Best match

Most recent

Nintendo

Switch

Newly Discovered Nintendo Filing Shows Off "Dual-Screen, ...

Best Nintendo Switch Couch Co-Op Games

Nintendo Life

Co-codamol

Co-codamol is a mixture of 2 different painkillers – paracetamol and codeine. It's used to treat aches and pains including headaches, muscular pain, migraines and toothache. It may help to take co-codamol if everyday painkillers, such as aspirin ...

Co-codamol for adults

NHS Choices

Cooperative gameplay

The best co-op games are designed to be enjoyed by at least two players, so grab a friend, colleague, family member, or loved one because you won't want to miss out on these. We've hand-picked ...

Best co-op games to play right now with friends and family

GamesRadar+

Horror

Be it on Halloween or in the middle of Summer, it is always the right time for a fun co-op horror game. Your browser does not support the video tag. GTFO is a popular ...

Best Horror Co-Op Games To Play With Friends

Game Rant

Best & Co.

Simulation

Even though simulation games tend to emulate real-life experiences, they can be fun to play, especially the ones with a co-op mode. Co-op simulation games offer a unique and enjoyable multiplayer ...

17 Best Co-Op Simulation Games, Ranked

BBC Sport

New Ireland coach Scott Bemand has appointed Sam Monaghan and Edel McMahon as co-captains for next month's WXV tournament in Dubai. The duo, whose appointment is initially for the Dubai ...

Ireland: Sam Monaghan and Edel McMahon named co-captains for Dubai tournament

Could you be saving money on your mortgage? Try out our refinance calculator to see if you could save on interest or get cash immediately by refinancing to a new bank.

Master your Mortgage

Mortgage Rates

TSB Bank

Get 4% back in Booking.com Travel Credits for your next holiday. Find stays to suit every budget, making your plans a reality. And get a little extra back in your Booking.com Wallet to spend on your next break. T&Cs apply. Access thousands of gyms, pools ...

Dive into a great deal

tsb.co.uk

Get the most recent info and news about CAASie.co on HackerNoon, where 10k+ technologists publish stories for 4M+ monthly readers.

CAASie.co is where you can plan, buy, and manage your own billboard ads, all over Australia and beyond. No strings or contracts. Pick your billboards, upload your ads, click play, and go.

CAASie.co

What Are the Top Startups in Oceania?

We’ve studied several zkEVM projects in the previously published article , among which the representative of native-based zkEVM is what is being researched by the AppliedZKP team of the Ethereum Foundation. The text below will guide you through the circuit code of the AppliedZKP zkEVM. Exploring Popular zkEVM Solutions The code is still in the initial implementation, and many crates have not yet stabilized. Thus, all discussions in this article are subject to change in the future. Please always refer to the latest code if you are to do code development and code study. This text aims to provide an easy guide to the basic framework and logic of the AppliedZKP zkEVM code base. Hopefully, more developers will join us in the further research of zkEVM. Code Base Path Branch Commit_ID Date zkEVM-circuits https://github.com/appliedzkp/zkEVM-circuits main 39a7bf7b30e25faa7db26a3cb94640240164542d 2021-10-18 Code Structure In the current code, there are three crates, namely , , and . The relationship among the three is shown in the figure below: bus-mapping keccak256 zkevm-circuits Also in , we pointed out that Bus-mapping is the state channel of and in zkEVM. The devs organize the code in the way shown in the figure above. Let’s look at each crate in turn. Exploring Popular zkEVM Solutions evmcircuit statecircuit 1. bus-mapping The objectives of bus-mapping are: Analyzes EVM execution path Constructs witness data following complete execution path analysis Provides a simple interface to gather and put witness data in the circuit, and witness it with a small number of functions So the basic function is to process witness data, regardless of whether it is the path data executed by the EVM or the state data that changes before and after. Let's look at each crate in bus-mapping specifically. 1.1 EVM Basically, EVM defines , , , , and include and contains data infrustractures like , , and . It can be seen that and are encapsulation to ; is the encapsulation of the array with a length of 32. ProgramCounter GlobalCounter EVMWord GasInfo GasCost stack memory opcodes pc gc usize EVMWord u8 is usize, but Stack is a dynamic array of . is also usize, which should be between 0 and 1023 while Memory is an array of u8. Developers use Rust macro cleverly to extract key memory information, such as traits of index and range, and implement them in a unified way. Storage is a HashMap and its key values are all . StackAddress EVMWord MemoryAddress EVMWord The opcodes crate defines the opcode traits of stack, memory, and storage, as well as the implementation of parts of the opcode traits. pub trait Opcode: Debug {
    /// Generate the associated [`MemoryOp`](crate::operation::MemoryOp)s,
    /// [`StackOp`](crate::operation::StackOp)s, and
    /// [`StorageOp`](crate::operation::StorageOp)s associated to the Opcode
    /// is implemented for.
    fn gen_associated_ops(
        &self,
        exec_step: &mut ExecutionStep,
        container: &mut OperationContainer,
        next_steps: &[ExecutionStep],
    ) -> Result<usize, Error>;
} Let’s take as an example to see how the trait is implemented. Here is the implementation of go-ethereum. mload // core/vm/instructions.go
func opMload(pc *uint64, interpreter *EVMInterpreter, scope *ScopeContext) ([]byte, error) {
    v := scope.Stack.peek()
    offset := int64(v.Uint64())
    v.SetBytes(scope.Memory.GetPtr(offset, 32))
    return nil, nil
} There are three steps of mload:
1. Take the value of offset from the last position of the stack;
2. Take the value of bytes from memory;
3. Write bytes in the last position of the stack. of mload is to do these steps. For detailed code, please refer to
bus-mapping/src/EVM/opcodes/mload.rs. [offset, offset+32) gen_associated_ops 1.2 exec_trace This is the most important crate in bus-mapping, including , , and . contains some constants of EVM, such as , , etc. BlockConstants ExecutionTrace OperationRef BlockConstants coinbase timestamp is the core of bus-mapping, which is the interface form of witness. contains content used in every execution such as memory, stack, current instruction, and Operation-related . implements its own function, which can call the corresponding according to the current instruction. ExecutionTrace ExecutionStep bus_mapping_instance ExecutionStep gen_associated_ops gen_associated_ops pub struct ExecutionStep {
    pub(crate) memory: Memory,
    pub(crate) stack: Stack,
    pub(crate) instruction: OpcodeId,
    pub(crate) gas_info: GasInfo,
    pub(crate) depth: u8,
    pub(crate) pc: ProgramCounter,
    pub(crate) gc: GlobalCounter,
    // Holds refs to the container with the related mem ops.
    pub(crate) bus_mapping_instance: Vec<OperationRef>,
} include all steps executed, and container includes each operation used in Trace. The most critical function in this structure is build, which is used to set and generate a corresponding Operation for each step in the trace. ExecutionTrace gc pub struct ExecutionTrace<F: FieldExt> {
    steps: Vec<ExecutionStep>,
    block_ctants: BlockConstants<F>,
    container: OperationContainer,
} 1.3 operation The operation crate is relatively simple, defining , , , , , and , where is of two types, i.e., read and write, and includes Memory, Stack, and Storage. The encapsulation of Operation will specify what kind of Operation it is. RW Target MemoryOp StackOp StorageOp Operation RW Target 2. zkevm-circuits This crate requires the use of halo2 to implement the zkEVM circuits, which is also a more complex part of the implementation of zkEVM. It temporarily includes the following subcrates: , , , and . evm_circuit gadget state_circuit util 2.1 util and gadget is used for tool implementations, such as converting Rust's basic data types, such as , , etc., to the elements of a finite field. Util bool u8 defines the basic tool chip. For example, is a constraint check for 256 bit (or 32 bytes), is to determine whether it is 0, and checks whether the advice column increases monotonically within a range. The implementation of these chip circuits follows the general architectural design. For more details, please refer to our previous publication . Gadget evm_word is_zero monotone Develop Circuits Using Halo2 2.2 state_circuit In , there are state proof and evm proof. State proof is used to help evm proof check whether all random read, write, and access records are effective. The records of random read and write are generated by state_circuit. State_circuit defines BusMapping and Config. From , we can know that config is the most critical part to write the circuit of Halo2. It contains all the information on the lookup table needed for evm proof. zkEVM design specs Develop Circuits Using Halo2 Take the code comments in state_circuit as an example. is the selector, where 1 represents the first line of each target, 2 represents memory, 3 represents stack, and 4 represents storage. As we can see, state is grouped by address first, and then sorted by . Note that the address here is not the Ethereum address. For memory and stack, the address is the and mentioned in the previous section, namely usize, so the values in the example are 0 and 1. is the counter, which will count different addresses separately. Value is the operation value, and flag marks read and write, where 1 is write, and 0 is read. padding is used to fill the gap between each target and has no practical meaning. is exclusive to the storage operation, which is the key, and is the value of the previous value in this target, which is 0 at the beginning. q_target global_counter StackAddress MemoryAddress global_counter storage_key value_pre /*
Example state table:

