Vivek Chauhan

Pov: Android Developer Working for 2hrs 1-hr Gradle Build time... (especially in release build+R8) 30 mins - Thinking & Fixing bugs 10mins - Creating Boilerplate classes (must in clean architecture eg: a use case for just calling repository) 10mins - Taking break 10mins - Actually Coding Ps: Gradle build times are very long when particularly working with compose and release build with minifyenabled (R8) #buildtimes #android #kotlin #android_studio

222

24 Comments

“Bhaiya at the end to react native ka code bhi jvm bytecode m hi compile hota h na?” This was a response i got from a developer when asked about how react native works in Android. But reality could not be farther. See, Android DVM is not capabale of running javascript code because these are inherently different technologies. Instead, what happenes is that a JSC (JavascriptCore) runtime that is bundled into android application alongside react native libs which actually loads the javascript code into memory at runtime. And then, there is a React Native bridge native library (.so file) which acts as a intermediator, translating the calls from java/kotlin code and js code. And because of all these intermediary layers translating instructions among languages, the performance for react native is not as good as native applications. Hope it helps.

22

Q) What is the difference between Static Dispatch and Dynamic Dispatch in iOS? Ans: Let's try to understand the difference from a practical example initially, and then we will go deep technically : Suppose there are two food delivery guys: One of them works for Uber Eats and the other one is a direct employee of McDonalds. Suppose they both stay in the same apartment and they both got an order in their respective application and from the same location. Now since Mconald's delivery guy is a direct employee, he is aware of their store and will reach out directly. Whereas, Uber Eat's delivery guy's restaurant is never fixed so he has to check the address/location of the restaurant each time, needing some inquiry as well from the map, the Mall's store map, or from the Mall's help center. Thus the McDonald's delivery partner is fast and direct just like the Static Dispatch, and the Uber Eats Partner is a bit slow even though their address was the same because of are not aware, like the Dynamic Dispatch. 1.⁠ ⁠Static Dispatch: Static dispatch occurs when the method to execute is determined at compile time. Swift heavily favors static dispatch due to its performance benefits, as it allows the compiler to optimize calls by directly referencing the function to execute, eliminating any overhead associated with dynamic resolution. Direct Access: Static dispatch accesses the function's implementation directly. Performance: It's the fastest method of dispatch available in Swift due to its ability to be optimized by the compiler. Usage: Most functions in Swift, including those marked as final in classes, those in structs, enums, and global functions, use static dispatch. Using final can enforce static dispatch for a method in a class by preventing it from being overridden by subclasses. 2.⁠ ⁠Dynamic Dispatch: Dynamic dispatch occurs when the method to invoke is determined at runtime. This approach is more flexible but comes with some performance costs due to the lookup overhead. Virtual Table: Swift uses a vtable for classes that require dynamic dispatch. A vtable is a lookup table used at runtime to resolve method calls. Performance: Slower than static dispatch due to the overhead of determining which method implementation to call at runtime. Usage: Non-final class methods in Swift use dynamic dispatch. This allows subclasses to override and provide their implementations, which the runtime system will invoke correctly. Please add more points and insights further in the comment section. #ios #swift #iosdevelopment #LinkedinLearning

186

36 Comments

Just going through the NamastheyJS series from Akshay Saini 🚀! Found out that 'Hoisting' is not an explicit term defined in ECMAScript! It's more of considered as a default behaviour of JS rather than some special feature. More on this: https://lnkd.in/gbW7aMFJ #JS #frontend

14

2 Comments

How does Swift memory management work for structures In Swift, memory management for structs works differently from classes due to the value type semantics of structs. Here's an overview of how Swift manages memory with structs: Value Type - Structs are value types in Swift. When you assign a struct to a new variable or pass it to a function, a copy of the struct is created. This means each instance of a struct is independent of others. Stack vs. Heap Allocation - Stack Allocation: In many cases, structs are allocated on the stack, making memory access faster. When a function or a scope ends, the stack memory is automatically deallocated. - Heap Allocation: If a struct contains reference types (e.g., classes, closures, or objects), those will be allocated on the heap, and reference counting (ARC) will manage their memory. However, the struct itself is still passed by value. Copy-on-Write (CoW) Optimization Swift uses a copy-on-write (CoW) mechanism to optimize memory usage for certain types like `Array`, `Dictionary`, and `Set`. For example, if two variables refer to the same data, a copy of that data is not made until one of the variables is modified. This optimization minimizes unnecessary copying. - When a struct is copied, Swift only makes a copy when you attempt to modify the data. Until then, both variables will reference the same underlying data. - If either variable modifies the data, Swift creates a new copy for that variable (hence the term "copy-on-write"). Memory Efficiency - Swift's memory management for structs is generally more efficient than classes because of the absence of ARC (automatic reference counting) for the struct itself. Structs don't need to keep track of references like classes do, so they avoid the overhead associated with reference counting unless they hold reference types. Summary: - Structs are passed by value, so copies are made when assigned to new variables or passed to functions. - Memory for structs is typically allocated on the stack, except when they contain reference types. - Copy-on-write (CoW) optimization helps avoid unnecessary memory duplication. - Structs containing reference types are managed via ARC for those specific reference types. #arc #ios #interview #apple #swift #developer #memorymanagement #structure #stack #heap

