@@codingwithjamalexcept that next constantly makes lives complicated (yea talking about cache here) while angular surprisingly started to get lives easier.
@@alexandertarasenko3038 Well, thats your point of view. I think nextjs has been progressively getting easier, and probably anyone who actually uses it thinks the same. Adding optional features doesn't make anything complicated (yea talking about cache here)
26:00 I don't think you really understand what standalone means. It's nothing to do with the DOM. It just means they don't need to be included in modules. Standalone means they are their own modules and can be imported directly into other components. This was the original plan back in 2015 but they had to change it before the official Angular 2 release because of some TypeScript issue.
I've done a 3 months Angular 18 side project and boy was I pleasantly surprised. Nothing like the Vue 3 project or Angular 6 project I did back in the days. Everything just felt like the best practices I've learned & implemented in React but out-of-the-box and done in an intuitive way. I had the absolutely same reaction as you Theo, wanting to hate it but slowly getting Angularized. Interesting times.
theo is the only channel that i'm actually excited to see the ad segment, i'm always getting introduced to cool new tools and stagehand looks amazing + thanks for showing some love to Angular and the team behind it
Yeah right... because react compiler really adds complexity
Місяць тому+31
I'm an Angular dev and I must say the last few releases have been a blast! Less code and less bugs. Easier to understand. Easier to pick up as a new dev. Easier for React or Vue dev to jump in (and Angular dev to use those). Way faster apps. Less RxJs code. I've seen a lot of push back from some Angular devs but there's no arguing these changes bring the framework in the right direction.
@ angular will always be behind the industry unfortunately. It’s has its own conventions and it’s way of doing things that not everyone wants to adopt. React is way more flexible and unopinionated.
Drag and drop as any other UI element is not "build in into framework". There is an Angular CDK with core primitives and Angular Material both officially maintained via Angular team. And that is actually a huge benefit, I hate material for many reasons, but it is the only reliable thing you can use to dtay up to date. Since with any major update of Angular you get material update with corresponding schematics.
Angular CDK is awesome, it lets you implement your own components and your own UI. It's a huge win compared to other frameworks that are stuck with opinionated component libraries
Angular dev here, the recent changes in Angular make me less jealous of the React ecosystem ❤ I lived through the AngularJS to Angular transition, this is a million times smoother.
React devs have been pretending for years that Angular doesn't exist (at least at its real popularity), meanwhile Angular quietly integrating the best features, while leaving out the worst, of all of the more hip frameworks. It has always been the framework for money-making business and applications, and I believe it is further solidifying it's position. The preliminary data from State of JS shows it's having a resurgence in both popularity and likability.
I don't think anybody was "pretending angular doesn't exist". But you are right in something, itr is for "money-making business", it has all the bullshit talk and style corporations like and go hard in this direction, don't change the fact it is only of the tools that are money-making, react is also one of those and angular people love to shit on react a lot as well.
@@diadetediotedio6918 A lot of business tech people look at React and discard it because the ecosystem is in constant flux, they want something that has most of the things included and doesn't change all the time. This is especially true if you work on projects that target generally more conservative customer bases (like banking, infrastructure,...).
I've been using Angular for over five years now. They have always made their migrations super simple to do and the documentation is superb. I really enjoy the Angular framework, especially coupled with an NX monorepo.
The more direct control over rendering types, like the route definitions and piping the hydration event, both really nice. I love Next but a lot of that behavior is obfuscated on their side
dude, you're clearly a nextjs fan, let me say you, angular introduced many many of the concepts out of the box way before nextjs was born and not the other way around. Angular 2 had universal for SSR since the start, around 2015.
Been an angular dev since before the red wedding and I'm very excited about these changes. It's very encouraging to see Google put their weight behind this. The future of Angular is very bright!
I've been working with Angular starting from 1.x series, and by far the core issue I've seen with young devs is the inability to think in terms of OOD and MVC design. It takes time to master it, still absolutely worth the effort.
Love to see this. I am a senior web dev in my company. We are using Angular 12v. We are block on the migration to the latest due to material UI 15 breaking all our styles and it doesn’t have a support for old version. Angular 18 has to use material UI 18, no backward compatibility. I already did the migration for angular 18 on our angular 12 and the material UI kill it. We decided to move to NextJS. I am happy that Angular is becoming for like NextJS
I like this a lot because I can bring a lot of the new stuff into existing projects and the other devs who might not pay as much attention can just use it without much hassle.
Being an angular developer for the last couple of years, for me, a regular ng update should not bring too many breaking changes or costly innovations. Having a large repo, quality and stability is very important, more than the productivity gains of a certain new feature.
In the past, Laravel frequently released major updates, making it challenging for developers to keep up. Angular is currently on a similar path. They should adopt a one major release per year cycle.