| q_target | address | global_counter | value | flag | padding | storage_key | value_prev |
-------------------------------------------------------------------------------------------
|    1     |    0    |       0        |  0    |   1  |    0    |             |            |   // init row (write value 0)
|    2     |    0    |       12       |  12   |   1  |    0    |             |            |
|    2     |    0    |       24       |  12   |   0  |    0    |             |            |
|    2     |    1    |       0        |  0    |   1  |    0    |             |            |   // init row (write value 0)
|    2     |    1    |       2        |  12   |   0  |    0    |             |            |
|    2     |         |                |       |      |    1    |             |            |   // padding
|    2     |         |                |       |      |    1    |             |            |   // padding
|    1     |    0    |       3        |  4    |   1  |    0    |             |            |
|    3     |    0    |       17       |  32   |   1  |    0    |             |            |
|    3     |    0    |       89       |  32   |   0  |    0    |             |            |
|    3     |    1    |       48       |  32   |   1  |    0    |             |            |
|    3     |    1    |       49       |  32   |   0  |    0    |             |            |
|    3     |         |                |       |      |    1    |             |            |   // padding
|    1     |    1    |       55       |  32   |   1  |    0    |      5      |     0      |   // first storage op at the new address has to be write
|    4     |    1    |       56       |  33   |   1  |    0    |      8      |     32     |
|    4     |         |                |       |      |    1    |             |            |   // padding
*/ 2.3 evm_circuit The is the core of the entire projec, including the implementation of opcode, the definition of Constraint, the definition of EVM execution results, the element of each execution step (ExecutionStep), and the most important . evm_circuit EVMCircuit For example, in the definition of Constraint, the name serves as an annotation. When the selector is a success, the gadget's selector - polys are the Constraint’s polynomials, and lookups are the lookup tables used by the Constraint. In the Halo2 circuit instruction manual, the constraints interface function in the OpGadget trait is an array that returns Constraint. struct Constraint<F> {
    name: &'static str,
    selector: Expression<F>,
    polys: Vec<Expression<F>>,
    lookups: Vec<Lookup<F>>,
} Lookup is divided into two categories. First, , which an array of 3 s; second, , which is an enum containing , , , , etc. FixedLookup Expression BusMappingLookup OpExecutionState Stack Memory AccountStorage pub(crate) enum Lookup<F> {
    FixedLookup(FixedLookup, [Expression<F>; 3]),
    BusMappingLookup(BusMappingLookup<F>),
} The difference between and is that the former is used in the call of BusMappingLookup, and the latter is used in the OpExecutionState of BusMappingLookup. CallField CallStateField The structure includes constraint expressions, columns, and the offset used by the selector. Cell pub(crate) struct Cell<F> {
    // expression for constraint
    expression: Expression<F>,
    column: Column<Advice>,
    // relative position to selector for synthesis
    rotation: usize,
} A word is represented by one expression and is positioned by an array of 32 cells. struct Word<F> {
    // random linear combination expression of cells
    expression: Expression<F>,
    // inner cells for synthesis
    cells: [Cell<F>; 32],
} The structure, , is often used when writing opcode circuits and is used for context caching in the assign implementation in evm_circuit. CoreStateInstance pub(crate) struct CoreStateInstance {
    is_executing: bool,
    global_counter: usize,
    call_id: usize,
    program_counter: usize,
    stack_pointer: usize,
    gas_counter: usize,
} contains the opcode executed in this step, the execution result, and the value used. Although the name is the same as in the bus-mapping crate, its purpose and structure are different. ExecutionStep pub(crate) struct ExecutionStep {
    opcode: OpcodeId,
    case: Case,
    values: Vec<BigUint>,
} contains the selector, lookup table, read-write table, fixed table, and op execution gadget for each step. EVMCircuit struct EVMCircuit<F> {
    q_step: Selector,
    qs_byte_lookup: Column<Advice>,
    fixed_table: [Column<Fixed>; 4],
    rw_table: [Column<Advice>; 7],
    op_execution_gadget: OpExecutionGadget<F>,
} Currently, opcodes of , , , , , and have been implemented. We take add/sub as an example to explain . arithmetic byte comparator dup pop push how to write an op circuit First, define the data structure required by op (when the operation is successful), which is also the operation data of the op in most cases. When it fails, no other data is needed, we just need to handle the corresponding case. For the structure, the selector is used for activation in different cases. Because and are implemented together, swap is used to distinguish add or sub. a, b, c are used to represent the three values of the op operation, satisfying a + b == c. When sub, a + c == b. AddSuccessAllocation ADD SUB struct AddSuccessAllocation<F> {
    selector: Cell<F>,
    swap: Cell<F>,
    a: Word<F>,
    b: Word<F>,
    c: Word<F>,
    carry: [Cell<F>; 32],
} Second, define the corresponding Gadget structure, mainly the results of the op in different cases. For add/sub, there are three states: , , and . success stackunderflow outofgas pub struct AddGadget<F> {
    success: AddSuccessAllocation<F>,
    stack_underflow: Cell<F>, // case selector
    out_of_gas: (
        Cell<F>, // case selector
        Cell<F>, // gas available
    ),
} The final step is to implement the trait for the op. To begin with, define several case situations covered by this op, as well as the number of values, the number of cells, and whether or not to halt. OpGadget const RESPONSIBLE_OPCODES: &'static [OpcodeId] =
    	&[OpcodeId::ADD, OpcodeId::SUB]; // This operation implements opcodes ADD and SUB.  
const CASE_CONFIGS: &'static [CaseConfig] = &[
	CaseConfig {
    	case: Case::Success,
    	num_word: 3,  // a + b + c
    	num_cell: 33, // 32 carry + swap
    	will_halt: false,
  },
	CaseConfig {
    	case: Case::StackUnderflow,
    	num_word: 0,
    	num_cell: 0,
    	will_halt: true,
	},
	CaseConfig {
    	case: Case::OutOfGas,
    	num_word: 0,
    	num_cell: 0,
    	will_halt: true,
	},
]; // Several case situations by ADD and SUB The function is relatively simple, which is to initialize our Gadget with . The function is a relatively critical interface, which needs the process of the current state and the next state to construct constraints. construct() case_allocations constraints fn constraints(
&self,
state_curr: &OpExecutionState<F>,
state_next: &OpExecutionState<F>,
) -> Vec<Constraint<F>> {
    let OpExecutionState { opcode, .. } = &state_curr;

    let common_polys = vec![
        (opcode.expr() - OpcodeId::ADD.expr())
            * (opcode.expr() - OpcodeId::SUB.expr()),
    ];

    let success = {
        // success constraint
    };

    let stack_underflow = {
        // stack underflow constraint
    };

    let out_of_gas = {
        // out_of_gas constraint
    };

    array::IntoIter::new([success, stack_underflow, out_of_gas])
        .map(move |mut constraint| {
            constraint.polys =
                [common_polys.clone(), constraint.polys].concat();
            constraint
        })
           .collect()
} If you want to know how to use halo2 to develop circuits, please refer to . Develop Circuits Using Halo2 3. keccak256 While we are studying the code, the crate of keccak256 is still in the construction stage. We will analyze it in the follow-up study. How to participate As written in Sin7Y’s early articles, zkEVM is the brightest jewel in the crown of Ethereum scaling, and we look forward to more developers participating in this great project. Of course, as an open-source project, you may make your contribution in whatever way you like. Fixing a typo, improving documentation, or implementing a certain function are all helpful. If you are more familiar with the development of Halo2 and want to implement a certain opcode, you can take the above ADD/SUB as an example. The implementation process complies with writing the design specifications first. After the design specifications are merged by the core developers, you can write the opcode circuit code. Finally, don't forget to write the unit test code for the newly added opcode implementation. References ZkEVM Specifications Circuits for zkEVM Develop Circuits Using Halo2 Exploring Popular zkEVM Solutions https://github.com/appliedzkp/zkEVM-specs https://github.com/appliedzkp/zkEVM-circuits https://hackernoon.com/sin7y-tech-review-develop-circuits-using-halo-2 https://hackernoon.com/exploring-popular-zkEVM-solutions-appliedzkp-matter-labs-hermez-and-sin7y-quick-publication-ltq37ah 📎 A little about us As a team of both vigorous and professional blockchain devs, Sin7Y has been doing constant research on zkEVM, layer 2, privacy computing, and cross-chain technology, delivering high-quality tech reviews on a regular basis. You can find us on or write to us at . Feel free to contact us by or if you have any queries. Twitter sin7y@web3.com DM Email

