쉐도잉 연습: Mastering iOS Development: Mock Interview with Ramiro Neto - 영상으로 영어 말하기 배우기

Right.

So thank you for joining, Romira.

And let's just quickly go through some agenda for our smoke interview.

So firstly, we'll just talk about, you know, some do some introduction, then going to be two sections.

First one, more like soft sections of skill, your behavior driven section in a way.

And then the technical part, the technical part, we will split for two parts.

It's one pure task, pure questions, both, you know, classical Swift, iOS, SDKs, questions, and the second part,

more like example testing, let's say.

So I will drop you some piece of quotes, and then we'll just talk about it.

No life -building, no kind of algorithms thingy, just let's say...

Discussion.

Yeah, just discussion, basically, like a classic thing, nothing more.

So that's it.

Let's start.

Let's see how it's going to go.

Yeah, let's start from the introduction.

Let me start it.

So my name is Ildar.

I'm now located in Sweden.

I'm working in the H &M as a lead engineer.

And I really like to drive such sections like mock interviews or in general interviews.

So always nice to see and talk with the new people and see how we can be used to support each other.

Yes, what about you, Amir?

Yeah, thank you for the opportunity.

So I am Amir, I'm from Fortaleza, Brazil.

I have been working with software development since 2012.

I started as Android, but after that, I did the Apple Develop Academy course.

And after the course, I worked with other technologists like PHP,

Muscle, React .js, and others but I got back to focus on iOS in 2017.

And yeah, I have worked with several projects and companies and like telecommunication companies,

banking projects, and also digital wallets, e -commerce.

Yeah, that's it.

Yeah, really, really nice experience, I would say.

You know, especially this kind of different experience from different areas probably gives you, you know, a sense of understanding how things work you know,

I draw it or I end up in a way.

So that's really nice, especially for, you know, agile teams when you need to be agile, you know, I guess.

Cool, cool.

Right.

Let's just then jump to the first part.

So behavior driven questions to warm up in a way.

And yeah, just to start all this.

Yeah.

Since you have some experience, I've seen that you have more than five years experience in the development right

or in general in development you probably had situations where you need to review code of your teammates um

so really interesting to understand how you're doing it

so for example um you have a team of three four people

and you need to review each other right

and you need to review someone's code

so how do you do it here you you have a rule

that uh a pure pr to be approved needed at least two likes.

The team needs approved at least for two two person to be approved.

Yeah, I have to work.

Okay, do you think two person is enough for approving PR?

What exactly?

Two person, not one or not three?

It depends because our team, it's not a big team, you know?

So as you have less numbers from people, I think it's okay in your case.

But of course, if you have a bigger team, it would be better and have more likes.

Okay.

Yeah, yeah.

Fair enough.

And for review, when you're reviewing some word code, you know, usually people trying to see for something specific

or have some we have some pattern to you know scheme the code

and catch some some of the things can you maybe tell

what usually is catching your eyes on the it's maybe syntaxes errors

or some kind of boss symbols in a way yeah here you have a some a partner for CodeStyle.

So we see it in the PR as guardlets, as possible iflets, trolls, or yeah, some things like that.

Camel case, uppercase too.

And also we try to see if the developer is following the clean architecture that it's necessary for the, the improvements.

So that's it what I do.

Interesting.

Right.

And then when you're working in the team, you usually have a designer and some backend developers and you need to,

let's say you need to work on some feature where all these parts involved,

what would you say is best way for communication about new feature

or how to be sure

that some specific feature is uh you know in the best shape

when you're finishing it or you when you you know

when you finish the work how to be sure

that uh all the parts in in the place

when you work yeah it's a it's a it's a very good question it's a complex one you know yeah

and i think uh the design maybe it's the the first parts of the idea.

We, as a mobile developer, works in all the parts from the idea.

And so when the design presents the idea for us,

it would be nice to see if you have some technical implications from the implementations.

And after that, when you are implementing the the the feature

that is approved for design and for us we need to see if the the the backend it's

it's everything all right or what you will need implementing the backends or to to

to develop it in the mobile yeah yeah okay and let's say design is finished for starting the work,

but back -end not.

Is there any way you can still start working or you will wait for back and fully finish the work?

No, honestly, we can start to work and create interviews at least.