5

The documentation for implementing shared element transitions using Jetpack Compose is very good. Do check it out. https://lnkd.in/dQbsW2Ke #androiddev #android #jetpackcompose

2

Just went through an interview for an Android SDE role, and it was quite the experience! Here are some of the questions I faced: 1️⃣ What architecture do you follow? 2️⃣ Explain MVVM architecture. 3️⃣ Breakdown the lifecycle of an activity. 4️⃣ Live machine coding round 5️⃣ Tell us about Jetpack components. 6️⃣ How does Retrofit work internally? Each question gave me a chance to dive deep into Android development. Excited to continue learning and growing! #androiddevelopment

8

𝐓𝐡𝐞 𝐫𝐞𝐚𝐥𝐢𝐭𝐲 𝐨𝐟 𝐂𝐒𝐄 𝐢𝐧 𝐏𝐫𝐢𝐯𝐚𝐭𝐞 𝐜𝐨𝐥𝐥𝐞𝐠𝐞𝐬 𝐨𝐟 𝐈𝐧𝐝𝐢𝐚 [Part-3] 𝐖𝐡𝐲 𝐒𝐨𝐥𝐯𝐞 𝐭𝐡𝐢𝐬 𝐩𝐫𝐨𝐛𝐥𝐞𝐦? This problem occurred in its first place due to colleges below 2nd tier not paying attention towards Product based company placements. They take fees of B. Tech and prepare them for BCA jobs. Many parents' hard-work is getting into the wrong hands. So, this is a crime that needs to be stopped. 𝐓𝐡𝐞 𝐒𝐨𝐥𝐮𝐭𝐢𝐨𝐧 Below are some role wise customised initiatives one can take to solve this problem. 𝐓𝐞𝐜𝐡𝐧𝐨𝐥𝐨𝐠𝐲 𝐞𝐱𝐩𝐞𝐫𝐭𝐬 (𝐆𝐃𝐄, 𝐌𝐕𝐏, 𝐞𝐦𝐩𝐥𝐨𝐲𝐞𝐞𝐬, 𝐢𝐧𝐝𝐮𝐬𝐭𝐫𝐲 𝐥𝐞𝐚𝐝𝐞𝐫𝐬) As you are an expert in some domain, you must be having your inspiring story worth sharing. You can work along with the Developer relation team to conduct in-person boot camps for such students where problem solving is taught first, with the help of examples. Specifically curated study material can be created for these students. Note that this has to be done just once. Once a community is built and we have true leaders, this problem will then be solved in an exponential manner. Just a spark is needed. 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐞𝐫 𝐑𝐞𝐥𝐚𝐭𝐢𝐨𝐧 𝐭𝐞𝐚𝐦𝐬 The following is intended for people responsible for monitoring developer relation programs like DSC, GDG, GDE, MSP, etc. We understand and appreciate the efforts of tech giants to start such amazing programs. As you know, the aim of these programs was to promote problem solving attitude, community building, helping each other and sharing knowledge. They were intended to bridge the gap between theory and practical. But the catch here is that colleges where these programs are just started, often fail at understanding these programs in the first place. They fail at community building. Let me give an example of my own college, where DSC started just to organize events, briefing about tech and not mastering them. Students don't sit together to solve challenging problems nearby. Not only this, they even think that merely collecting course completion certificates and incentives will help them improve their CV. Also, people often misuse incentives for trading. We have seen students developing useless Google Actions just to get a Google Home. Not only this, they ordered for peers too. Massive misuse of developer programs!!! 𝐂𝐞𝐫𝐭𝐢𝐟𝐢𝐜𝐚𝐭𝐞𝐬 𝐭𝐡𝐚𝐭 𝐩𝐫𝐨𝐯𝐞 𝐢𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐬𝐤𝐢𝐥𝐥 : In this competitive world, how can one distinguish between one with actual skills and the other just pretending, both holding certificates of course completion? We strongly believe that there must be a certificate that actually proves one's ability to implement those skills. Course completion and implementation of skills acquired are completely different things. 𝐓𝐨 𝐛𝐞 𝐜𝐨𝐧𝐭𝐢𝐧𝐮𝐞𝐝 Solution Link:- faangx.com [Follow and check] #btech #bca #mca #collages #reality #faangx #CSEeducation #privatecolleges #highereducation #careeropportunities #CSengineers