AppliedZKP zkEVM Circuit Code: A Guide

: Yes you should use a premium CDN because 1) premium CDNs are easily 3 times faster than free CDNs — 1+ seconds vs 0.3 seconds, a huge difference for traffic as valuable as ecommerce, 2) it doesn’t even cost you that much. Extensive cost analysis later in this essay. It’s only a few dollars unless you’re running a very large traffic website. Short answer I’m on and the . Twitter @SpeedTestDemon email list is here EDIT: I just read on this random tech website (TechRadar) “no doubt prices can be high”. Ugh, it’s just trash content. It’s written by someone who just wants to fill up word count. It’s not written by someone who’s actually operated a website. Come on and actually do the math. If you’re the typical small-time website, you CDNs cost you close to nothing. If you actually have a lot of traffic, it end up getting to $100/month — hardly more expensive than whatever you’re already paying for hosting. Apologies for the mini-rant. know might EDIT 2: just looked up techradar’s monthly traffic and holy crap they have 52 million visitors a month for writing that kind of trash content? Unbelievable. Makes me feel like ranting. Why do people put up with trash content. I will write a rant about this later. : rest of this essay. Longer answer The main focus of the rest of this essay is going to be discussing the pricing advantage of free CDNs over premium CDNs. Regarding the other dimension of performance — well I’m already going to be discussing that in 20+ articles analyzing 20+ free CDNs. A brief summary thus far: Premium CDNs are obviously way faster than free CDNs — easily 3 times better. It’s not even close. And it makes sense, right? A fast CDN like CloudFront is hugely expensive to maintain high performance across terabytes of data, thousands of servers. This is really hard. Imagine debugging and maintaining thousands of computers. And then making sure the servers have high availability. Really hard. So the only advantage that free CDNs have over premium CDNs is pricing. It’s true: a premium CDN will cost you money. But how much is it really? Let’s explore the cost question for the rest of this essay. To find out how much a CDN will cost you, all we need to find out is how much data your website transfers. This is because the metric that CDNs charge on is the amount of data transferred by the CDN. Cloudfront, for example, is around USD $0.11 per GB (BunnyCDN is a cheaper alternative at $0.04 / GB. Not sure whether it’s actually faster, I will be speed testing Bunny CDN shortly!). Let’s estimate how much data you’re going to be using if you’re in either one of these two buckets: You already have a working website You’re still building out your website How to find how much a CDN would cost you if you already have a working website If you’ve already got a website setup, there’s a simple test you can do. Go to your website. Open up Chrome Developer Console. Click the Network tab. Ctrl+R or Cmd+R in order to reload the website. Then you see a number where it says how much data got transferred over the network. This is how much data the CDN will charge you. Here’s an example screenshot of where to find the data transferred in the Chrome Developer Console tab. Look for the red circle. This example is for facebook by the way. Please note you can safely ignore the “X MB resources” number. That number is usually far larger than the “X MB transferred” number because “X transferred” is highly compressed. For example, a 1 MB photo is usually first compressed (probably gzip, although brotli compression recently released by Google is starting to gain adoption) before your browser downloads a 500 KB photo, which gets decompressed back to 1 MB. Web servers do this in order to speed up how long it takes for the web page to render for you. OK once you’ve got the number, jump down to the section . analyzing the cost of the premium CDNs How to find how much a CDN would cost you if you are planning to build a website If you’re still planning out your website, we can still get a good estimate of how much data you might transfer by looking at a few representative examples. The first best thing you could do is to do the “data transfer” test (written about in the previous section for a working website) on somebody else’s website already out there that is most similar to what you want to create. Another good rule of thumb is to assume your average web page has a size of 2.07 MB. Pingdom has conducted a study on the average web page size. They have all this data since millions of people use Pingdom to test how long their website takes to load. They found the average page size is 2.07 MB. Lastly, you could go through a bunch of case studies below where we analyze how much data was transferred for each example. You can pick out the example that represents your goal most closely and use that number to calculate your estimated CDN cost. Note: case studies are conducted with an adblocker, which means data transferred could be much smaller than without an adblocker. Case Study: How much data does a Shopify store use? Example: (they have awesome olive oil, you should buy from them) https://seloolive.com/ Data transferred: 1.2 MB Case Study: How much data does a minimal wordpress website use? Example: (my own website, haha, which runs on WordPress) https://speedtestdemon.com/ Data transferred: 203 KB (wow that’s small!) Case Study: How much data does a complicated custom ecommerce solution on WooCommerce use? Example: the product page of an ecommerce store doing 7 figures in revenue per year. Link: https://www.solioswatches.com/product/solar-black-black-mesh?caseColor=matte-black&dialSize=36mm&gender=men Data transferred: 4.6 MB Case Study: How much data does a simple sports betting web app use? Example: this is a simple web app that does optimizing. Link: https://www.topflightfantasysports.com/pga/pga-fd-optimizer/ Data transferred: 4.4 MB Case Study: How much data does a big complicated web app like Facebook use? Example: facebook.com Data transferred: 3.8 MB Case Study: How much data does a simple landing page use? Example: https://sizzy.co/ Data transferred: 516 KB Case Study: How much data does a portfolio website use? Example: http://www.baileyritchiephoto.com/ Data transferred: 1.8 MB Case Study: How much data does a magazine website use? Example 1: typical article techcrunch.com Link: https://techcrunch.com/2018/09/11/comma-ais-george-hotz-ousts-george-hotz-as-ceo/ Data transferred: 2.4 MB Example 2: oil price news site Link: https://oilprice.com/Geopolitics/Middle-East/Could-Energy-Ties-Bring-Saudi-Arabias-Conflict-With-Iran-To-An-End.html Data transferred: 650 KB (oh interesting, MUCH smaller than techcrunch) Example 3: bitcoin news site (because who doesn’t love bitcoin??) Link: https://news.bitcoin.com/latin-american-countries-bitcoin-adoption-el-salvador-bitcoin-law/ Data transferred: 3.6 MB So how much does a premium CDN really cost you? Once you’ve got the number of how much data you’re transferring per pageview, now find how many pageviews you’re getting per month (or per day, your preference). The math is just a simple multiply: (number of pageviews * amount of data per pageview) = total data per month (or day). OK now what is the cost of transferring that data? In the case of Cloudfront, you can just use the highend number of $0.11 / GB (they vary their pricing per region). If you rather use other premium CDNs, here’s the full list of the most popular ones (it’s a crowded market, there’s probably many premium CDNs I missed. Drop me a comment on twitter or email if you want me to include your favorite CDN). CloudFront: $0.11 / GB Cloudflare (the premium version, not their free tier which also has bad performance like every other free CDN): cheapest premium plan is $20/month. Fastly: minimum charge $50/month. BunnyCDN: $0.01 / GB for North America & Europe, $0.06 / GB for Africa. KeyCDN: minimum charge $0.04/GB to $0.11 depending on your location. StackPath (MaxCDN): Starting from $15 Akamai: flexible pricing model starting from $0.087 / GB Microsoft Azure CDN: Starting from $0.081 to $0.233 depending on location. CDN77: starting at $199/month ($0.033 / GB). Leaseweb: minimum of $49/month ($0.06 / GB). Have you finished calculating how much that premium CDN costs you? If you’re a small-time website, you should probably go with CloudFront (or KeyCDN or Azure) which allows you to pay a much smaller upfront cost than some of these other premium CDNs like Fastly. $50/month is ridiculous to pay for a CDN if you’ve got barely a few thousand traffic. Let’s say you’re a website doing 10,000 pageviews/month, which is an amount most websites don’t even get to. The math is 10,000 pageviews/month * 2 MB/pageview = 20 GB/month of data. Then 20 GB/month * $0.11 /GB = $2.20 / month. Basically a pittance. Chances are pretty good you’re already paying $5/month for hosting along with many other widgets and plugins. In that example we just quickly analyzed, you’re getting 3 times faster page loading times for 10,000 people for just $2.20/month. It’s an incredible deal. Are there exceptions where a Free CDN is as good as a premium CDN? Actually, yes. I’m currently doing an investigation on free CDNs and I’ve been impressed with some Free CDNs so far. The best one I’ve seen so far is , where Jetpack’s image CDN performed very admirably (as well as CloudFront) while also having a great image compression algorithm. However, very limited to images, when you probably would also like to put HTML, JS, CSS, video, and audio on the CDN. Jetpack CDN (a.k.a. Photon CDN) Another good one is Cloudinary but they have a fairly limited free tier of 25 GB in a month, after which you need to pay $99/month. I’m currently writing an article on Cloudinary right now. So the answer is yes, premium CDNs are worth it In my humble opinion, the premium CDNs don’t cost that much. The only problem is if you’re the kind of person who has trouble understanding how to set up a CDN. Am I wrong? Feel free to argue with me on or in email (hello at speedtestdemon dot com). I’m willing to change my mind if your argument’s good. Twitter

Is it worth it to pay for a premium CDN?