Looks like an easy upgrade for huge apps and thus makes React less important as we were considering moving to Svelte (because no one wants React cancer).
Before this changes i was seriously considering going all in on react, ive only used angular and vue and im concerned about open roles using angular. Im hoping this makes more angular roles available outside of just huge enterprise applications.
Standalone and isolated components ARE a good thing, and I am really baffled w/ Angular devs who yearn for ngModules. Angular modules were introduced before we had an industry standard module system on the web (ESM). The Angular team recognised 4 years ago that they have outlived their usefulness, and over 2.5 years ago (and since) they have made all features work without ngModules, leaving them in for interop as syntactic sugar.
Yeah, parts of my code base written with standalone components instead of modules has a much better dev experience, with unit tests running faster, etc. because of the better tree-shaking. It's also so much simpler to see what your dependencies are at a glance. The 2 things I'm not sure are a good things are inject instead of a more common constructor injection, and the functional guards (having to use assertions to make sure the guard functions will not be used in a context where there cannot be injection seems worse than the previous guards with services implementing the interface).
UA-cam is not going to be rewritten in Angular :) rumors of wiz and Angular mergin were greatly exaggerated by one of the parties. What is going to happen is that they will share more ideas and some dependencies (like signal implementation)
Yep, they'll just take the best parts from each other. And it won't happen immediately - that's their long-term plan, unfolding over the next 10 years.
Regarding stagehand, I will be always in control of the code of the right side and it's guaranteed to work always. I never gonna use stagehand for testing purpose. Yeah definitely agreed, it has it's use case in other places
@@sohamnandi7526 yeah I didn't want to be a naysayer without actually investigating it but if I was doing the advertising for it then very first point to be made would not be shorter set up code it would be how they solved the repeatability issue so the ai generated tests are actually the same every time.
The autocsp is huge, and should be making a way bigger jump up and down as that will make angular app’s security wise improve majorly and from the start of a project having that makes a great dev experience as I have wondered about csp things in typescript/js apps for a long time
Although I almost exclusively live on the back end, Angular has been the most approachable framework to land into. It seems like there was just a natural right way to do things. Dependency Injection, templating and RxJS were great. With react there seemed to be a footgun around every corner blowing up the reactivity of the DOM, or making elements render a million times too many. I'll never be a javascript person, but if I have another go at a more frontend-leaning app than HTMX can offer, I'll pick Angular again :D
Angular user here. Started learning between 17-18 and one of first adopters. I've got a couple critiques on new stuff hope it reaches the right people: - linkedSignal breaks the reactivity naming convention ( signal, computed, effect, resource ... ) and "linked" is enough imo. - resource (and rxResource) are just inferior versions of tanstack/query (and ngneat/query) and I'd advise to use them instead while resource is still in dev preview and may get better
31:04 I actually think if an app I was using had filtering like this I wouldn't want them to drop my selections just because they are filtered out of my current screen. Cause if I know I want a few things from one category and a few from another, it'll be shorter to look for them when I only have to look through their category. You are probably right that not being able to undo the selection is an issue though. But I'd say put the selected values up at the top of the page(if they were filtered out), and the values that still remain after all the filters just continue as they are now.
As a mostly end-to-end testing relying blazor dev, I've been looking for a tool like stagehand for a while and it seems really cool! Can't wait for it to support local llms like llama with ollama.
I‘m hyped and worried at the same time. It‘s nice to see angular evolve and care about performance. On the other hand I fell like they might destroy some of the core benefits of angular and there are also some long existing bugs that are not beeing fixed. I hope they don‘t just follow hypes and trends blindly and move in different direction once the direction changes.
19:45 you said you hate the syntax. Was this just instinctive? in that you prefer react style vs traditional angular? Or did you not notice that angular has new control flow syntax? It felt like you probably were already decided you don't like angular syntax, and didnt grok that it was actually very react style now.
@@TayambaMwanza ah is it? Thanks for the clarification. I am quite limited with experience of other libraries. I've just stayed with Angular as I wanted to become good at one rather than distracted by the shiny things and just know bits of all of them.
I first got exposed to Angular after working on big Ruby on Rails projects. And the cool thing at the time was the similarities in scafolding between RoR and Angular. But now, with the simplicity and control you have with the JS ecosystem wether it be node, deno react and so on, I find the boilerplate and paradigm of Angular quite frustrating
We run a ecommerce shop on angular and it's way faster and snappier then the demo you showed. If you profile the code and avoid patterns that tank your performance, then angular could always be fast. zone.js and animations are the worst performance offenders, and we got rid of them 4 years ago already and use observables and native css animations instead.
@@professor_ozzy SSR solves that. Our product pages answer in 220ms - html generated by angular included. But angular SSR isn't that fast by default. We wrapped SSR to work like an SPA as well. We bootstrap common pages ahead of time, then when a request arrives, it just applies the changes, as it would in the browser, yielding in SSR answering in ~40ms.
