give a try at detailing the way i want to do it, but it also rely on whole system of type class like system and module/component/interface between app & blockchain. Multichain have implementation of storing structured data with signature & timestamping, but I think my system is potentially better. |
|
|
|
Even 80% of the market prefers to only goggle over Justin Bieber, there are significant markets that want more diversity. For example, Shawn Grunberger who created BandCamp was a guy that I encouraged back in 1995 to pursue his dream (when I was the programmer and he was working the customer support dept). We are headed into a fracturing of society where humans break away from a monolith and assert their differences. Some years I was involved with developping a c++ streaming module based on ogg/ogm and independant distribution platform, was still involved with this in the beginning of last year with blockchain / html5 based things. But we saw very quickly it's very locked, artist and author have literally gun on their temple from their producer, and even signifiant part of them has been heavily brainwashed on hating p2p in the 2000 Can hope on independant artist or content creator, but they will rarely bring high commercial interest, and still large part will prefer to stick to traditional edition/promotion companies, even with the bad conditions, but they dont think any other way to make good revenu with content creation than being signed with universal or Sony. We have somehow à plan with decentralized music streaming, and made quick prototype with multi chain, and home made ogg streamer in the node + html5 player and some thinking on this kind of system. But not sure it's 100% viable commercially and if enough artist who are not already chain signed to a big one would jump the step. And explaining block chain to artists good luck already 2+2 they have an head ache |
|
|
|
I'm just not sure about your organisation, and how to get involved with your ideas, if you are looking for people to develop stuff
I am trying to get to that stage asap. Sorry I am slow for the reason I explained in my prior post. Also I can't think entirely clearly yet about optimal (organizational, etc) strategy for acceleration of progress. No pb, I can totally understand this I know this kind of issue, and also I prefer to stay out of hot code and decision when im not 100% :)I need to kick myself in the but sometime again when I mess up code trying to advance when im not 100% now it's clear I avoid to rush in long coding project and prefer to take time, think things out, and get into coding & décision when im confortable enough doing it, after i absorb as much of theory as possible, watch other project, and solidify the idea two week in hospital bed with a pneumonia was enough for me i can totally understand this  |
|
|
|
Decentralized social media is your first target ? it's one of mine too, i have also a revenge to take on youtube and deezer Do you have precise idea of the characteristics for decentralized social media ? But the problem is always to have commercial grade content posted on it, it's same with blogs & social media & streaming, and 95% of the market is locked by majors and big editors. |
|
|
|
And, the more confidence you lose in the brand, the more you will be interested in the technological aspects, development, research and adoption (by people who actually matter).
That would be good news |
|
|
|
Coming back to this some months later, I will now add some details.
The blockchain is now named OpenShare (domain opensha.re) which has been publicly well received. Xxxxx (domain xxxxx.us) remains a good potential name for one of the public facing apps that will run on and be the public face of the project to the masses.
We need a way to motivate others to develop apps for the system as well as to motivate users to join and use the system, because we know we can't possibly scale up this thing doing all the development and marketing by ourselves. And we'd like to provide so many money making and development flexibility opportunities for others to minimize the jealously and power struggles that occur in the cryptocurrency sphere. Sufficient adoption inertia by both devs and users can minimize the viability of threats from copycoins and hardforks. In the whitepaper, I even mentioned the technical design facets such decentralized introduction of new data formats and separately smart contracts merged-mined isolated from each others' state machines.
By [redacted], we can fund four class of actors involved in ecosystem promotion, marketing, and development:
1. Developers of apps which attract the users.
2. Power users who [redacted]
3. Providers of content which attract the users.
4. [redacted]
...
So the idea is we can grow the marketcap ranking on Coinmarketcap.com, fund developers, content providers, and power users ...
...
But even in this paradigm I have proposed, the app developers have full autonomy! And let a 1000 apps bloom! They don't need to do an ICO to raise funds! They don't have to incriminate themselves! Now finally app development can go mainstream.