We are six weeks into 2019, and I’ve already been to three conferences that were all unique in their own way. I wrote about my , which was billed as “Bitcoin Not Blockchain” and featured mostly self-described Bitcoin maxmillists and HODLers. I was in Mexico for an invite-only conference, which I attended as a guest but was not directly invited to, so out of respect for the rules I didn’t write about it. Finally, I just got back from , which happened to take place in one of my favorite cities- HotLanta. The Atlanta Blockchain, or TAB, is an east-coast boulder in the Blockchain earth and even features their own Bitcoin Embassy. The organizers did a great job and I have to give a shout out to Michael Tidwell, who went to great lengths to accommodate all the egos and personalities who landed in the great state of Georgia. Vegas experience at the Unconfiscatable Conference TAB Going into the year, I could easily say I was most excited for the TAB conference. Of course, it didn’t hurt that I was invited to be a speaker. But, in reality, my fiancee and I had decided to go to the conference the minute we heard about it because our homie Dan Anderson was going to be setting up a booth there and introducing the world to . Sasha and I didn’t exactly meet through Bitcorns gaming but we certainly forged our relationship during the early days of the weird crypto-centric gaming platform and thus, it has a special place in our hearts. Which leads me to what you are probably thinking… Bitcorns Bitcoin Weird Behold the Boosters at the Bitcorns Booth. Yes, this term has been thrown around for years, but we missed the Bitcoin weirdness during the 2017 bull-run and then last year’s crushing bear market. You see, the VCs, bro-vestors and normies invaded the space and kind of ruined Bitcoin gatherings for me. Blockchain conferences became the norm and with it came scam coins, money grabs, pump and dumps and off the chart hype that fueled some of the most manic behavior I’d seen from grown people. I literally cringed at a local Washington D.C. event during the ramp up in 2017 as 40–60 year old men stood on stage and shouted with excitement how blockchain was going to change the world. As my comrade Rob — a Bitcoin weirdo himself — exclaimed, “this is like an MLM (Multi-level marketing) event. These guys are the ones who missed out on the internet pumps of twenty years ago and are certain not to miss this boat.” I’m proud to report that I believe most people in the space have sobered up or have left to go drive their Lambos for Lyft. So let’s dive into the event… I would reckon about 400 or so people attended this conference and Tidwell reported that it nearly doubled in capacity from last year’s inaugural event. That is quite phenomenal considering that last year in February the price of Bitcoin, Ethereum and Monero were several times more valuable than they are today. A wide range of views were set to be on display during this conference, speakers and panelists like Jameson Lopp, Chris DeRose, Dan Anderson (or as I like to say- Danderson,) Carol Van Cleef, Paul Sztorc, MikeInSpace, Daniel Krawisz, and many more were booked to give up the goods to the crowd. Of course, I also would be dropping the cryptoart jewels (toots the ham horn.) Now, keep in mind this conference had four ballrooms and they all had panels and presentations going on simultaneously. I had to pick and choose what I saw and therefore, you are now entering my TAB silo. TAB Silo. Get it? The first talk I saw was around 10:30 am on Saturday morning. It was in the meme-centric Ballroom A and featured the aforementioned Dan Anderson aka Droplister presenting, “How I Stopped Worrying and Learned to Love the Mob.” This presentation, although very enjoyable, more than likely had some college students in the room scratching their heads at first. There wasn’t a mention (that I can recall) of blockchain or crypto and it featured clips of Dr. Strangelove and analogies of crosswalks. But, in a true Bitcoin-weird twist it all started to make sense…I think. What I took away from the talk was that a story can be projected a number of ways and trusting in one silo to make a decision is dangerous. Don’t profess that you are “willing to die on this hill” unless you’ve truly dove in. How it related to Bitcoin for me? Bitcoin Cash, BitcoinSV, Bitcoin, Bitcoin Private and the rest of the forks all have communities that can paint the picture that their Bitcoin is the best. Proceed with caution. The next talk at 11:00 was “Blockchain and Art Beyond the Hype Panel” moderated by Theo Goodman. The CEO of Portion, Jason, the head of the Artists’ Liberation Front Johnny, Theo, who runs theos.gallery, and I were all grouped stage-left in a tight half-moon of blue chairs you may find in a cafeteria. We all know one another, but that didn’t stop us from speaking our minds while fielding Theo’s questions. Some divergences on our thoughts in the space were that of ERC versus Counterparty tokens and whether or not physical art could or even needed to be tokenized. We got to shill some of our cryptoart at the end and that was fun. After the panel we fielded some questions and audience member Chris DeRose asked us to give our thoughts on the whitewashing (or as I like to say “greenwashing”) of Rare Pepe Trading in the space. We all agreed that it was a thing. It’s hard to write skeptically about this panel because I was on the other side but the feedback was good. Johnny Dollar making cents (sense) on the Blockchain and Art Beyond the Hype Panel. After the morning panels and a quick lunch with the wifey, it was back to Ballroom A for me. Most people who aren’t in this space would definitely consider the morning panels I attended weird. Well, shit was about to really hit the fan in the afternoon. Insert mindblown.gif here… The infamous Chris Derose was up next with “Dont Just Stand There And HODL, Do Something!” Of course not being one to back down from anyone and their opinion, I went into the room proudly sporting my HODL dad-hat. (The HODL meme is a whole other debate about how people define it’s meaning- maybe a new article coming soon?) Chris has the ability to really ruffle feathers and probably has collected more blocks on the Twitter Blockchain (a pun about crypto people blocking on twitter) than any other meme-ber of the Bitcoin society. I’d bill him as a meme philosopher, twitter troll, shameless promoter and a polarizing and thought-provoking Bitcoin leader. When I was mulling around this space a few years back I discovered “Bitcoin Uncensored” with Junseth and Derose and it really was the craziest podcast/youtube from the crypto-sphere. If you’re over 21, admire 1990’s style comedy, are not afraid of fart jokes and want to learn Bitcoin then you have to check out the old episodes. I’m not as up-to-date on the newest content but it doesn’t seem to matter as Chris, according to his presentation, is going to start doing more live trolling and IRL debating. He is sick of the saturation of social media and his “block-chain” on twitter is limiting his engagement. He repeated a famous Gil Scott-Heron line a couple times, “the revolution will not be televised” and it really resonated with me in terms of the arm chair social justice-ing going on all over the internet. Oh yea, he took the stage in a train conductor’s jumpsuit then a few minutes into his talk he proceeded to strip down into a mankini looking wrestling outfit. Bitcoin-weird 10/10. Not the Super Hero We Asked For But Bitcoin God Works In Mysterious Ways. “Is Bitcoin Car Talk Fake News?” was the next panel I caught. I felt like I was having deja-vu (because as a fan I’ve seen all of the episodes) throughout the half hour long laugh fest but I loved it. You see, Mike basically video-archived his way through his presentation with a self-promoted display of clips and episodes from his youtube series of short funnies called . This was shilling at its finest and the room was in hysterics as his unique brand of humor cut right through the divided room of Blockstreamers, Craig Wright admirers and Roger Ver fans. Mediation through comedy could be the answer to making Bitcoin great again and I nominate Mr. InSpace to write the plot. Bitcoin Car Talk MikeInSpace Looking At Himself Looking At The Crowd. Next up was a panel that featured a man with a bullet-proof vest, the self-proclaimed “Emperor of Bitcoin,” a self-described “toxic, Bitcoin extremist” and a Bitcoin Cash ABC evangelist. Bitcoin is fucking weird and this panel, “The Battle for Bitcoin” was only missing Jose Canseco to clear all of the bases. Quotable after quotable spewed from the panelists and I really don’t know where to begin or how to describe this panel as it was both entertaining, gut-wrenching, thought provoking and interesting to hear the arguments made in favor of the Bitcoin Cash chains. BitcoinSV was the dominating topic as Daniel Krawisz and Kevin Pham — who rocked the kevlar vest — were the most charismatic of the bunch. The Bitcoin extremist, Zack Voell, ended up mainly moderating the panel and that left Collin Enstad to defend his Bitcoin Cash fork against the two SV supporters. Kevin, who only a couple months ago was a Bitcoin maxmillist and Vitalik Buterin’s number one hater, has now pulled a Hulk Hogan and gone to the New World Order side of this twitter-finger wrestling league. He claimed he is “public enemy #1” and was wearing the bullet proof vest to prove a point… I think it was about the dangers of speaking out against the elite and how high stakes this space is getting. He mentioned how he was amazed that he was considered a leader in the space since all he discovered he’d been doing was parroting ideas from his oracles. If Zack is a BTC extremist then Daniel is a BSV extremist and said, “the reason ppl care about the name BTC is because the name is all they have” and “Bitcoin is a social order.” He added that the exchanges are Bitcoin’s enemy because Bitcoin will take over the world and exchanges won’t be needed. Collin mentioned that Bitcoin’s decentralization was a spectrum and “Bitcoin is defined as money.” The idea of marketing was the only thing really splitting up Daniel and Kevin’s viewpoints. Kevin and Collin think marketing the name is part of the process and Daniel thinks marketing is bad because it’s just people wanting to use the marketing to sell one’s own Bitcoin. The whole panel was a real mind bending Tour de Bitcoin forks and although entertaining and somewhat educational it didn’t seem to sway many in the room to the dark side of the forks. At the end of the panel they took a vote by having the audience raise hands of which Bitcoin chain was the “real” Bitcoin and the BTC ticker was still the overwhelming victor. A Bitcoin Extremist, A Bitcoin For Coffee Guy, a Big Blocker in a kevlar vest and The Emperor Of Bitcoin Walk Into A Bar… Because we were part of the event, Sasha and I were invited to have a VIP dinner with the speakers and panelists. A sweet 5 course meal consisting of salad, salmon, steak, chicken and pecan pie on top of a ten story rooftop overlooking Atlanta was a great way to finish the day. We sat next to Jameson Lopp, Wayne Vaughn, Paul Puey, Brian Hoffman and James Poole so the conversations were less meme-centric than the TAB silo I had been in earlier in the day. I listened mostly as lightning, side-chains and wallet security was discussed and then proceeded to have a nice chat with Lopp about Lightning and how people need to learn Bitcoin before we see a Lightning network effect with payment processors. The afterparty was lit, I heard it involved shots, shots, shots and a few edible lollipops. Sasha and I went home because we were running on fumes and she had a to finish up her big presentation that was scheduled for the following day. On Day 2 of the event we arrived a bit later and Justin Wales was the first presentation I saw. I’d seen Justin speak a few weeks prior in Tampa and really enjoyed his points and hot takes that seemed to differ from some of the other panelists. He is a first amendment and blockchain lawyer with Carlton Fields and his presentation “The First Amendment as Applied To Bitcoin and its Lightning Network” was a real interesting dive into Bitcoin as speech. He hammered home some interesting points and I’d say it was probably my favorite hour of the conference in terms of learning new things. For instance, Bitcoin’s Lightning network can allow people to transact 1/236th of a penny- capturing a whole new way of expressing payment. My man Cryptograffiti just did an interesting art experiment with Lightning by selling the cheapest artwork in history. To be clear, this is stuff that is not doable without Bitcoin. Justin drove home a point that regulators need to be cautious of how far they go because some of the experimentation going on could be ground breaking. He really explored Bitcoin as speech. The forks kind of painted a picture for him that “the bitcoin network is, at its core, an association of individuals” and if you don’t agree with something in the protocol then you are free to run a new or different version. I recommend everyone watch the video of his presentation when it drops. UCIT Lunch was way better than I expected. I inhaled a small ocean of curried fish, and got busy on some turkey sandwiches, chips and salad. “Blockstream and Bitcoin” was the next panel and I got a nice summary of what is doing from the Product guy, Allen Piscitello. Lightning, Liquid side-chains, Green Address wallet, Schnorr signatures and Bitcoin via satelite were the main points he discussed. All super interesting topics and Allen was thorough with his delivery. Before he started the presentation he pointed out that he had a hater in the crowd- MikeInSpace, which I translated as he had been trolling the Blockstream boys lately. The basic narrative I left with was that Blockstream is trying to make Bitcoin more secure and that they’re all about improving financial privacy. Blockstream Sasha Hodder dripping with Cryptoart memes. The final panel, was my favorite. I may be biased but it was presented by Sasha Hodder who happens to be both a supporter of mine and a cryptoart enthusiast. (Oh yea, she’s agreed to marry me as well.) Her talk started off with a new and edgy video, “ ” (notice how it’s never written as DJ Pepe folks,) in which DJPEPE (a carboard cut-out) jokingly schools the SEC “officers” (played by Tatiana Moroz and Sasha.) She proceeds to make the argument that DJPEPE and many more cryptoart projects possibly should fall under the UCC’s definition of a tangible good and not the SEC’s definition of an investment contract. Her performance even earned her a new CORNLAWYER asset for Bitcorns.com by a farmer named Bench. “When Howey Met Forman- Questioning The Appropriate Legal Classification for Crypto-Art” DJPEPE vs. The SEC Graffiti rules. Willing to die on that hill. The conference featured a couple dozen booths for sponsors and vendors outside of the 4 Ballrooms. The most popular booth seemed to be the Bitcorns booth as it was set up to display how assets could be used in idle gaming using the Bitcoin blockchain. They had hand-painted artwork of corn memes, Upper Deck-quality trading cards made by Roger Fliporian, stickers and more all on display. I got a taste of the new Haven mobile app by and Brian Hoffman. One of my favorite cryptoart projects was set up outside Ballroom A and talking to attendees about owning fractions of a video that is on display in the MOMA in New York. Bitpay, Bitcoin.com, Johnny Dollar, Monero and many more all had quality booths set up as well. OB1 Snark.art I ate like 20 of these things. Maybe this is the real Killer App for Bitcoin. All in all, this was a great experience and the bear market blues were not anywhere in sight as high level chats took place all over. Bitcoin is weird and wonderful and I look forward to next year’s event. stuttering air horn sound* #FreeRoss and the Johnny Dollar Booth had some sick hand painted joints.