But of course, it depends from the case.

But as I assume that you have a BFA, we can close a contract,

you know, like I agree and agree be agreeing it in a contact contact contact sorry

and yeah it started develops the view

and maybe with uh uh uh creating a block or something like

that right right yeah sounds sounds great

100

and let's say the final question for this you know the classic situation

101

when you have some technical depth as unit tests or some refactoring

102

and you also have your features that needs to be delivered.

103

How do you find a balance in that?

104

Or how do you maybe as a senior developer can recommend to your management or team how to resolve it?

105

So you're not stuck fully in the back block with all this technical depth, but also you're delivering the features.

106

Do you have any tips, tricks, how you find the balance between technical depth and actual features.

107

Can you repeat the first synthesis, please?

108

How do you find the balance between the technical depth?

109

So basically tests or refactorings and the features that you need to deliver because, you know, usually management pushing for features, right?

110

Or business pushing for new features to release.

111

But you know that you need to cover something with a unit test or refactor something because in the future, there is some consequences for that.

112

So how do you find this balance?

113

How do you kind of sell it?

114

Honestly, I think that's for a mature team.

115

We need to spend at least like time to cover some tests and fix some bugs.

116

And improve also our code bases.

117

I think I find the balance for it.

118

Like I haven't taken priorizations from the features between the others,

119

things that you need to improve the codes, like the tests and the others factors that you need to do, as I said.

120

Right.

121

So I think it would be a priorization between these ones.

122

Yes.

123

Yes.

124

All right.

125

So I guess we can wrap up with these behavior questions and then smoothly go to technical part,

126

but I'll just give you some feedback after each section.

127

So not to forget about the things.

128

This section, I would say that all gifts for behavior questions, your answer is solid about how you behave, how you do, especially part about.

129

you know, how do you find the balance in the technical part between these features and technical depth,

130

because you always need, you have such experience.

131

I mean, this is something where you can recognize that person has some experience to actually work on technical depth, not just work purely on the features,

132

because you really need to think about it and you really need to

133

reserve some time in your daily work or your sprint or whatever you call it for specifically for technical depth.

134

You could be a bit maybe descriptive in a way for how exactly you can do this.

135

Let's say that you can reserve 20 % of sprint

136

or technical depth or have some iterations and indication in the team that that if we will not do this,

137

then there is a consequence for businesses to sell it in a way for the team and present it.

138

It's not just some updated work or not needed work, but something really important, not only for us as the developers, but first of all, to the business.

139

Yes.

140

Right.

141

Let's jump to the technical part.

142

Yeah.

143

Let's just start from the, some, you know, easy, easy ongoing questions.

144

So where in the Swift, we have the structures and classes.

145

Can you explain to me what is the differences between these two?

146

Yeah.

147

The main difference is because the struct, it's a value type and class, it's a heifer type.

148

So it's created a heifer in the memory.

149

So it's the main difference.

150

All right.

151

Maybe, can we, for example, inherit structure from another structure?

152

And no, you can't, I think.

153

Can we inherit classes?

154

Because it's like a structure.

155

It's a copy one from the object.

156

So yeah, it's immutable.

157

Yeah, right.

158

Can we inherit classes?

159

Yeah.

160

Yes.

161

Can we inherit, let's say, two classes?

162

So one, you know, we have class A and then we inherit class B, class C.

163

Yeah, you can.

164

Okay.

165

So how we can do this? so i just decided to describe

166

or you meant something else yeah you can create like a final class

167

and like in here to other like a network uh protocol class

168

and they are that would be inherit from a protocol

169

and you have a network service class and a mock services networks okay

170

so so you meant protocols

171

but i mean can you inherit you know classes just after you know comma class you have class a

172

and then you have just a pure class b

173

and then comma class c comma class d can you do this yeah you you can do it like

174

and create like an um let's see um a knowledge case

175

and y view controller

176

that would inherit some some functions to to analysis to to send to the crash leads

177

or send to the Firebase analytics some functions to analytics.

178

Or you can create a view that would have a background like rounded and grey.

179

I don't know, and every view that your design systems would be narrated from this one.

180

It can be created in the final class.

181

And yeah, like a super class.

182

Got it, got it.

183

All right, let's move further a little bit.

184

So in the Swift, we have different types, right?