Hello I am motivated to make blockchain based application What do you have as framework and base to build the applications ? I think i'm on the same line than you with many things, with typeclass, json like data definition, javascript, DOM, and i have good experience in many domain you are talking on the git I'm probably about to pull out the basics of script engine to program blockchain node based on my framework with the binary module components, json data definition, and i already have all the working code base for this I'm just not sure about your organisation, and how to get involved with your ideas, if you are looking for people to develop stuff, i'm on the same line for theory in language definition and decentralized code, and i'm rather pro efficient in programing in many languages, and i'm totally aware of the problematics you are pointing in the zen script git hub discussion I'm aware also of health problem, had a pneumonia some years back, and still have some amount of neuropathy/fibromalgy, but i keep myself healthy now I try to do sport and eat healthy and manage stress and anxiety better But i'm aware with this problem and concentration needed to do good code, need to find to good pace |
|
|
|
Hello Some news from the developpment i have put online the new block explorer : http://iadix.com/ico/blocks/enWhich include a filter system, the api looks like this : http://iadix.com/ico/api/blocks?height<86000&tx*>2&size>400 https://github.com/iadix/purenode/blob/master/block_explorer/block_explorer.c#L212
tree_manager_set_child_value_i64(result, "height", height);
tree_manager_set_child_value_i32(result, "size", size);
tree_manager_set_child_value_hash(result, "hash", block_hash);
tree_manager_set_child_value_i32(result, "confirmations", nblks-height);
tree_manager_set_child_value_i32(result, "time", time);
tree_manager_set_child_value_i32(result, "version", version);
tree_manager_set_child_value_i32(result, "bits", bits);
tree_manager_set_child_value_i32(result, "nonce", nonce);
for (qv = 0; req->query_vars[qv].value.len > 0; qv++)
{
if (!tree_node_keval_i64(result, &req->query_vars[qv]))
{
release_zone_ref(&block);
return 0;
}
}
I also added a mechanism to buffer blocks that are potentially orphaned, and being able to put a node on testing to test a few blocks and rollback the current chain to the fork point if the other chain is longer. https://github.com/iadix/purenode/blob/master/purenode/main.c#L1182if (testing>0)
{
hash_t lh;
int new_len,cur_len;
if (!tree_manager_get_child_value_hash(node, NODE_HASH("last_header_hash"), lh))
node_get_hash_idx(testing-1, lh);
tree_manager_create_node ("log", NODE_LOG_PARAMS, &log);
tree_manager_set_child_value_hash (&log, "ph", prevHash);
tree_manager_set_child_value_hash (&log, "lh", lh);
log_message (" testing block '%ph%' -- '%lh%'", &log);
release_zone_ref (&log);
if (memcmp_c(prevHash, lh, sizeof(hash_t)))
{
release_zone_ref(&header);
return 0;
}
new_len = node_add_block_header(node, &header);
cur_len = get_last_block_height();
tree_manager_create_node("log", NODE_LOG_PARAMS, &log);
tree_manager_set_child_value_i32(&log, "new_len", new_len);
tree_manager_set_child_value_i32(&log, "cur_len", cur_len);
log_message("testing chain at height : %new_len% - %cur_len%", &log);
release_zone_ref(&log);
if (new_len > cur_len)
{
mem_zone_ref last_blk = { PTR_NULL };
int n;
uint64_t lost_reward;
lost_reward = 0;
n = cur_len;
while ((n--) > testing)
{
char chash[65];
hash_t h;
uint64_t reward;
if (!node_get_hash_idx(n, h))continue;
n = 0;
while (n<32)
{
chash[n * 2 + 0] = hex_chars[h[n] >> 4];
chash[n * 2 + 1] = hex_chars[h[n] & 0x0F];
n++;
}
chash[64] = 0;
if (is_pow_block(chash))
reward=get_blockreward(n);
else
get_stake_reward(n, &reward);
lost_reward += reward;
}
tree_manager_create_node("log", NODE_LOG_PARAMS, &log);
tree_manager_set_child_value_i32(&log, "testing", testing);
log_message("truncate chain to height : %testing%", &log);
release_zone_ref(&log);
if(node_truncate_chain(testing))
{
if (tree_manager_find_child_node(&self_node, NODE_HASH("last pos block"), NODE_BITCORE_BLK_HDR, &last_blk))
{
unsigned int pBits;
if (compute_last_pos_diff(&last_blk, &pBits))
tree_manager_set_child_value_i32(&self_node, "current pos diff", pBits);
else
tree_manager_set_child_value_i32(&self_node, "current pos diff", 0x1FFFFFFF);
release_zone_ref(&last_blk);
}
sub_moneysupply (lost_reward);
log_output ("switching to new chain \n");
tree_manager_set_child_value_i32(node, "testing_chain", 0);
}
}
release_zone_ref(&header);
return 0;
}
if (!tree_manager_find_child_node(payload, NODE_HASH("txs"), NODE_BITCORE_TX_LIST, &tx_list)){ release_zone_ref(&header); return 1; }
if (!tree_manager_find_child_node(payload, NODE_HASH("signature"), NODE_BITCORE_ECDSA_SIG, &sig)){ release_zone_ref(&tx_list); release_zone_ref(&header); return 1; }
tree_manager_get_node_istr(&sig, 0, &signature, 0);
release_zone_ref(&sig);
ret = node_is_next_block (&header, &lastBlk);
if (!ret)
{
ret = check_node_chain(node, &header, &keep_block);
if (!ret)
{
free_string(&signature);
release_zone_ref(&header);
release_zone_ref(&tx_list);
return (keep_block==1)?0:1;
}
}
And also 2 command line parameters to rebuild local index and re validate local block chain, or looking for corrupt data. To recheck the block headers and tx indexes. To recheck block reward and proof of stake/proof of work to rebuild the money supply. Additionally i started to write some more in depth documentation about the memory model, and the binary module system : http://iadix.com/articles/viewarticle/en/purenode%20memory%20and%20component%20programming%20model_0/blogs/purenodeI'll probably update it to add more example and better page formatting Still no date from the ICO, nothing is really heavily tested for the moment, don't hesitate to mentions bugs or errors or comments Normally i just have to finish the page with the 3D demo, and we can start to get more precise with the ICO details, with the escrowed coins and the official detail, and the bounty campain. Thanks for your patience |
|
|
|
none of those alts solve the fundamental problems that bitcoin has so why whould i want to get on any other alt but hold btc?
this guys claims to have the be all end all of cryptos solving every posible problem seen in bitcoin and every other alt, i dont believe it. why should I anyway?
There are solution that exists, in the things I posted from e lang, they already offert solution to have trust less decentralized network with crypto and there is no blockchain or pow/pos. They are making à system of decentralized e cash with it. In theory the solutions exists, but the system with pow/pos is still the simplest, anyone can install and run a node relatively easily and the model of security is simpler in blockchain, but truly trustless decentralized system are already being worked on. And the theory is already there and according to them their code has already been audited and reviewed by independant groups. |
|
|
|
|
Show Posts