Regarding the 2700 thumbs-up on github: This isn't the bar to clear for an issue to be addressed, this is the cumulative number of thumb-ups across all issues that are addressed in this release.
I am an angular enjoyer, and no I do not use Next.js, I use React and sometimes Angular ,dabbled with Swelte and that's about it, other than that I'm doing Go and Java and the thing I love about Angular is that the way its developed is extremely opinionated by the devs of the framework unlike other frameworks where we can go crazy, I do not necessarily want to go crazy I love the fact that everything has a way for you to do it, ish~ Also I hate things like React the more i interact with React codebases because all of them are so different, and its pissing me off, but in Angular, its always the same but maybe with a bit of flawour.
Angular Community Member here, we had a lot of push back a few versions ago from people who didnt want Angular to change, as time has gone on and the benefits have become clear 99% of Angular devs are on board, id say its practically a non issue by now. Remember you're seeing only V19, theres been several RFC and conferences talks/announcements that Angular devs have been exposed to, so we've passed the pushback phase already.
@hengkeatyam3700 silly question in the software world why would a brand new version that's not even had hot fixes yet have more downloads than the older one?
@@TayambaMwanza Angular v4 was the first stable version. Then, Angular rewrote the rendering engine. Angular v9 was the first stable version of the new Ivy rendering engine. After that, internal conflicts emerged within Angular. By Angular v14, new coding styles began to appear. In Angular v19, these new styles became over 50% stable. Angular Material v18 doesn't include a single line of code related to Signals. Angular Material v19 includes a small amount of code using Signals. Angular Reactive Forms are still not related to Signals. Angular Router and HttpClient still rely on RxJS. Why did we learn and write Zone.js, decorators, RxJS, and NgModules, when now they can be optional or even no longer considered best practices? Was it because of new technological breakthroughs in front-end development in recent years? No, it was simply because Angular chose the wrong path. They spent too much time on rewrites and migrations.
Angular defer directive is closer to what react lazy does than to ssr=false. To be honest i don't know if they do SSR of it or not, so i guess maybe it's both ssr=false and lazy? :)
Queues can be dangerous for input, but I'm guessing that the angular devs have somehow addressed it. For example, imagine a user clicking a button and seeing nothing happen, so then they click it like twenty times in a row and/or get frustrated and just start clicking things all over the page. If everything were queued suddenly all this unintentional input is captured. I can't remember the context (what language/project, don't think it was even web) but I realized this when using a queueing approach to handle missed input. The better option was just to let the interface not do anything until it could be handled. Anyways, like I said I'm sure there is more sophisticated logic in angular.
6:57 Just try angular, make a relatively big project with it, and watch your development server explode with updates taking 1-60 seconds each update you make to components that are used in a lot of places. The experience is definitively astonishing.
I like the SSG, been using the SSG recently and even if it still feels weird to configure with Vercel compared to Next.js which is without custom shit, its good and working properly
This won't affect 80% of angular projects. Most angular projects are stuck in 12-17. Their managers need a business reason to do this and all the in-house optimizations they made to make it work fast will break when they bump the version. Happened to all angular projects I worked on. I don't discredit that this is a bad company culture tbf.
I am fine with this as an Angular Dev and I have been around since the AngularJS (hated it back then). I am not a NextJS (I see its advantages) person and like HTMX/AlpineJS & Astro for most of my personal stuff. ZoneJS issues is that you would click a button to update some counter on the same page and it would update the whole page. With the way it should work 17+ it can be it just updates the counter. Old APPs in Angular are not getting updated most likely, no matter what it takes about 1 year (depending on size) to update from Angular 3 to 11 let alone 19. Also there are things like Bootstrap or PrimeNg behind some of the components which is more of a pain to convert most of the time.
“…just write an English instruction and it does the right thing”. English is ambiguous and not only that but AI can’t interpret it as good as a human dev yet. The right side example is longer because it’s explicit and way more “deterministic” if you will. The magical thing about programming is the ridiculously high exactness of control you have over the computer. (wrote at 3:26, not a dev I am at uni programming as a hobby, I critique words and opinions not people)
i enjoy angular a lot... but i understand the fallout from the red wedding. at work, half our code is still angularjs since our upgrade path is unclear... to big for us to re-write. tried to get hybrid conversion.. (had done it before with smaller app)... dont think its do-able.