185

So, and we have so -called primitive types that usually variables.

186

and we have reference types that usually is something bigger than just pure variables.

187

Can you tell me what is in this system?

188

What is closures?

189

Are they reference types or are they value types?

190

Yeah, honestly, no. I don't know about it.

191

Not sure.

192

Closures.

193

You know closures, right?

194

Yeah, I know closures.

195

I'm not sure where to.

196

Okay, no problem, no problem.

197

Right, and then maybe if we go a step further, let's maybe speak about the memory management.

198

So we have a memory management.

199

Can you explain how it works in the iOS?

200

Yeah, we have a GCG on iOS and it works with a lot of threads.

201

Yeah, we have ARC 2 that's that have the memory management for count heiferis.

202

And yeah, it's the automatically way that Swift works with memory management.

203

Right, right.

204

And for ARC, you have, you know, so can you maybe explain in the details a bit how ARC works for us?

205

Yeah.

206

How that works?

207

Every time that you create a new object, it counts one reference from this object, this one.

208

When the object would be deallocated, they decrease this reference from the memory management.

209

But of course, if you create some strong references, it could not decrease it and maybe crash the app,

210

create a memory leak.

211

Do you know how it's called this kind of bad situation when you have one strong reference to another strong reference?

212

Two classes have strong references to each other.

213

Do you know how this situation is called?

214

How this situation is called?

215

Yeah, I know the XETLIN name, but I forget.

216

Sorry.

217

No, no, it's not a problem.

218

It's called Retain Cycle, right?

219

Retain Cycle, yes.

220

Because you can't retain actually.

221

Yeah, yeah.

222

So I thought how to resolve it, do you know how to resolve such situations?

223

Yeah, you can use weak or unknowable, so two options to resolve it.

224

And what's the difference between these two?

225

Because weak is an optional one and unknowable you guarantee the app that it's,

226

you have, it's like an unfortunate in others.

227

Right, right.

228

And that's interesting.

229

So when we're working with this, right?

230

So I see that almost no one uses a null.

231

Do you know any scenario where you can actually use a null and be locally safe with that? questions?

232

Well, I think you can use it another way like when you have a reference between one,

233

just one class, like when you have a protocol inside in a cell,

234

for example, and the view controller has the table view.

235

You know that this protocol just will exist in this case because the cell,

236

it's just in there in this view controller.

237

It's not being used in anyone, in some somewhere, sorry.

238

And so you can use anodes in this case, that would be a problem.

239

Right.

240

Okay.

241

Yeah.

242

Yeah.

243

Interesting.

244

So let's then jump to next one with the, just maybe about the, you know, coming back to the closures.

245

So we have the closures, right?

246

And they have some specific, I would say, working with the references.

247

When we're working with the references inside the closures, can you tell what we need to be aware of?

248

You need to be aware of the routine cycle that you want to know?

249

I don't know if I understood everything, sorry.

250

No, no, I mean, that's one thing.

251

I agree.

252

But for inside the closures, how we can be safe with this kind of routine cycles,

253

what we should you can use wix itself as i as we said and guide let's to okay the the view

254

all right and then so for closures sometimes

255

when i'm using it for the functions from time to time

256

and it has a big different syntaxes in the you know

257

in the function name sometimes called escaping sometimes it's not escaping

258

can you tell me the difference between this what's what's the

259

point to put this yeah i already use it it's before of course

260

but yeah i can't exactly remember

261

that as i think escape as i can remember escaping i

262

think you can pass a code of block to be be executing in the in the order

263

when they the user would be call this function with the closer

264

so So you have a complexion one that I can remember, sorry, but yeah, my knowledge about that, it's not good.

265

It's not very close by the answer, but yet.

266

Okay.

267

Got it.

268

Got it.

269

Right.

270

And then, so the closures, uh, really, you know, they are, they are very powerful, right?

271

But is there, do you see, do you think it's a best practice to use closures as a parameters for functions today?

272

you know in a protocol oriented object it would be better Right.

273

But in the case when you need to, let's say, when I have a function and after that function,

274

after, you know, finished, we want to call another function.

275

Yeah, you can use a guarantee sequence.

276

That's right.

277

As we use, you know, with closures as a parameter, it's usually because when it's the parameter, we're calling it callback or, you know, completion block.