My TAB Recap- Making Bitcoin Weird Again (That’s A Good Thing)

Those who understand the benefits of NFTs and how they give you provable ownership of digital items might be surprised to know that gamers, or those who might benefit the most from NFTs THE MOST, become outraged whenever gaming companies even mutter the letters N-F-T. At the same time, it’s easy to see the concerns as gamers are increasingly to extract more money while providing less value, and relatively few understand the benefits of NFTs and instead see them as a cash grab. targeted by predatory tactics Let’s see why gamers are outraged by NFTs and why they should actually greet NFTs with open arms. 1. Smells Like a Scam Crypto has been called a scam and shady finances at best ever since its inception. It also doesn’t help that crypto was often first used by criminals . to buy illegal items from Silk Road Repeated scandals, scams, and rug pulls in the crypto industry have damaged its reputation and made it difficult for outsiders to trust the legitimacy of the technology. Cryptocurrencies like Bitcoin and Ethereum, while different from NFTs, are all based on blockchain technology. For those unfamiliar with the differences, Bitcoin is a form of digital currency that is interchangeable and indistinguishable from one another, while NFTs represent unique digital assets that cannot be replicated and serve as proof of ownership. The may be unsettling to some, but they are actually a natural evolution and offer great potential for the gaming industry. concept of NFTs Many games feature gacha elements where players can obtain virtual items such as skins, costumes, and weapons, but these items do not truly belong to the player. NFTs provide a way for gamers to truly own the digital assets they have purchased. While there are ways of selling the items (there are often marketplaces for selling Fortnite, TF2, WoW, DOTA, and items from other popular games), you don’t actually own them. Selling is and can get your account banned, and your items are completely reliant on the games. If the game dies, then you have nothing proving that those items belonged to you or even existed. often against the ToS Despite the negative media coverage and justified criticisms of the crypto industry, the technology itself is not inherently bad and has the potential to be game-changing. It is important to separate the actions of some bad actors from the potential of the technology itself. 2. Just Another Revenue Source I don’t want to pull the “back in my day” card, but I’m going to. When you bought retro games like Sonic for Genesis, Mario for NES or SNES, or Resident Evil for PS1, you bought the entire game. It wasn’t a half-finished game that wouldn't get patched for months, there was no DLC, and there certainly weren’t any items, loot boxes, or other microtransactions. Then came the dark days of early microtransactions when companies knew they could make more money, but many didn’t know how to do it. Remember when essential items would be kept behind large paywalls, or when you’d have to grind monotonously for hours before getting the smallest amount of in-game premium currency? Most developers now understand that it’s best to give you additional items, costumes, skins, and other goodies as microtransactions. In fact, last year, gaming companies made more than . $40 billion by selling in-game items The digital cosmetics aren't essential to gameplay, but they add to it. However, many gamers find themselves spending tens or hundreds of dollars on these items (even though the game itself is usually free), and it’s become a necessary evil that gamers have grown accustomed to. While the potential for increased microtransactions is a concern for gamers, NFTs offer the possibility of true ownership of digital assets and the ability to prove ownership through the use of blockchain technology. NFTs, which can offer added security, censorship resistance, and interoperability to a digital asset, could potentially be very beneficial for gamers. This is because it would allow them to sell their in-game NFTs once they are done playing the game while also reducing the risk of getting locked out of an account and losing access to assets. It is not yet clear if developers will take advantage of this by charging a premium for NFTs, but it presents an interesting opportunity for both players and developers. 3. Over-Financialization Takes Away Fun Some gamers already understand what NFTs are and how they’re beneficial, but they are still worried about them. If you own an NFT, you have the right to sell it. That means that there will likely be large marketplaces where gamers can sell items en masse as they own them. While cashing in on your in-game achievements or selling items you no longer want sounds great (and it might be), the problem is that this can turn entertainment into work. Getting a rare item feels great. Maybe it’s stronger than other items, or maybe it looks so unique that other players will know how special you are. In either case, these items make the game more enjoyable. If NFTs become widespread and your main goal is to sell items as soon as you get them, then games become work. You aren’t farming items to improve your character. You’re farming items, beating bosses, and doing quests for your daily living. Suddenly having fun in a fantasy world turns into a colorful office space. While some players might experience this, the truth is that it's already happening. Like I said before, you can sell items, but it’s technically against the terms and might even be illegal. NFTs will just make it easier to sell items and it will give you more rights, but that doesn’t necessarily make it a job. Final Thoughts Gamers are showing their anger whenever developers mention NFTs, but this would actually be highly beneficial for them. Not only does it give you true ownership of a digital item and give you more rights in selling and owning these items, but it also feels good to know you truly own the item. It’s not just some random item floating in digital space. It’s an item that you lawfully own. If you felt the sting of NFT rage, then I recommend you reconsider your position as this can be a huge benefit to gamers.