I think the constant change of Angular is not a good thing. It means constant breakage and need to rewrite old parts to the new ways of doing things. It puts additional load on the developers to constantly learn new things and discard them after the new best thing appears. It makes you feel that everything you learn about Angular can soon become useless.
angular dev since .js ... actually angular updates broke my apps every version up to maybe 4 years ago... it almost kick the ecosystem... but i somehow keept the trust (probably also sunk cost fallacy)
zonejs was annoying because it was a 'huge' payload and even required if you used onPush (alternative way of change-detection - not relaying on it). Also it wasn't possible to configure it probably without other 3rd party libs - for example default change-detection would trigger updates - for example on mouse-move even if there was no reason too and you couldn't opt-out of that behaviour easily
I like that you got a bit over the unreasonable Angular hatred. Nothing is ever perfect of course and Angular has some messy parts. But c'mon - next.js is SUCH a mess too. Using unstable versions of React has lead to serious issues and headaches. React 19 relased last month FINALLY, but before that it absolutely was BLEEDING edge to even use next.js. It really IS a good sign that they look to other frameworks and copy some of their best ideas, while still keeping their own identity.
The Angular team doesn't care about the number of users. Even if 5 developers use it, it gives a major update every 6 months. The project I started with Angular 10. Now it is in angular 19. Only the stanalone component took me some time. I only came up to 19 with migration scripts. This was not the angular I was used to that I learned at the beginning. The project became like someone else's project. This is the negative aspect.
Idk how many viewers are hate watchers like me. Mind you I like Theo as a person/youtuber, the thing I hate is web dev and seeing him just casually mentions hundreds of frameworks and APIs a second with somehow no overlap is a beautiful reminder of why I'm never going back to working in this world.
Sadly your concerns with the angular community are real. The angular subreddit has been a shitshow of insults towards anything that isnt "traditionnal angular" as of recently. They even insult newcomers coming from other frameworks that are trying to learn Angular, just because they were "on the enemy team".
Theo, if you're going to become an Angular person you need to know that we get a major version every six months!
Same with next
Shh, don’t scare them
@@codingwithjamalexcept that next constantly makes lives complicated (yea talking about cache here) while angular surprisingly started to get lives easier.
@@alexandertarasenko3038 Well, thats your point of view. I think nextjs has been progressively getting easier, and probably anyone who actually uses it thinks the same. Adding optional features doesn't make anything complicated (yea talking about cache here)
I think us Angular devs have a very low bar though. Like, it's been so bad for soooo long that literally any scrap of performance is a game changer
26:00 I don't think you really understand what standalone means. It's nothing to do with the DOM. It just means they don't need to be included in modules. Standalone means they are their own modules and can be imported directly into other components. This was the original plan back in 2015 but they had to change it before the official Angular 2 release because of some TypeScript issue.
I've done a 3 months Angular 18 side project and boy was I pleasantly surprised. Nothing like the Vue 3 project or Angular 6 project I did back in the days. Everything just felt like the best practices I've learned & implemented in React but out-of-the-box and done in an intuitive way. I had the absolutely same reaction as you Theo, wanting to hate it but slowly getting Angularized. Interesting times.
Always appreciate the acknowledgement for Analog 🤝
Deserved!🎉
With incremental hydration, code can be loaded not only on interaction, but also on “hover,” “idle,” or when an element becomes visible to the user.
Welcome, you late but welcome in the club of frameworks that can do that.
@@ivan.jeremic you mean angular or OP?
@@ivan.jeremicYou might be confusing deferred loading with deferred hydration.
theo is the only channel that i'm actually excited to see the ad segment, i'm always getting introduced to cool new tools and stagehand looks amazing + thanks for showing some love to Angular and the team behind it
They learnt a lot from angularjs to angular transition.
React getting complicated with new release whereas the angular is getting simpler.
Yeah right... because react compiler really adds complexity
I'm an Angular dev and I must say the last few releases have been a blast! Less code and less bugs. Easier to understand. Easier to pick up as a new dev. Easier for React or Vue dev to jump in (and Angular dev to use those). Way faster apps. Less RxJs code.
I've seen a lot of push back from some Angular devs but there's no arguing these changes bring the framework in the right direction.
Angular had a form of server side rendering called Angular Universal. It has been there for quite a while.
Around 17/18 already on @angular/ssr
I know angular universal was a thing even before spa frameworks went full circle and started ssr again 😂
And more recently, AnalogJS
The Angular team didn't push Universal enough for developers to care. Now Angular is years behind and trying to catch up.
@ angular will always be behind the industry unfortunately. It’s has its own conventions and it’s way of doing things that not everyone wants to adopt. React is way more flexible and unopinionated.
Drag and drop as any other UI element is not "build in into framework". There is an Angular CDK with core primitives and Angular Material both officially maintained via Angular team.
And that is actually a huge benefit, I hate material for many reasons, but it is the only reliable thing you can use to dtay up to date. Since with any major update of Angular you get material update with corresponding schematics.
Angular CDK is awesome, it lets you implement your own components and your own UI. It's a huge win compared to other frameworks that are stuck with opinionated component libraries
Angular Material is great but it's definitely not perfect. I've ran into a number of issues with it over the years.