278

And then we're calling it and then putting some piece of code there to be called after the function or, you know, after we complete it.

279

Yeah.

280

Today, I think it would be better using the triassincurate in this case.

281

Right.

282

Yeah.

283

Good, good.

284

Assignuate is a modern way for threading, right?

285

And working with the threads.

286

Do you know any old style way for working with threads?

287

What do we have to use in a normal multi -threading without a sync weight?

288

Dispatch queries, you mean?

289

Yeah, for example, dispatch queries, something else?

290

Yeah, dispatch queries, as you said, the completions in some case should be...

291

I'm specifically about, you know, working in the multi - Dispatch group.

292

Dispatch group.

293

But that is just a part of GCD.

294

So we have a GCD.

295

We have a Syncquate.

296

Something else?

297

Do you know about something else?

298

No, that I can't remember.

299

Sorry.

300

Have you heard about the operation queues?

301

Oh, operation queues.

302

Yes, but I never worked at Zwift.

303

Yeah, but I have...

304

Right, right.

305

So for.. for these things to work.

306

So this AsyncAway and GTD in general, we're saying that it's multi -threading, right?

307

In our project.

308

But does it really clear multi -thread in our apps?

309

Or, I mean, how does it work that we can do things on different threads?

310

If we have, you know, iPhones with adjusted one or two core, actual physical core, I mean, do you know how

311

but i can create you know hundreds of threads i can create hundreds of calls

312

and gcd different priorities

313

or how does it work do you know general what what

314

does it mean for us this multi -threading for the apps is it the same as nvidia you know

315

when you have a thousand of separate threads where each works separately

316

or it works differently for us let me see we have a concurrence and also, oh my gosh, I forgot the...

317

That was my thing.

318

Yeah.

319

And you can create, you can work with that.

320

If you create other, like you said, it would be just to be more organized to us or just to,

321

but it's not, we work like in really so many different strategies.

322

It would be more for organize the codes for, we have this, these two ways to, to, to work with it at the,

323

the simultaneous and the asynchronous ones.

324

Yeah.

325

I don't know if you, you want to know about it.

326

Sorry.

327

Yeah.

328

I mean, okay.

329

It's everything, everything fine.

330

Let's then talk about, that's a good thing.

331

So we talked about the threats and concurrency, you know, how things works.

332

And the part of this, you talked about the async way, how this works.

333

But part of this topic is also so -called actors.

334

Do you know what is actors?

335

How do we work with this actors thing?

336

Yeah, you mean these annotations that you...

337

Yeah, for example, annotations.

338

Yeah, you're using some annotations to put this whole object or just the functions in a specific branch,

339

like a main actor.

340

okay but uh do you know

341

that we can actually create actor ourselves yeah you can create

342

actors like this yeah we have class we have structure

343

and then we have actor as a you know a keyword

344

we can create actor something do you know what what what

345

it give us to create such in such a way for example some class user manager

346

or something like this yeah actors uh as i said to change the block

347

or the objects and its markets that it would be due in a...

348

Well, it's markets the code or the function on a code block to do something that you want,

349

that you like a main actor, it would be that you mark the code block to be executed in the the main QE,

350

let you create an actor for other...

351

Like at this table, you would be due that the code of blocker is the, the, the, the, the stable from other, from the, the current app.

352

Okay.

353

But, but I'm, I'm saying that, so you're saying about the marking something as attribute like main actor or something like that for the class, right?

354

Mm -hmm.

355

Or, or property in the class.

356

Yeah.

357

What I'm saying that we can create actor as, you know, as you, for example, you have a, let's say,

358

network manager or user manager that manages users.

359

You have a class, user manager, something, something.

360

In the same way, we can create actor, user manager, something, something.

361

Do you know what would be the difference or what is the scenarios when would be nice to create actors such?

362

yeah honestly no i never worked with this okay type of actor got it got it

363

and uh probably last thing about um uh just in this you know topic

364

so we have so -called thing solid have you heard about the solid yeah

365

and there is some you know something called the dependency injection

366

just interesting how you would do it in the project without any with

367

or without frameworks If you have worked with some frameworks, just tell me with which, how you worked and the scenario when you don't have any frameworks, how you would do it.