Three Reasons Why Gamers Are Outraged at NFTs

has become the norm among review sites. I don't know if a review site exists without any affiliate links. On one side, people hate this because the reviews are poisoned by green. But, on the affiliate's side, why would they write something without getting compensation? Affiliate marketing Let's look at why affiliate marketing gets a bad rap, and the major difference between the good and unethical sites. Why People Hate Affiliate Marketing You are curious about buying a product, but you don't know if it's the right fit for you or if the product lives up to expectations. So, you do what anyone else does and search for a review. However, all you find are articles saying how amazing the product is. Most of them are only rehashing what was said on Amazon and not adding anything new. Did they even use the product? This is the most common type of affiliate site and it’s . You’re looking for a real review, not to be sold something. While this is the most common type of affiliate, not all are like this. why people are frustrated with affiliate marketers The truth is that affiliate marketing makes it possible for many writers to make a living online. I'm not condoning artificial or fake reviews (as I will explain later), but it's a double-edged sword. If you're careful about your source, then an affiliate website isn't necessarily bad. However, you have to make sure the person used the product and is giving you an honest review. Good Affiliate Marketing I’ve been focusing on review sites up until now because those are the most common affiliate sites. I’ll still be using them as an example, but I also want to bring up here as they are doing a wonderful thing. cashback sites Cashback sites , Ibotta, Honey are all affiliate advertisers. However, there is a major difference between their model and review sites. With a review site, a potential buyer clicks on a link and buys a product. The commission goes solely to the affiliate. Cashback sites do things differently. like Rakuten They start in the same way. They direct you to a website or store, you buy stuff, and they get a commission. Here's where it changes. Instead of taking the whole chunk of the pie, they share 50% with the buyer. If a cashback site earns a $10 commission, the cashback site keeps $5, while the buyer gets $5 back. Depending on the cashback service, this can work with physical stores, online stores, or a mix of the two. There are also honest review sites that do good affiliate marketing too. These are people who have bought and used the product. You can tell the difference because they'll use photos and videos of them holding and using the product (not just stock images that are taken from Amazon), and they'll tell you if something is bad, even if it hurts their end commission. A good affiliate adds value to the reader's buying decision. Bad Affiliate Marketing These are the affiliate sites you're , unfortunately. These are the majority of affiliate sites, but they often get pushed down and are relegated to finding keywords with very little competition (that's how they get traffic, not from quality content). These websites are known for two things: dishonest reviews and pushing any product or service. more accustomed to Dishonest reviewers typically don't buy or try the product, they only write about its features and make it sound like they've used it. Some marketers take info from other honest review sites while others will just rehash what they found on Amazon. If they aren't adding anything new or are just fluffing up the review, then it's likely that the writer didn't use the product. They'll also push any product or service, even if it's bad or useless. This is , get rich quick, and health products (raspberry ketones, anyone?). Web hosting commissions are insane. $100 referrals are the norm. This pushes many affiliates to climb over each other for your attention. They don't care if the service is good, they just care about that commission. especially common with web hosting Aside from rehashing, you’ll notice that every product they review is THE BEST EVER, NO FLAWS! That simply isn’t true. Even the best product has one or two small errors. The only time these reviewers will say something is bad is if it makes a product with a higher commission look better. Sometimes it’s hard to tell, but often these marketers will spend more time talking up a product than actually reviewing it. Ethics I've already said that honest reviews are from those who buy and use products, but let's go a little deeper. Can anyone honestly review a product when money is on the line? An affiliate marketer is screwing with their commission check if they say a product is bad. All that time put into buying the product, using it, and writing up a review just so you won't click the "buy now" link seems like a stupid idea. A real review can take hours to write, not to mention all the time placed into actually testing the product. While this might enhance their authority, it won't improve their commission check. This is a person by person debate, but what do you think? Do you think anyone can be honest when money is on the line? I believe affiliate marketers can be, but it’s a minority. However, despite the wasted time, I think authority is worth its weight in gold. I'd rather be believed and honest than promoting some garbage product that doesn't help anyone. Sometimes this means getting several products that won't make any money, but that's the price you pay for honesty and building long term value. Conclusion Affiliate marketing gets a bad rap and it's well deserved. There are far too many people pushing bad products (or never even testing the products) just to get a commission. While it's hard to be honest when money is on the line, I believe more affiliates can do this. It might mean investing in products and sometimes saying that certain products are bad, but that's what a review is all about. You need to be honest and tell people what the truth is.

Affiliate Marketing Can Be Dangerous But Is It Always Bad?

Smart devices are great. They help you search the internet with only your voice, allow you to order essentials without using your phone or computer, and let you see who's at the door. But they also do one other thing. They are listening, all the time. While Alexa is best known for this morally dubious listening, it's not just them. Google has just been shown to do the same. It seems like every smart device is listening in to your conversations. Let's talk about potential concerns and why these companies are doing this. Every Conversation and Interaction Unless you set everything properly, which we’ll discuss later, every conversation and interaction around the smart device is being recorded. This includes routine conversations you have with friends and coworkers to fights with spouses and top-secret conversations you wouldn’t want anyone to know. Also, remember that much of your interactions with smart devices include searches. Do you really want a data worker going through your online searches? It's a scary thought. This also includes video on devices like Ring or anything else with a camera. Anything you’re doing around the device is being recorded and, more importantly, reviewed by that device’s data team. Activating the System According to these companies, your conversations with the smart device should only be sent and reviewed by real people if you activate the system. This includes your common commands like "OK Google," “Alexa,” or "Hey Siri." However, it also activates whenever you say something similar that might "trick" the system to think it's needed. There seem to be conflicting reports about how often Google, , Alexa, and Siri are listening to you. Some reports say it’s all the time while others say it’s only when you activate the system. In either case, you’re allowing a lot of your interactions and conversations to be recorded. Ring Anonymized One comforting thought is that , to a point. The information should be wiped of any personal information when it comes to passwords and login information. It also shouldn’t collect your name. the data is anonymized However, reports show that many of the conversations being recorded have personal information recorded like names, addresses, and less than savory details you wouldn’t want others listening to. , they were able to easily identify people in some of the recordings they listened to and even confronted some of those people with their recordings. According to VRT News Reason for Listening If you’re being recorded, then you should at least know why. Accordingly, even though Google, Amazon, and Apple are “inadvertently” collecting the juicy details of your life, they don’t care about what you’re saying. It’s not what you’re saying, but how you’re saying it. All of these devices rely on AI and algorithms to deliver the right information to you. Human speech can be difficult to conceptualize by a robot. These companies are so that they get the right responses. This should lead to a smoother interaction with your home device and make them more useful overall. attempting to better attune smart devices to how people ask questions This can also be used to improve search engines and website relevancy. By understanding how you search for products or information, these services can improve to bring you more relevant results. That’s great for our digital world, but scary that very personal conversations are being used in this manner. Stop the Recordings While limited, there are a few ways to stop or minimize the recordings. Because this issue has exploded in the last few months, these companies are making it a little easier to stop the recordings. For example, Amazon recently of their conversations. This doesn’t stop the recording, but it prevents other humans from hearing them. announced that Alexa users can opt-out of human review Many of these companies ask you to opt-in to the recordings being stored and listened to, but this is buried under other terms and most people just skip over them. Be sure to go over the terms and see if there's anything you can opt-out of. Unfortunately, some devices like Google Home will punish you by prohibiting certain features from activating if you opt-out. Another thing you can do is turn the microphone off whenever you’re not using the device. Much like losing out on certain features, this is inconvenient. It means you’ll be turning the microphone on and off, plus it will still record any conversations that deal with commands or anything that activates the system. Agreeing to Change Since people are very worried about their conversations being recorded, Apple, Amazon, and Google have all apologized and have agreed to change how they review these conversations. Apple said that they will be stopping manual reviews for a while until they can get the review process up to their high standards. There have also been promises by and instead only use their own employees to listen to conversations. While changing the process is great, it doesn’t change that your conversations are still being recorded and have already been recorded. Plus, many of the promises are either vague or only switching one bad thing for another (ie: subcontractors to employees, it’s still people listening to your conversations). Apple to stop using subcontractors Staying Safe This doesn’t mean that smart devices are terrible or going to destroy your life. The truth is that less than and unless someone wants to search you out, they aren’t going to confront you with the recording. 0.2% of conversations are ever reviewed It’s scary, but if you review the terms and just make sure you don’t say anything personal around the device, you should be fine. While it’s not ideal, if you’re careful around the device, then there shouldn’t be any problems. Conclusion They make it easy to search with your voice, see who's at the door, or order products with ease. However, they are recording conversations and a very small percentage of them are reviewed by other people. Many of these companies have promised to change their process of collecting and reviewing recordings, but none have promised to stop them entirely. Just be careful around your device, opt-out when possible, and you should be perfectly safe. Smart devices are incredibly convenient.