Angular dev here, the recent changes in Angular make me less jealous of the React ecosystem ❤ I lived through the AngularJS to Angular transition, this is a million times smoother.
React devs have been pretending for years that Angular doesn't exist (at least at its real popularity), meanwhile Angular quietly integrating the best features, while leaving out the worst, of all of the more hip frameworks. It has always been the framework for money-making business and applications, and I believe it is further solidifying it's position. The preliminary data from State of JS shows it's having a resurgence in both popularity and likability.
Solid is a dead framework@日本語だけを聞くために-n3o
I don't think anybody was "pretending angular doesn't exist". But you are right in something, itr is for "money-making business", it has all the bullshit talk and style corporations like and go hard in this direction, don't change the fact it is only of the tools that are money-making, react is also one of those and angular people love to shit on react a lot as well.
@@diadetediotedio6918 A lot of business tech people look at React and discard it because the ecosystem is in constant flux, they want something that has most of the things included and doesn't change all the time. This is especially true if you work on projects that target generally more conservative customer bases (like banking, infrastructure,...).
@日本語だけを聞くために-n3o the creator of solid tried angular and likes it.
Idk why you say that, I've been an angular dev for at least 5 yrs and angular devs have always had an amazing community and super good documentation
Good documentation…? Lol
"super good" documentation
@@edlmiranda yes
Theo joins the dark side, leaves react and dives head first in Angular, lets go 🤣🤣🤣🤣
I am more blown away by your sponsor than Angular 19
Brob was I the only one who noticed it, that is so good
Same here man, I was just looking if someone else has mentioned it 😜
Yeah that looks amazing, I think it could actually help me solve some things I have been wanting to do for a while.
Looks amazing, it's just the reproducibility concern that's bothering me.
Just dont check their pricing tab
They've been on fire for a while now. Just keeps getting better.
been an angular dev for a few years now and these updates go so hard
I've been using Angular for over five years now. They have always made their migrations super simple to do and the documentation is superb. I really enjoy the Angular framework, especially coupled with an NX monorepo.
The more direct control over rendering types, like the route definitions and piping the hydration event, both really nice. I love Next but a lot of that behavior is obfuscated on their side
Angular is 1000% slept on
dude, you're clearly a nextjs fan, let me say you, angular introduced many many of the concepts out of the box way before nextjs was born and not the other way around. Angular 2 had universal for SSR since the start, around 2015.
Ben Awad gonna be so mad.
Angular dev here. Thanks for the video :)
Been an angular dev since before the red wedding and I'm very excited about these changes. It's very encouraging to see Google put their weight behind this. The future of Angular is very bright!
I've been working with Angular starting from 1.x series, and by far the core issue I've seen with young devs is the inability to think in terms of OOD and MVC design. It takes time to master it, still absolutely worth the effort.
Man... I did ssr on angular before React devs even learned word "ssr"
17:30 this is one of those Wiz dependencies that Angular can use now
The best web framework
Love to see this.
I am a senior web dev in my company. We are using Angular 12v. We are block on the migration to the latest due to material UI 15 breaking all our styles and it doesn’t have a support for old version. Angular 18 has to use material UI 18, no backward compatibility.
I already did the migration for angular 18 on our angular 12 and the material UI kill it. We decided to move to NextJS. I am happy that Angular is becoming for like NextJS
Material has a backwards compatibility mode.
angular also has migration tools
Next time try to update not only in 10 years
I like this a lot because I can bring a lot of the new stuff into existing projects and the other devs who might not pay as much attention can just use it without much hassle.
Thanks for the praise and review! We're delighted! 😃 Will you update the FAQ on your website now?
I had entirely forgotten about that lol, just updated!
Time to master Angular!
Start from 2:42
You asked how many new angular applications get made and let me assure you that especially in the government sector its quite a lot.
Being an angular developer for the last couple of years, for me, a regular ng update should not bring too many breaking changes or costly innovations. Having a large repo, quality and stability is very important, more than the productivity gains of a certain new feature.
i am an angular dev. and i love what they are doing. they address so many pain points i was frustrated with right now.
In the past, Laravel frequently released major updates, making it challenging for developers to keep up. Angular is currently on a similar path. They should adopt a one major release per year cycle.
Angular is the best Frontend Framework to ever exist.
Except it isn't
Looks like an easy upgrade for huge apps and thus makes React less important as we were considering moving to Svelte (because no one wants React cancer).
Before this changes i was seriously considering going all in on react, ive only used angular and vue and im concerned about open roles using angular.
Im hoping this makes more angular roles available outside of just huge enterprise applications.
I'll have you know that we don't live in Ivory towers. They're marble and porcelain, thank you very much!