368

Yeah, I have worked with Swinject,

369

but also with products without Swinject and we create our own factories.

370

Sometimes builders are necessary to inject some dependencies too

371

and sometimes just it's a it's a complex subject you know

372

like just an in meets with others it's already an injection dependence

373

so yeah all right yes yes i mean cool framework right

374

let's wrap up with this part yeah you've been really great

375

for this part i i have just maybe a few comments

376

so just about the first question you remember was about the inheritance in the suite

377

and can we inherit multiple classes and there is might be some misunderstanding, but the thing just to, you know, be clear here in the iOS,

378

in the Swift and Objective -C, we don't have multiple class inheritance.

379

Just as example, as just I said it, so we have class A, then you have inheritance from class B, then comma inheritance class C, you can't do it.

380

So that's why we're using product protocols, that's why we're using, but we can, we can use like simulated in a way.

381

So in a protocol oriented programming, when you have multiple protocols, we have, you know, infinite amount of protocols that we can implement, but it's a different one, right?

382

So between inheritance and implementation or something, but in a way we can simulate it because what we can do,

383

we can extend the protocol and then do default implementation for each function, right?

384

Of the protocol, for example.

385

That meaning that we have a default implementation for functions.

386

And whoever inherits or implements these protocols actually has default implementation for functions.

387

Same way as it would have during the inheritance of these functions from another class.

388

You see what I mean, right?

389

So this is the trick where we could use to have

390

this kind of multiple inheritance doesn't mean it's always need to be in place, doesn't mean it's always good, but just the fact that we don't have multiple inheritance,

391

but we can simulate it in a way with a protocol

392

or in the programming with extending it and giving it default implementation for the functions.

393

Right, the other one about the actors.

394

So basically just my comment that, you know, check it out, how it works.

395

It's something that now on the, you know, on the top of discussions, where to use it, where not to use it.

396

Yeah, I would say it's helpful for some specific group of tasks.

397

So the difference would be that the actors makes your data inside secure for threats,

398

for concurrency, but it has some drawbacks or I mean, it has some tricks.

399

So you need to be aware of something there.

400

So it doesn't mean that you guarantee that everything that runs inside the actors, threat safe.

401

If something comes from outside, if you have some parameters and functions that comes from outside, there is no thread safe anymore,

402

even though it works inside the actors.

403

But whatever inside class as property in the class, as some variable, I mean, in the class, some counters,

404

some other classes, it's safe.

405

So that's just cool to know.

406

And for scenarios like maybe network manager, maybe user manager, maybe some credentials or other parts

407

when you need threat safes that's a good idea to create actor

408

but yeah it's it has some definition it has some behavior

409

so you know the part about the part about

410

that in general about the actors is set right

411

that we have this kind of main actors

412

and to run everything even there you could be a bit

413

more district this descriptive in a way why we need it

414

or how it help us yes i didn't ask it

415

but just you can just tell it

416

that okay we have this kind of main actors it's it

417

helps us to run everything on the main thread on the main queue

418

and we save there so that's really helpful

419

because before we needed to run in dispatch queue main async everywhere for example to guarantee

420

that this code works and now if you just mark it with the main actor.

421

And then in most scenarios, it runs on the main queue.

422

Again, there is some, you know, some scenarios when something could go wrong, but it's a bit harder scenarios, not just a normal when you're just calling some functions or,

423

you know, changing some variables.

424

That should be pretty, pretty fine.

425

Right.

426

That's pretty much it.

427

Otherwise, yep, we took clear.

428

Was something, ah, yeah, the closures.

429

I asked you about the closures, right?

430

It's a class as a reference type of course there are closures

431

because you can pass them I asked if the closures value type or reference type remember this?

432

Yeah I remember this but I didn't understood.

433

Yeah yeah it might be again misrepresenting in a way so just saying that the closures

434

reference types

435

so that's why we can pass them in the parameters of

436

functions that's why we can in general working with them in such way that we can call them afterwards.

437

Same for escaping, non -escaping.

438

You answered in a way for this, but the clear answer would be that escaping closures,

439

they basically escaping current context, meaning you have a function that returns something.

440

So escaping closures can be called after function returned.

441

And non -escaping will be always called, you know, just sequent.

442

It's called and then something returned.

