Sure, this was recorded 5 years ago and the APIs have changed. I believe the github repo reflects these changes.

Trimmed copy is just a means of making it so that you aren't loading the entire transaction into memory every time you copy it. It would be too expensive to load all of that data instead of just the inputs and outputs. Since you only need the inputs and outputs to track the flow of the currency, it makes sense to trim out the rest of the metadata for that purpose. Block verification is somewhat straight forward once you get past all of the cryptography. First, we take some publicly known data, in this case, its the block header. We add a counter to it, then we get the hash of the data plus the counter and check to see if that hash meets certain requirements. If it does, you are done, if it doesn't you move on to the next counter and keep incrementing until you find a hash that satisfies the conditions. For verifying transactions, we want to verify the signatures of the transaction inputs. Its basically just a simple ecdsa verification and if it succeeds, we know the blockchain hasn't been tampered with. If you have any more questions, feel free to ask. I may remake this tutorial series at some point, but for now there isn't much else I can do but answer your questions.

​@@TensorProgramming thank you for your reply! I hope you do remake this tutorial for the current state of things :) this is very very informative. Unfortunately it takes a lot of time for me to figure out the technical implementation for blockchain, but luckily there's this tutorial so i can dive deep into it :)

Because some of the outputs in the transaction may not belong to the address, my bad

Many android devs use Kotlin these days. Its cleaner than Java though I wouldn't say it takes queues from python as much as its just more like C# with some hints of bits of Scala, Groovy, JavaScript and Gosu in there.

glad you liked the content.

The speed is going to be based on the difficulty you set. Given I built this blockchain almost 5 years ago, I would have either been using a 3960X threadripper or an amd fx 8350. It was around this time that I replaced my CPU.

I am using indenticator mixed with indent-rainbow. The later is the one that is giving them all different colors.

For Genserver specifically? The Hex Docs on it are pretty comprehensive and do a good job explaining what the abstraction is about. Just keep in mind, its basically just a pattern that you can follow to easily make processes and message passing APIs. Elixir School would be another resource that also has some good explanations on GenServer and other OTP concepts.

Each of the processes on the BEAM starts at 2kb and then increases from there based on the data structures that you assign to their heap. They can get pretty large depending on how stateful they are. Fortunately, you have per process garbage collection which you can invoke manually if you really need to. That being said, I've only really had to do this on one app that was pulling millions of timeseries entries from an influxdb. Eventually, I ended up using a process pool for this task and just dumped the process after it was finished instead of using the GC directly. This is also an effective way to keep the processes low in cost; killing them and passing the state off to a new one after minimizing its size. Its almost always useful to treat your elixir/erlang processes as disposable. You can see the process size using the observer and the IEX repl.

@@TensorProgramming Thanks a lot Tensor for such a great explanation!

@@Rishabh0501 Not a problem. As an amendment to said explanation, each process has its own isolated heap. I kind of implied as much but thats how they can start at 2kb.

Well, good luck. I do recommend maybe going through some of the fundamental videos first however as you will need to learn fundamental golang concepts to get into this program.

You'd want to swap over to UDP instead of TCP and then you would need some kind of relay system to find the peers. Raft for example would be very easy to implement on top of this code.

Thats something I've been meaning to do for awhile and its certainly still on my Todo list. Dart is a powerful language even outside of flutter but many people just overlook it.

Proof of work.

@@TensorProgramming can you please tell me how can I add pos

@@Moddingmonster-v6f Thats not really something that I can describe in a KZbin comment as its vastly different from PoW.

Good o' insomnia...

Glad you found the content useful.

It should be. I will be replacing it at some point soon though.

Glad you like the content. I should maybe do a rerelease of this project at some point.

I don't see the point. Serverpod is plenty fast and you could add gRPC endpoints to your serverpod instance.

I use a glove80. But was using a corsair keyboard when I wrote this video.

Not really sure what you mean.

@@TensorProgramming There is a diagram in the thumbnail with boxes labeled "Port" and "Adapter". The adapters live on the very edge of the system, providing access to the ports for a given scenario (e.g. a "web" adapter accepting HTTP request and forwarding the data to the "create order" port). But in your diagram, the ports are on the outside, and the adapters are on the inside.

@@flflflflflfl Ahh right I see what you are saying. If I recall, I just grabbed the diagram off of google to make the thumbnail. Unfortunately, that means someone else made the mistake and is proliferating that bad information. Its so small that I am surprised you noticed but yes you are right. Ports are the boundaries of the application where as adapters are the components that allow the application to interact with specific technologies. I'm not sure its worth fixing as my explanation should be correct in the video and again the image is so small in the thumbnail.

It really just depends on the usecase. Msgpack might be generally faster and smaller but thats not always the main concern. Json APIs tend to be easier to implement since you have access to many different libraries.

Definitely a nice setup for debugging. Obviously you would want a filter if you pushed something like this to prod so that you don't flood the client with information that they don't need.

You can call rest apis from a gRPC client via certain plugins. Theres one that lets you encode the data to/from Json. You won't get the advantages of using protobuf from these calls, but it doesn't add any extra overhead to a normal rest call. In other words, you don't really need to build your own logic for this, just use one of the many plugins.

@@TensorProgramming wow, I must admit, I feel honored for receiving a response from you and so quickly! Really appreciate, master. gRPC feels so cool and I am just going through your go microservices series. And I was thinking... I have a microservices system: a client app that sends through graphQL requests to my main server which does of gateway for my microservices through gRPC and thus, my microservice that consumes a vendor REST API gets touched from gRPC. I was thinking... Well, I just need to do the regular REST JSON HTTP 1.0 POST request to the vendors API inside my microservice gRPC method, retrieve the response and propagate it back through gRPC to my gateway server and then it can answer to my client graphQL request. Does that architecture make sense? I am new to this, I am sorry if what I am saying sounds stupid.

@@TensorProgramming Thank you so much for your response, I feel honored.

Elixir doesn't change that much; nothing in this tutorial is depreciated.

@@TensorProgramming Thank you sir 💚

Glad you found the content useful.

I usually get Norm McDonald but I suppose I could hear some Jonah Hill in there as well. Thanks mate, glad you found the content useful.

Spacemacs dark with fira code font.

No, its just code ligatures. Ligatures are just a means of taking multiple font characters and combining them into a single character. In this case, either 2 or more equal signs.

Nice, glad you found it useful.

In go they are very similar.

It just has to do with the basic conversion that happens inside of the repl for the sake of simplicity. In reality, all of these values have equivalencies as far as data goes.

Its showing the input and the output. The first line is just what I typed where as the 2nd line is the input in the repl and the 3rd line is the output. If you did this in the CLI instead of emacs, it would just show one line for input and one for output.

Elm as a language hasn't really changed that much. I believe the version is still 0.19.1 which is the release that came out after this series.

Glad you found it useful. Good luck, that kind of migration can get hairy if you aren't careful.

I hear you. I went from Elixir to erlang and it definitely was different.

Right, its the magic of editing.