We had angular 19 before react 19 😂
Standalone and isolated components ARE a good thing, and I am really baffled w/ Angular devs who yearn for ngModules.
Angular modules were introduced before we had an industry standard module system on the web (ESM). The Angular team recognised 4 years ago that they have outlived their usefulness, and over 2.5 years ago (and since) they have made all features work without ngModules, leaving them in for interop as syntactic sugar.
Yeah, parts of my code base written with standalone components instead of modules has a much better dev experience, with unit tests running faster, etc. because of the better tree-shaking. It's also so much simpler to see what your dependencies are at a glance.
The 2 things I'm not sure are a good things are inject instead of a more common constructor injection, and the functional guards (having to use assertions to make sure the guard functions will not be used in a context where there cannot be injection seems worse than the previous guards with services implementing the interface).
UA-cam is not going to be rewritten in Angular :) rumors of wiz and Angular mergin were greatly exaggerated by one of the parties. What is going to happen is that they will share more ideas and some dependencies (like signal implementation)
Yep, they'll just take the best parts from each other. And it won't happen immediately - that's their long-term plan, unfolding over the next 10 years.
@JLarky for whatever reason I see you everywhere, even on Ryan's Solidjs channel
@MengLinMaker what do you mean even? That's my main place
“They’re angular devs for a reason”…..that hurt 😂
It's true too xD
Regarding stagehand, I will be always in control of the code of the right side and it's guaranteed to work always. I never gonna use stagehand for testing purpose. Yeah definitely agreed, it has it's use case in other places
@@sohamnandi7526 yeah I didn't want to be a naysayer without actually investigating it but if I was doing the advertising for it then very first point to be made would not be shorter set up code it would be how they solved the repeatability issue so the ai generated tests are actually the same every time.
The autocsp is huge, and should be making a way bigger jump up and down as that will make angular app’s security wise improve majorly and from the start of a project having that makes a great dev experience as I have wondered about csp things in typescript/js apps for a long time
if youtube is made using the faster one I don't want to know how bad would be using angular
Although I almost exclusively live on the back end, Angular has been the most approachable framework to land into. It seems like there was just a natural right way to do things. Dependency Injection, templating and RxJS were great. With react there seemed to be a footgun around every corner blowing up the reactivity of the DOM, or making elements render a million times too many.
I'll never be a javascript person, but if I have another go at a more frontend-leaning app than HTMX can offer, I'll pick Angular again :D
Angular user here. Started learning between 17-18 and one of first adopters.
I've got a couple critiques on new stuff hope it reaches the right people:
- linkedSignal breaks the reactivity naming convention ( signal, computed, effect, resource ... ) and "linked" is enough imo.
- resource (and rxResource) are just inferior versions of tanstack/query (and ngneat/query) and I'd advise to use them instead while resource is still in dev preview and may get better
I build my first startup using angular and Ruby. It’s working since 2015 without any issues
dude you have the coolest sponsors
🫡
31:04 I actually think if an app I was using had filtering like this I wouldn't want them to drop my selections just because they are filtered out of my current screen. Cause if I know I want a few things from one category and a few from another, it'll be shorter to look for them when I only have to look through their category.
You are probably right that not being able to undo the selection is an issue though. But I'd say put the selected values up at the top of the page(if they were filtered out), and the values that still remain after all the filters just continue as they are now.
As a mostly end-to-end testing relying blazor dev, I've been looking for a tool like stagehand for a while and it seems really cool! Can't wait for it to support local llms like llama with ollama.
I‘m hyped and worried at the same time. It‘s nice to see angular evolve and care about performance. On the other hand I fell like they might destroy some of the core benefits of angular and there are also some long existing bugs that are not beeing fixed. I hope they don‘t just follow hypes and trends blindly and move in different direction once the direction changes.
Style could be iterated in browser's dev console. Probably that's why it was low priority for HMR.
damn fresh video, 3k views
I love angular
That sponsor seems really cool. Really cool. Wow.
19:45 you said you hate the syntax. Was this just instinctive? in that you prefer react style vs traditional angular? Or did you not notice that angular has new control flow syntax?
It felt like you probably were already decided you don't like angular syntax, and didnt grok that it was actually very react style now.
Svelte style*
@@TayambaMwanza ah is it? Thanks for the clarification. I am quite limited with experience of other libraries. I've just stayed with Angular as I wanted to become good at one rather than distracted by the shiny things and just know bits of all of them.
@@TayambaMwanza yeah I don't know why they decided to use @ instead of # . Why not just reuse Svelte's syntax? Haha
@marcuss.abildskov7175 community requested it actually in the rfc.
Angular team was into # but after community suggested @, they did further research and changed their minds to what community wanted.