443

So in the moment when you have some thread work or some async work, where you're calling this kind of closure,

444

the compiler, it has compiler check.

445

Compiler will tell you to put escaping there because you will call this closure afterwards,

446

after the function finished, just to cover up this topic.

447

Right.

448

Yeah, that's pretty much it.

449

Let's now jump to interesting part, practical part.

450

So I will send to the chat some part of the codes and if it's fine to you, you can maybe share your screen with the playground.

451

If you can just start it.

452

I've sent some code in the chat.

453

If you see it.

454

Nice.

455

Yeah.

456

Just pass it there and then let's discuss.

457

Right.

458

So you can maybe, yeah, just ignore it.

459

It's fine.

460

So you can maybe just increase the font size here.

461

Increase font size?

462

Yeah, just or comment plus maybe.

463

Yep.

464

Nice.

465

Right.

466

So let's just ignore this.

467

Let's just thread safe checks.

468

It's fine.

469

But in general, let's just discuss this code.

470

How do you feel about this code?

471

Do you see any problems with this code?

472

what kind of followers you see?

473

What would you improve?

474

Yeah, I think here I will check in this way.

475

You can say in this case, if they're this way, you pass the data.

476

What do you...

477

You need this one too.

478

Right.

479

And you need to put a return to avoid call tries that I can see.

480

Right.

481

What do you think about the first two constants that we have in the class?

482

First constants?

483

Ah, yeah.

484

Do you have any comments about this?

485

You'd say that if it's a charge,

486

would it be uh states it needed to be no it's needed to be staged but not

487

Same day both.

488

That's fine.

489

I mean, you don't need to mark anything for now.

490

Let's discuss, you know, just a singleton here.

491

So you have a shared singleton, right, for network manager?

492

Yeah.

493

Any comments on that?

494

Do you feel it's fine?

495

I think we need to create and initialize it to

496

that's assumed that it would be just one

497

and in it but yeah I can't see any problems here sorry I think okay for the session

498

Is it fine to have such a session as shared one here?

499

Just use the shared one?

500

Yeah, I don't see any problems.

501

Okay.

502

Right.

503

So if today you would do something like that, let's say you need to load data, would you do it like this kind of function?

504

Or would you do it differently?

505

If today I do, I think I wouldn't create a network chart.

506

I will create a final class with a protocol.

507

Okay.

508

So why would you need the protocol?

509

No, with some get, the other ones, like put and everything else.

510

and the load data would be my network would be in the my network service

511

I think but the way it implemented do you like it the way the implementation way of this function

512

or you would change it to something else for the function

513

you mean yeah for the function itself the searching yeah

514

so you basically need to load data right by url some some url

515

that coming it could be let's say images whatever

516

and you have some urls

517

and then updates you got it you need to do something

518

with this data would you use function like this for your real project

519

or how we would do it no in my real products of course not because it's a it's a simple function.

520

It just puts an URL.

521

You don't know if the URL is a UTF string.

522

Also, sometimes you need methods,

523

parameters, and I think who else to do this one.

524

But for a simple function, I don't see in big project in a huge problem, you know?

525

Okay.

526

Yes.

527

Right.

528

I mean, there could be, of course, different, different way for improvements to this function.

529

One of the way would be using a sync way instead of completion blocks, right?

530

So would you use the, how you, you know, today in your work or in your pet project,

531

why would you prefer using the completion blocks for such functions?

532

No, I prefer using a SyncAway.

533

A SyncAway, yeah.

534

So this could be also mentioned in a way, but I know that you know it.

535

So just to be clear here.

536

You see inside the function, there are multiple completions, right?

537

Mm -hmm.

538

And multiple places.

539

Do you know any way to move this completion just in one place?

540

That guarantees that it's always called?

541

Yeah, you can like put that. So it's like that.

542

I got it.

543

Yeah, that could be one way.

544

Let's say that inside this task, we have more if else's or garbs inside.

545

For example, when you need to parse do so or do something else. So sorry?

546

Yeah.

547

It's fine.

548

Sorry.

549

Can you repeat, please?

550

Yeah.

551

So let's assume that you have multiple if -else's inside this task.

552

Yeah.

553

Right?

554

So when you need to call...

555

Like it.

556

Yeah.

557

Like here.

558

So let's leave this scenario when you have this.