Are Smart Devices Worth the Privacy Risk?

I have figured out a way to develop the popular pong game on the Arduino board. The game is not complicated but interesting to develop and play. Here 0.96 inch OLED display and two buttons have been used. The display is small but enough for our project. Two buttons will be used to move our racket up and down. The The first and last 16 pixels will be used to display scores. display size is 128x64 pixels. The scores will be displayed vertically. One racket will be controlled by the player, and the second one, the opponent will be controlled by the Arduino. We will write code so that it will try to move towards the ball. At the beginning of the game, it will move slowly, and then, it will gradually accelerate. We will control the ball direction through ball_direction_X and ball_direction_Y variables. The ball will move in a specified direction each moment. If the ball hits a wall, ball_direction_Y will be reversed and the same logic for ball_direction_X and rackets. The project code is published on my GitHub . ‘Arduino Ping Pong Game’ project page Connection Buttons’ one leg will be connected to the GND. The second legs will be connected to pins 6 and 5 of the Arduino. The display will be connected by I2C pins: V - 5V, GND - GND, SCL - A5, SDA - A4. Developing Pong on Arduino Let’s include libraries and initiate all variables we need. Here we have buttons, scores, player, opponent, and the ball-related variables. //oled libraries:
#include <SPI.h>
#include <Wire.h>
#include <Adafruit_GFX.h>
#include <Adafruit_SSD1306.h>
//oled vars:
#define SCREEN_WIDTH 128 // OLED display width, in pixels
#define SCREEN_HEIGHT 64 // OLED display height, in pixels
#define OLED_RESET     4 // Reset pin # (or -1 if sharing Arduino reset pin)
#define SCREEN_ADDRESS 0x3C ///< See datasheet for Address; 0x3D for 128x64, 0x3C for 128x32
Adafruit_SSD1306 display(SCREEN_WIDTH, SCREEN_HEIGHT, &Wire, OLED_RESET);

//Button pins:
const int buttonUP = 6;
const int buttonDOWN = 5;
//button vars:
int lastButtonStateUP = LOW;   // the previous reading from the input pin
int lastButtonStateDOWN = LOW;   // the previous reading from the input pin
unsigned long debounceDelay = 10;    // the debounce time; increase if the output flickers

//GAME vars:
//scores:
int player_score = 0;
int enemy_score = 0;

//player:
int player_position_X = 19; // static
int player_position_Y = 0;
int player_width = 16;
int player_thickness = 4;

//enemy:
int enemy_position_X = 104; // static
int enemy_position_Y = 47;
int enemy_width = 16;
int enemy_thickness = 4;
long enemy_last_move_time = 0;
long enemy_speed_of_moving = 2000;//update time in ms


//ball:
//void fillCircle(uint16_t x0, uint16_t y0, uint16_t r, uint16_t color);
int ball_position_X = 63;
int ball_position_Y = 31;
int ball_radius = 1;
int ball_direction_X = 3;
int ball_direction_Y = 3;
int ball_speed = 8;//9,8,7...1
long ball_last_move_time = 0; In the , we will initiate buttons, Serial, randomSeed, and display. After that, we will show a starter screen and start the game: setup void setup() {
  pinMode(buttonUP, INPUT_PULLUP);
  pinMode(buttonDOWN, INPUT_PULLUP);

  Serial.begin(9600);
  Serial.println("Start");

  //initiate random
  randomSeed(analogRead(0));
  ball_direction_X = -3;
  ball_direction_Y = random(-5, 5);
  //ball_direction_Y = -5;//test

  
  // SSD1306_SWITCHCAPVCC = generate display voltage from 3.3V internally
  if(!display.begin(SSD1306_SWITCHCAPVCC, SCREEN_ADDRESS)) {
    Serial.println(F("SSD1306 allocation failed"));
    for(;;); // Don't proceed, loop forever
  }

  // Show initial display buffer contents on the screen --
  // the library initializes this with an Adafruit splash screen.
  display.display();
  
  // Clear the buffer
  display.clearDisplay();
  
  
  //draw lines:
  display.drawLine(16,  0,  16, 63, SSD1306_WHITE);
  display.drawLine(111, 0, 111, 63, SSD1306_WHITE);
  display.display();
  
  //scores field init:
  display.setTextSize(2);    
  display.setTextColor(SSD1306_WHITE); // Draw white text

  player_score = 8888; // test
  enemy_score = 8888; // test
  print_score(player_score, 0);
  print_score(enemy_score, 115);
  
  display.setTextSize(3);
  display.setCursor(28, 0);
  display.write("Ping");
  display.setCursor(28, 31);
  display.write("Pong");
  display.display();
  
  display.setTextSize(2);
  delay(2000); // Pause for 2 seconds
  
  //NEW GAME:
  
  // Clear the buffer
  display.clearDisplay();
  
  //draw lines:
  display.drawLine(16,  0,  16, 63, SSD1306_WHITE);
  display.drawLine(111, 0, 111, 63, SSD1306_WHITE);
  display.display();
  
  //Write scores:
  player_score = 0; //reset player_score
  enemy_score = 0;  //reset enemy_score
  print_score(player_score, 0);
  print_score(enemy_score, 115);
  
  //Display players:
  //void fillRect(uint16_t x0, uint16_t y0, uint16_t w, uint16_t h, uint16_t color);
  display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_WHITE);
  display.fillRect(enemy_position_X, enemy_position_Y, enemy_thickness, enemy_width, SSD1306_WHITE);
  display.display(); // Update screen with each newly-drawn rectangle

  
  //Display the ball:
  display.fillCircle(ball_position_X, ball_position_Y, ball_radius, SSD1306_WHITE);
  display.display();

  delay(500); // Pause for 0.5 second  
} In the , we have three main functions which will be explained in detail below. loop void loop() {
  buttons_check();

  move_the_ball_and_check_for_collisions();

  move_enemy();
} In the function, we will check if the UP or DOWN button is pressed and update the position of the player's racket accordingly. The racket position change is done in three main steps: Drawing a black rectangle in the old position, changing the position variable, and drawing a white rectangle in the new position. buttons_check void buttons_check(){ 
  if (!digitalRead(buttonUP) && !lastButtonStateUP) {
    lastButtonStateUP = true;
//    Serial.println("UP pressed");
    if(player_position_Y > 0){
      display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_BLACK);
      player_position_Y = player_position_Y-3;
      display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_WHITE);
      display.display(); // Update screen with each newly-drawn rectangle
    }
  }
  if (digitalRead(buttonUP) && lastButtonStateUP) {
    lastButtonStateUP = false;
  }