Interesting to see ".. I'd pick Angular" with enthusiasm of passing kidney stones (19:50) 😀
I first got exposed to Angular after working on big Ruby on Rails projects. And the cool thing at the time was the similarities in scafolding between RoR and Angular. But now, with the simplicity and control you have with the JS ecosystem wether it be node, deno react and so on, I find the boilerplate and paradigm of Angular quite frustrating
Theo talks about Angular devs like a primatologist talks about babboons
It even pains him to complement angular at all.
In the near future the signal based forms, routes and new component author. it will be a totally different framework
@defer doesn't seem any weirder of a syntax than throwing a random string like "use client"; in the code and having that do something meaningful.
SSR was added in angular 16
Or v5 if you used the community package @angular/universal
Cool to see all the improvements in Angular but I still like Nuxt better.
We run a ecommerce shop on angular and it's way faster and snappier then the demo you showed. If you profile the code and avoid patterns that tank your performance, then angular could always be fast. zone.js and animations are the worst performance offenders, and we got rid of them 4 years ago already and use observables and native css animations instead.
But angular is a SPA. you don't have problems with SEO?
@@professor_ozzy SSR solves that. Our product pages answer in 220ms - html generated by angular included. But angular SSR isn't that fast by default. We wrapped SSR to work like an SPA as well. We bootstrap common pages ahead of time, then when a request arrives, it just applies the changes, as it would in the browser, yielding in SSR answering in ~40ms.
They put an artificial delay on that demo though...
@@alexandervonweiss2927 Cool!
@@TayambaMwanza Ok, didn't know that. My bad.
Regarding the 2700 thumbs-up on github: This isn't the bar to clear for an issue to be addressed, this is the cumulative number of thumb-ups across all issues that are addressed in this release.
I am trying to understand - is the goal to eventually build its own version of NgRx and RxJS with signals or equivalent?
I am an angular enjoyer, and no I do not use Next.js, I use React and sometimes Angular ,dabbled with Swelte and that's about it, other than that I'm doing Go and Java and the thing I love about Angular is that the way its developed is extremely opinionated by the devs of the framework unlike other frameworks where we can go crazy, I do not necessarily want to go crazy I love the fact that everything has a way for you to do it, ish~
Also I hate things like React the more i interact with React codebases because all of them are so different, and its pissing me off, but in Angular, its always the same but maybe with a bit of flawour.
Angular glazing? Insane
Remix: Why do we have to wait for Javascript just to go to another page or submit a form?
Angular Community Member here, we had a lot of push back a few versions ago from people who didnt want Angular to change, as time has gone on and the benefits have become clear 99% of Angular devs are on board, id say its practically a non issue by now.
Remember you're seeing only V19, theres been several RFC and conferences talks/announcements that Angular devs have been exposed to, so we've passed the pushback phase already.
which Angular version was most downloaded in the last 7 days?
v9 or v19 ??
@hengkeatyam3700 silly question in the software world why would a brand new version that's not even had hot fixes yet have more downloads than the older one?
@@TayambaMwanza
Angular v4 was the first stable version.
Then, Angular rewrote the rendering engine.
Angular v9 was the first stable version of the new Ivy rendering engine.
After that, internal conflicts emerged within Angular.
By Angular v14, new coding styles began to appear.
In Angular v19, these new styles became over 50% stable.
Angular Material v18 doesn't include a single line of code related to Signals.
Angular Material v19 includes a small amount of code using Signals.
Angular Reactive Forms are still not related to Signals.
Angular Router and HttpClient still rely on RxJS.
Why did we learn and write Zone.js, decorators, RxJS, and NgModules, when now they can be optional or even no longer considered best practices?
Was it because of new technological breakthroughs in front-end development in recent years?
No, it was simply because Angular chose the wrong path.
They spent too much time on rewrites and migrations.
sorry, I misunderstood your word "99% of Angular devs are on board", my false😅
@hengkeatyam3700 oh, no problem :)
Angular defer directive is closer to what react lazy does than to ssr=false. To be honest i don't know if they do SSR of it or not, so i guess maybe it's both ssr=false and lazy? :)
Question: will this start to steal React’s thunder/popularity?
Crazy how anglular not really solves problems I didn't have in the first place. 🤯
Queues can be dangerous for input, but I'm guessing that the angular devs have somehow addressed it. For example, imagine a user clicking a button and seeing nothing happen, so then they click it like twenty times in a row and/or get frustrated and just start clicking things all over the page. If everything were queued suddenly all this unintentional input is captured. I can't remember the context (what language/project, don't think it was even web) but I realized this when using a queueing approach to handle missed input. The better option was just to let the interface not do anything until it could be handled. Anyways, like I said I'm sure there is more sophisticated logic in angular.
They using a library form Google internal framework to handle it, it's the same one used in the Google search page apparently.