559

But do you know any way to still call just...

560

you know this completion just once inside the function but without changing the

561

signature without changing yes some way some blog of code

562

that will be guaranteed then there is guarantee

563

that it will be executed before function is returned when everything is finished it will be executed.

564

Yeah, I just want one.

565

Have you heard about that defer?

566

Defer?

567

Yes.

568

Uh -huh, you know what?

569

Let's see how we would do it.

570

I already know about in -out and defer,

571

but I can't apply in this situation.

572

No, no, it's fine fine just so you know

573

that in some scenarios

574

when you need to return after all the code is you know run best practice to use return

575

or defer it's just you know as just a simple block that will be called

576

um after function returned that could be better

577

if you're using still such the completions

578

when you need to you know do this but

579

and how are you use here yeah yeah exactly here you would return here

580

but you need to provide then you create before defer a variable

581

that holds inside possibly data so you would as you would work with the data it would be nil or something

582

And if you would have an error as well in the part of the completion block, then you would need to do the same.

583

So it's either exists or by default just an ilk and it will be executed after this task finished.

584

Right.

585

And then let's just maybe check one more example.

586

Yeah.

587

Just wondering, have you worked with the SwiftUI?

588

Yes, my main background is SwiftUIKit.

589

But have you really worked with SwiftUI in a way?

590

Yeah, in just views.

591

Okay, got it.

592

Let's maybe combine it.

593

Okay, so UIKit you said, right.

594

Gonna be simple example, too simple for you.

595

Just drop it to you.

596

This one?

597

Yep.

598

Any problems here?

599

Any problems?

600

Yeah.

601

We have chains in global QA.

602

It's what's being the main QA and also it would be better

603

if you have thread safe

604

or thread safe nor everything cycle safe to to view i would say

605

that you don't need it here you can

606

because you can assume that's it just

607

but yeah i prefer using every time for safety

608

but then what would be

609

if you hear instead of async main async will use main sync would be any difference yeah

610

when you use sync you are like stuck the the codes the the view

611

and and the codes will would be as acute

612

so the layout it's it's what's stuck in the the view

613

but is it is it always true that it will be stuck.

614

Oh, it's true?

615

Yeah, I mean, why even this sync exists then?

616

So, can we use it sometimes?

617

Or is it always?

618

Yeah, if you have a very important info or you need like

619

code to be executing in the same type that you want, it would be better use sync.

620

Okay.

621

Right.

622

But let's say today, can you, can you maybe change this code in a way

623

that you don't need to use the async dispatch queue main async at all?

624

What we should change here?

625

So we know that this function called of the data loaded, right?

626

So inside, let's say some background queue where we did some requests.

627

How you can change it. So.

628

Because I would say that we don't need to use this dispatch code main thing at all.

629

Yeah, because you mean that it's...

630

If the functions wouldn't be called from an API, you mean?

631

Like, let's say you're using...

632

In today's code, you have some background work, right?

633

You have background request.

634

And then you have such view controller with this kind of load data.

635

How to change this code so you don't need to use dispatch queue main async.

636

So it's always executed on the main.

637

Because, you know, dispatch queue main async just creates this kind of doom effect, like spacing, you know, you just wrapped up it to some part of code.

638

You mean like put some main actor here? For example.

639

Part of here?

640

Okay, interesting, yes.

641

Right, and the interesting question would be, does view controller here in the class view controller,

642

is it main actor or not?

643

It can be, but in this case, no. It's just the function.

644

If you didn't put it here, yeah.

645

So if it's just like that.

646

Yeah, it's just like

647

that just this code block is all being the main actor

648

i got it i got it yeah i mean then we can wrap up

649

so just for for comment for this one would be nice to know for you

650

that in this example view controller is main actor really yeah that's the interesting thing to remember

651

because if you go inside UIViewController, if you open it, somehow in your project inside,

652

you will see that in today's iOS development, all the UIViewControllers, all the UIViews,

653

they are inherent from or they have main actor basically.

654

So all your views, all your controllers, if they are part of the UIViewController, then they are by default,

655

they have main actor, meaning you don't need to put main actor at all.

656

It will always run this function on the main.

657

So the only thing that you would need to do to be sure that your changes...

658

If you're using async await, nothing you need to do.