10

2 Comments

Knock Knock!!! Here is the latest video on "Getting Started with Kotlin Multiplatform" Have a look at it and if you like the video don't forgot to subscribe the channel. Thanks Kashif Mehmood for delivering the session. Many more to come stay tuned. https://lnkd.in/dVMP8zuU

13

Pushed a bunch of updates to AutoProject. Caution: The repo is not stable by any means and there will be breaking changes with every commit. Here's the update. 1. Ditched the entire ExpressJS backend in favour of Sveltekit's server side API routes. Yes, we can use Sveltekit as a fullstack framework too! This simplifies the project structure and maintaining by a lot. The POC has 3 API endpoints, which I moved into the API routes natively provided by Sveltekit. So, now, you don't need to run the backend server separately. Just running the webapp is enough. 2. Environment variables are now handled directly within Sveltekit via Vite. This is still secure and no part of this data is exposed to the client side. 3. Revamped the entire outline of the UI with a sidebar navigation, an about page and a settings page. Will be pushing a bunch of configurable settings soon, so that users can tune the application better. Roadmap ahead. 1. Support for Jira, Asana and Plane. 2. Prompt architecture customization (will talk about it in future posts). 3. Support for local LLM inference via LM Studio. 4. UI rendering update for PRD markdown and Tasks JSON to look better. Also, planning to upgrade TailwindCSS to the experimental version 4 alpha release. 🤞🤞🤞

20

1 Comment

Kotlin 2.0 smart casting improvements 🚀 💻 Kotlin 2.0 was released just recently and there are quite a few improvements with the new compiler. One such advancement is the upgraded smart-casting. Now, whenever you check instances of a common supertype - the data automatically gets cast into that type. See my example below 👇 Prior, in Kotlin 1.9 you had to manually do the casting to access the functions of a common supertype. #androiddevelopment #kotlin #tips

422

18 Comments

ABGaugeViewKit now supports Swift Package Manager alongside CocoaPods! Also, I've also updated the codebase to Swift 5 for better compatibility and performance. Check it out on GitHub: https://lnkd.in/dQQ2zCGv #iOSDevelopment #SwiftLang #GaugeView #SPM #CocoaPods

19

1 Comment

That moment when you include different, but hidden, markup for mobile / desktop with different LCP candidates on each. Your desktop LCP candidate is not present on SSR content, only mobile candidate is (css hidden) which makes the Desktop not part of the critical rendering path and gets buried under loads of React hydration events, GTM, data fetching for component and site wide stuff. You get up to 11s LCP on desktop (depending on connectivity and processing power). There's a bug on this page which we are fixing. It's one of those things that are not obvious at first glance. The image is found around 9s and fully loaded and processed around 11.5s.

11

Asked in Fintech Company Frontend Interview Question. Do solve it here by forking this code - https://lnkd.in/g7V7wSEq I was asked to build a better fetch() Javascript web api with following functionalities:- 1. Auto Retry on api failure 2. Max Timeout of apis calls Also, i was asked various followup questions in depth about fetch that i didnt know at the time of interview, but i have shared them in the video. We will be doing full hands on coding and build a better fetch with the following concepts to solve the problems:- - Promises (race or any) - AbortController in JS - Recursion to auto retry ⬇️ Video link is in comments. Do code your own Better fetch and master your Javascript Maching Coding Interview Round Questions for your next Dream Company Frontend Interview. with this clear and practical tutorial! For more questions dont forget to subscribe to my youtube channel (Link in comment)

83

2 Comments

Learnings from Budget 2024 : - have less/no hopes from the govt - keep working harder to beat inflation - hustle & earn more money to afford a good lifestyle Incase you're a Frontend Engineer and want to learn JavaScript, React or Frontend System Design, then checkout NamasteDev.com and upskill yourself today! Upskill Yourself --> Keep Hustling --> Make more money YoY 🚀