Browserbase is not available in Malaysia
6:57
Just try angular, make a relatively big project with it, and watch your development server explode with updates taking 1-60 seconds each update you make to components that are used in a lot of places. The experience is definitively astonishing.
Wow how bad are you in coding Angular?
@@pakoottox262
I don't think the extremely large loading times are due to me being "bad at coding in angular", feel free to prove your statement tho.
They introduced HMR now. It's working quite well and should be enabled by default soon.
I don’t like the standalone side of things. I’m still using non-standalone because it just feels more familiar and organised
25:43 I was using browser dev tools for that as well :(
I like the SSG, been using the SSG recently and even if it still feels weird to configure with Vercel compared to Next.js which is without custom shit, its good and working properly
This won't affect 80% of angular projects. Most angular projects are stuck in 12-17. Their managers need a business reason to do this and all the in-house optimizations they made to make it work fast will break when they bump the version. Happened to all angular projects I worked on. I don't discredit that this is a bad company culture tbf.
I am fine with this as an Angular Dev and I have been around since the AngularJS (hated it back then). I am not a NextJS (I see its advantages) person and like HTMX/AlpineJS & Astro for most of my personal stuff.
ZoneJS issues is that you would click a button to update some counter on the same page and it would update the whole page. With the way it should work 17+ it can be it just updates the counter.
Old APPs in Angular are not getting updated most likely, no matter what it takes about 1 year (depending on size) to update from Angular 3 to 11 let alone 19. Also there are things like Bootstrap or PrimeNg behind some of the components which is more of a pain to convert most of the time.
I am an Angular dev with a next js side project. Got me there :))
“…just write an English instruction and it does the right thing”. English is ambiguous and not only that but AI can’t interpret it as good as a human dev yet. The right side example is longer because it’s explicit and way more “deterministic” if you will. The magical thing about programming is the ridiculously high exactness of control you have over the computer.
(wrote at 3:26, not a dev I am at uni programming as a hobby, I critique words and opinions not people)
i enjoy angular a lot... but i understand the fallout from the red wedding. at work, half our code is still angularjs since our upgrade path is unclear... to big for us to re-write. tried to get hybrid conversion.. (had done it before with smaller app)... dont think its do-able.
I think the constant change of Angular is not a good thing. It means constant breakage and need to rewrite old parts to the new ways of doing things. It puts additional load on the developers to constantly learn new things and discard them after the new best thing appears. It makes you feel that everything you learn about Angular can soon become useless.
The new version is 100% backwards compatible with the previous version. The old code still works.
As good as the new updates are I'd instantly trade this just for React like statefull fast refresh
angular dev since .js ... actually angular updates broke my apps every version up to maybe 4 years ago... it almost kick the ecosystem... but i somehow keept the trust (probably also sunk cost fallacy)
zonejs was annoying because it was a 'huge' payload and even required if you used onPush (alternative way of change-detection - not relaying on it). Also it wasn't possible to configure it probably without other 3rd party libs - for example default change-detection would trigger updates - for example on mouse-move even if there was no reason too and you couldn't opt-out of that behaviour easily
I like that you got a bit over the unreasonable Angular hatred. Nothing is ever perfect of course and Angular has some messy parts. But c'mon - next.js is SUCH a mess too. Using unstable versions of React has lead to serious issues and headaches. React 19 relased last month FINALLY, but before that it absolutely was BLEEDING edge to even use next.js.
It really IS a good sign that they look to other frameworks and copy some of their best ideas, while still keeping their own identity.
Browserbase ad maybe more interesting than the Angular news 😄
Mini Series - What Happened To Theo? 😁
Angular 1 999 🙌🙌
The Angular team doesn't care about the number of users. Even if 5 developers use it, it gives a major update every 6 months. The project I started with Angular 10. Now it is in angular 19. Only the stanalone component took me some time. I only came up to 19 with migration scripts. This was not the angular I was used to that I learned at the beginning. The project became like someone else's project. This is the negative aspect.
Tried stagehand is hilariously slower than writing the code yourself, too early
i thought Angular died, how is it still going on, i used this pointless thing at version 6 to 7, i hated it
"The rumors about my death were greatly exaggerated"
Good for you.
11:30 RSCs seem weak compared to resumability. Not interested in them at all.
Idk how many viewers are hate watchers like me. Mind you I like Theo as a person/youtuber, the thing I hate is web dev and seeing him just casually mentions hundreds of frameworks and APIs a second with somehow no overlap is a beautiful reminder of why I'm never going back to working in this world.
Sadly your concerns with the angular community are real. The angular subreddit has been a shitshow of insults towards anything that isnt "traditionnal angular" as of recently. They even insult newcomers coming from other frameworks that are trying to learn Angular, just because they were "on the enemy team".
The rest of the community is not like that by now, haven't been on reddit in ages though.