659

It becomes hard.

660

Yeah, absolutely.

661

So you'll just call it then because actors and async await knows how to switch the context,

662

knows that this part of code should be run on the background and this part of code.

663

Now, you know, returning back from the context to this actor and running this part of code here.

664

But that's purely safe.

665

The only moment when you need to put these kind of marks, main actors, when you have your personal classes, like some view models, you know, something like that.

666

Yeah, I understood.

667

So then you, of course, need to think, should you put it there or not?

668

Usually, yes if you do something related to view you just put main actor on top of the class

669

but now in the modern iOS I think that comes with a new release

670

all the classes if I remember correctly will be by default

671

everything will be by default main actor as it's supposed to be all the properties

672

and you can turn it off for some functions if you need very nice I didn't know about it Yeah, so it's interesting.

673

Yeah, I think it's useful to remember for your code.

674

Yeah, very useful.

675

If you just put it as, sometimes people are putting dispatch queue main as sync or running main actors for no reason, basically.

676

Because every time you're putting it to the code, meaning that you're just delaying it for some amount of, you know, nanoseconds, whenever, this part of code.

677

At the same time, it could be executed instantly.

678

And sometimes it affects your, it could affect your animation or some presentation, you know, in a weird way.

679

Instead of being a smooth, it could just be a little bit delayed and then presented for no reason.

680

When you can actually just, yeah, you know, use it.

681

To remove it, yeah.

682

Yeah.

683

Let's wrap up in general, Ramiro.

684

I think we did a great job.

685

So how do you feel about the...

686

running through this interview?

687

Was it hard for you?

688

Was it easy for you?

689

Yeah, it was hard to me.

690

Now I know that I need to study some points.

691

I have some lack of knowledge.

692

But it would be very nice to me to see what they want for me

693

and what I need to study more and I did some notes that you told to me.

694

Yep, sure.

695

You can always, if you have some questions or something,

696

I can reconcile what was the problem, what was the questions.

697

But in general, I think it was a really good session.

698

It's very different for interviews in Brazil, you know?

699

Yeah, yeah, it's very different because in Brazil,

700

they wouldn't know more about the real job you know like uh

701

and you this interviewer knew about the the knowledge of the

702

the the the the deep knowledge of the swift was very nice to to to have this interview thank you

703

so much yes thank you i i should say that all the all the interviews,

704

all the how it behaves, it really depends on the company, right, on the situation, because usually when we're interviewing someone to the company,

705

we're not just interviewing some random person to random position, there is actual position actual team,

706

actual expectation from this person and that gives more context usually to the

707

to the interview here on the mock interview on the other

708

side we have to go generally about things you know just pick more general topics of course

709

if you would interview to you know some audio company we

710

would talk about the audio you know core core core kids you know some

711

very different frameworks audio frameworks

712

and how you would work with this it's really depends to become where you're going

713

but i think so that that was a great

714

going through all these topics sections so i hope it was useful for you and uh

715

that's that's pretty much it thank you very much thank you for your time yes bye

쉐도잉이란? 영어 실력을 빠르게 키우는 과학적 방법

쉐도잉(Shadowing)은 원래 전문 통역사 훈련을 위해 개발된 언어 학습 기법으로, 다언어 학자인 Dr. Alexander Arguelles에 의해 대중화된 방법입니다. 핵심 원리는 간단하지만 매우 강력합니다: 원어민의 영어를 들으면서 1~2초의 짧은 지연으로 즉시 소리 내어 따라 말하는 것——마치 '그림자(shadow)'처럼 화자를 따라가는 것입니다. 문법 공부나 수동적인 청취와 달리, 쉐도잉은 뇌와 입 근육이 동시에 실시간으로 영어를 처리하고 재현하도록 훈련합니다. 연구에 따르면 이 방법은 발음 정확도, 억양, 리듬, 연음, 청취력, 말하기 유창성을 크게 향상시킵니다. IELTS 스피킹 준비와 자연스러운 영어 소통을 원하는 분들에게 특히 효과적입니다.

쉐도잉 연습: Mastering iOS Development: Mock Interview with Ramiro Neto - 영상으로 영어 말하기 배우기

이 레슨에 대해

쉐도잉이란? 영어 실력을 빠르게 키우는 과학적 방법