  if (!digitalRead(buttonDOWN) && !lastButtonStateDOWN) {
    lastButtonStateDOWN = true;
//    Serial.println("DOWN pressed");
    if(player_position_Y < 64-player_width){
      display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_BLACK);
      player_position_Y = player_position_Y+3;
      display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_WHITE);
      display.display(); // Update screen with each newly-drawn rectangle
    }
  }
  if (digitalRead(buttonDOWN) && lastButtonStateDOWN) {
    lastButtonStateDOWN = false;
  }
} In the function, we will move the ball and check for collisions with horizontal walls, rackets, and vertical walls (player win or lose). Before moving the wall, we will check the required time passed from the last move of the ball. I have set the required time as ball speed multiplied by twenty. To display the ball we will draw a filled circle. In general, the ball position update will be done in the same way as the racket position update. move_the_ball_and_check_for_collisions To check for collision of the ball with one of the horizontal walls we need to make sure that the add up of ball_position_Y and ball_direction_Y is in the range between 0 and 63 (not less than -1 and not greater than 64). If so, the ball_direction_Y will be reversed. To check if the player’s racket missed the ball we need to check if ball_position_X is less than player_position_X and start a new round of the game. The same logic for the enemy's racket, only since it is placed on the other side of the field the check will be for more.To check if the player’s racket hit the ball we will check if the ball is inside of the racket. If so, the ball_direction_X will be reversed and a random value for ball_direction_Y will be given. As always the same logic for hitting the ball with the enemy’s racket. void move_the_ball_and_check_for_collisions(){
  //move th ball:
  if(millis() > ball_speed*20+ball_last_move_time){
    //erase ball on old position:
    display.fillCircle(ball_position_X, ball_position_Y, ball_radius, SSD1306_BLACK);
    display.display();
    
    //set new posion of the ball:
    ball_position_X = ball_position_X + ball_direction_X;
    
    if(ball_position_Y + ball_direction_Y < -1) ball_direction_Y = ball_direction_Y * -1;
    if(ball_position_Y + ball_direction_Y > 64) ball_direction_Y = ball_direction_Y * -1;
    
    ball_position_Y = ball_position_Y + ball_direction_Y;
    
    //draw ball on new position:
    display.fillCircle(ball_position_X, ball_position_Y, ball_radius, SSD1306_WHITE);
    display.display();
//    Serial.print("ball_position_Y: "); Serial.println(ball_position_Y);
    ball_last_move_time = millis();

    //Check for player loose:
    if(ball_position_X < player_position_X){
      Serial.println("Player lose!");
      newRound("enemy");//player
    }
    //check for collision of the ball and the player:
    if(player_position_X <= ball_position_X && player_position_X+player_thickness >= ball_position_X
    && player_position_Y <= ball_position_Y && player_position_Y+player_width >= ball_position_Y){
      Serial.println("Collision of the ball and the player");
      //send the ball to enemy with random values:
      ball_direction_X = 3;
      ball_direction_Y = random(-5, 5);
      display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_WHITE);
      display.display(); // Update screen with each newly-drawn rectangle
      
    }
    //check for enemy loose:
    if(ball_position_X > enemy_position_X+enemy_thickness){
      Serial.println("Enemy lose!");
      newRound("player");//enemy
    }
    //check for collision of the ball and the enemy:
    if(enemy_position_X <= ball_position_X && enemy_position_X+enemy_thickness >= ball_position_X
    && enemy_position_Y <= ball_position_Y && enemy_position_Y+enemy_width >= ball_position_Y){
      Serial.println("Collision of the ball and the enemy");
      //send the ball to player with random values:
      ball_direction_X = -3;
      ball_direction_Y = random(-5, 5);
      display.fillRect(enemy_position_X, enemy_position_Y, enemy_thickness, enemy_width, SSD1306_WHITE);
      display.display(); // Update screen with each newly-drawn rectangle
    }
  } 
} As you noticed above, the function takes as input winner (player/enemy). The input is required to increase the appropriate score. In the function, we will reset all the required game-related variables, print updated scores, display all game objects, and if required update the speed of the enemy and ball. The game will become more difficult as you progress through it. newRound void newRound(String winner){
  // Clear the buffer
  display.clearDisplay();
  
  //draw lines:
  display.drawLine(16,  0,  16, 63, SSD1306_WHITE);
  display.drawLine(111, 0, 111, 63, SSD1306_WHITE);
  display.display();
  
  //Update scores:
  if(winner == "enemy"){
    enemy_score++;
  }else{
    player_score++;
  }
  print_score(player_score, 0);
  print_score(enemy_score, 115);

  //reset gaming vars:
  //player:
  player_position_X = 19; // static
  player_position_Y = 0;
  player_width = 16;
  player_thickness = 4;
  
  //ball:
  ball_position_X = 63;
  ball_position_Y = 31;
  ball_radius = 1;
  //set random direction for th ball:
  
  ball_direction_X = -3;
  ball_direction_Y = random(-5, 5);
  //ball_direction_Y = -5;//test
  ball_last_move_time = 0;
  
  //Display players:
  //void fillRect(uint16_t x0, uint16_t y0, uint16_t w, uint16_t h, uint16_t color);
  display.fillRect(player_position_X, player_position_Y, player_thickness, player_width, SSD1306_WHITE);
  display.display(); // Update screen with each newly-drawn rectangle

  //enemy:
  enemy_position_X = 104; // static
  enemy_position_Y = 47;
  enemy_width = 16;
  enemy_thickness = 4;
  enemy_last_move_time = 0;

  //checking for if we need to update enemy_speed_of_moving and ball_speed
  if((player_score+enemy_score)%5 == 0){
    //5,10,15 and so on
    if(ball_speed > 3) ball_speed = ball_speed - 1; //10,9,8...
    Serial.print("ball_speed: ");Serial.println(ball_speed);
  }
  if((player_score+enemy_score)%10 == 0){
    //10,20,30 and so on
    if(enemy_speed_of_moving > 1) enemy_speed_of_moving = enemy_speed_of_moving * 0.9; //2000,1800,1620,1458...
    Serial.print("enemy_speed_of_moving: ");Serial.println(enemy_speed_of_moving);
  }
  
  delay(500); // Pause for 0.5 seconds
} Printing scores vertically is a little tricky. So, I decided to split it into a separate function. Each digit of the score will be displayed one by one and the cursor needs to be updated accordingly. void print_score(int temp_num, int X){ //0/115
  for(int i=48; i>=0; i-=16){
    
    int num = temp_num % 10;
    char cstr[16];
    itoa(num, cstr, 10);
    
    display.setCursor(X, i);
    display.write(cstr);
    display.display();
//    Serial.println(cstr);
    temp_num = temp_num/10;
    if(temp_num==0){
      break;
    }
  }
} And the last one is . It will try to hit the ball. And for this, it will try to equalize its center with the center of the ball. Here we have some supplemental checks to stay the racket inside the screen. In order for such an enemy to be won, its movement speed will be slow in the beginning and will accelerate as the game progresses. moving the enemy racket void move_enemy(){
  //enemy:
  if(millis() > enemy_speed_of_moving+enemy_last_move_time){
    display.fillRect(enemy_position_X, enemy_position_Y, enemy_thickness, enemy_width, SSD1306_BLACK);
    
    if(ball_position_Y < enemy_position_Y+enemy_width/2){
      enemy_position_Y = enemy_position_Y - 3;
    }else{
      enemy_position_Y = enemy_position_Y + 3;
    }

    //checking if enemy is within the wall:
    if(enemy_position_Y > 64-player_width) enemy_position_Y = 64-player_width;
    if(enemy_position_Y < 0) enemy_position_Y = 0;
      
//    Serial.print("enemy_position_Y: "); Serial.println(enemy_position_Y);
    display.fillRect(enemy_position_X, enemy_position_Y, enemy_thickness, enemy_width, SSD1306_WHITE);
    display.display(); // Update screen with each newly-drawn rectangle
    enemy_last_move_time = millis();
  } 
} Conclusion I hope you enjoyed both the development of the game and playing the game. The game is familiar to many since childhood and nostalgia are felt when playing. Developers have a unique opportunity to get to know the game not only from the outside, but to see how everything works from the inside. Here are some ideas for enhancements to the project: Currently, the game is endless. It would be great to implement the Official Rules of Table Tennis to determine a winner. As the speed of the ball increases, make a visual effect so that the ball has a tail. The angle of flight of the ball during the rebound should be made dependent on the place where the ball touches the racket.

CAASie.co

CAASie.co

6 employees

Since 2019

COMPANY RANKING

CAASie.co

EVERGREEN INDEX #1721

See more notable links about CAASie.co

Apply to get your Tech Company News Page live on HackerNoon.com today!

Read More Tech Stories Related to #CAASie.co

CAASie.co

CAASie.co

CAASie.co

6 employees

Since 2019

COMPANY RANKING

CAASie.co

EVERGREEN INDEX #1721

Related CompaniesRank

CAASie.co Story MentionsComplete List

See more notable links about CAASie.co

Apply to get your Tech Company News Page live on HackerNoon.com today!

Read More Tech Stories Related to #CAASie.co

CAASie.co