5,707

132 Comments

Your Uber ride is possible because of these kinds of services working together: 📍Google Maps - Locates the place 💰 Paytm - Helps with payment 🔥 Firebase- sends the notification 📩 Gupshup - Send SMS confirmation 📤 Amazon SES - Sends the bill to your email This orchestration and communication between the systems is possible because of APIs or "Application Programming Interfaces" ⚡️ APIs are just a mechanism by which one tech system can talk to another. APIs are helpful in multiple avenues: 👉🏽 Extending use cases of your product: Google Maps and Uber work together to help you locate your cab. 👉🏽 Communication between internal systems: Frontend and Backend. 👉🏽 Providing API as a service: Stripe offers payment gateway API as a service. OpenAI offers LLM API as a service. How do APIs work? One system sends a request to the API of another system and in return, the API sends the response. The request has a type and a payload: 👉🏽 Type signifies what should the API do: Add something (POST), Delete something (DELETE), Update something (PUT), or Fetch something (GET). 👉🏽 The payload contains any important information that the API needs. 👉🏽 Some APIs also have headers, which contain the authentication information, so that only the right systems can access the APIs. This request is sent to the end-point of the API: a place on the internet where this API lives. [it's generally a url]. When API receives the request, it takes some action, generates the response, and sends back a response. A response along with the returned data also contains a status code, which signifies if the API request was fulfilled or not. That's scratching the surface of the API world. Original Post Credits: Ankit Shukla Khushal Jhalani #productmanagement #API

42

1 Comment

How does Xcode magically convert print(“Hello, World!”) into an iOS App !!? 1. Writing the Code: Lets write print(“Hello, World!”) in a file say main.swift 2. Lexical Analysis: The compiler breaks down the above code into 5 tokens- ‘print’, ‘(‘, ‘“Hello, World!”’, ‘)’, ‘;’ 3. Syntax Analysis: Abstract Syntax Tree(AST) is created     FunctionCallExpression       ├── FunctionName: print       ├── Arguments:               └── StringLiteral: "Hello, World!"  4. Semantic Analysis: The compiler performs checks on AST to ensure print function exists and can accept a string argument in Swift Language.  5. Intermediate Representation(IR) Generation: The compiler translates the AST into a low level, platform-independent representation of the code:  %0 = load StringLiteral, "Hello, World!"  call @print(%0)   6. Optimised IR: The IR undergoes various optimisations like constant folding, dead code elimination and loop unrolling. Here the optimisation is to directly use the string literal in the call. call @print("Hello, World!") // Optimisation: directly use the string literal in the call 7. Code Generation: The optimised IR(platform-independent) is translated into assembly instructions that can be executed by the target processor. Object file main.o contains below machine code:     mov r0, # HelloWorldString // Move the address of the string into register r0     bl _print         // Branch to the 'print' function     .data     HelloWorldString:        .asciz "Hello, World!"  // Define the string in the data section 8. Symbol Resolution: The Linker finds the definition of ‘print’ function in the Swift Standard Library(‘libswiftCore.dylib’) 9. Address Binding: The ‘print’ function is assigned a specific address(say 0x1000) in the final executable file. 10. Library Linking: Swift Standard Library libswiftCore.dylib is referenced from executable file and not copied into it so that the library can be shared across multiple Swift programs and is loaded to memory at runtime. This keeps the size of executable file small. 11. Handling Dependencies: The linker builds a dependency graph ensuring all required libraries and frameworks are included. Here libswiftCore.dylib depends on libobjc.dylib so both are referenced from the final executable file. 12. Creating the Executable: The linker combines object file, libraries and frameworks into a single executable file(say-HelloWorld) in mach-o format. Above steps are executed when we run below command on Terminal:   swiftc main.swift -o HelloWorld 13. The iOS App(.app file) is packaged as a directory containing the executable file(HelloWorld), resources(like images and storyboards) and metadata(like Info.plist) #KeepCoding #KeepLearning #KeepGrowing

11

#AndroidBytes Published a new video on Mobile System Design for Building Offline-first application. With the example of a TODO application, tried to cover HLD and LLD from Mobile design point of view. Checkout the video here: https://lnkd.in/gT4WGuUc #androiddevelopement #systemdesign #mobile #kotlin #programming #offlinefirst

76

1 Comment

TCA 1.13.0 got released 🚀 What's Changed 👉 UIKit navigation and binding tools. The UIKitNavigation library has been released and is now integrated in TCA. ⚒️ Update Shared to invoke reportIssue. ⚒️ Address a few compiler warnings. ⚒️ Avoid crash when writing to shared user defaults on a background thread. 🏗️ Typo fixes. 🏗️ Update swiftlang repository link. 🏗️ Updated a bunch of DocC references. PS: Please find the link to the full changelog in the comments. --- If you're looking to improve your TCA skills, check out my free interactive tutorial on TCA best practices: https://lnkd.in/dJHaxm4E

21

1 Comment