Can you wrap a component with a directive with ng content wrapped with defer block just for not typing defer on viewport prefech on idle multiple times? Just put a directive and wohohoho, it is posible to do?
Some developers are afraid that their jobs will soon become obsolete because of AI. This is especially the case for Frontend developers, as it is believed that user interfaces will soon be easily generated with AI even by non-developers. What do you think?
Hi, Angular is often used for apps in the finance industry where you can't use tools like Gemini due to privacy concerns. I have two questions: Is there a way to use Gemini privately without sharing code with Google? Will there be any open-source LLMs from Angular (or related) team to use locally on a machine?
When can we expect Angular Resources to move from the 'experimental' status to 'developer preview'? Additionally, is there a timeline for when guidance will be available on using Resources for POST requests (server-side mutations)?
Thank you for your nice presentation! Really helpful! :-)
My question is for Younes Jaaidi.
What architectural measures you will suggest to reduce flaky tests in large testing suites?
And, What are possible integration available for Playwright into CI/CD pipelines ?
Thank you in advance!!
When can we expect Angular Resources to move from the 'experimental' status to 'developer preview'? Additionally, is there a timeline for when guidance will be available on using Resources for POST requests (server-side mutations)?
Could you divulge a little bit of the process when you release or you work on a feature, is the update/migration of the google project automatic? To have perspective could you give us a timeframe of how long that release/deployment (development and production) takes?
For the angular team. Really glad that Vitest will be supported by CLI and default from the new version coming. We also saw today a talk about writing playwright tests. Do you recommend it, are there plans to integrate it? What strategy do you use in your team and in google in general for e2e?
Can I somehow patch the compiler, so I can replace one component with another during the build? Assuming those components would implement one interface. The main purpose would be in large monorepos, when eg. I have multiple applications and I have the form input component to have different implementations for web and for mobile apps, but in my form components, I would like to use only the “common” form input component. So I do not have to have duplicated forms for every application, just because I want to render some inputs differently
Can we dynamically create signal stores, form stores, and validation schemas at run time based on metadata (required, read-only, etc.) for each field defined on the backend?
For NgRx signals, we expect that Angular devtools will support signals, which can be used to track state. I don't know if this has already been mentioned! But I'm mentioning it here. Thanks.
New architecture looks like completely new framework. Is it better than React? If I need to learn everything from scratch, isn’t it better to learn React instead, which has broader market share and more positions to apply for job?
How to get to Angular core team? Do you have to be at Google for some years in other teams or is it easier to contribute open source (for free) for few years and then be hired?
Would anyone be able to provide more insight on why Angular Material was refactored to reuse some kind of shared (abstracted) implementation of Material web components: I am mostly interested how the Angular team sees interoperability & compatibility of Angular with web component based UI libraries / design system implementations (e.g. Adobe and SAP went web component first).
Hello, have already started using the NgRx signal store implementation for new features, on the other hand, I have a lot of old NgRx store implementations. Is there any documentation or playgrounds to facilitate migration to signal store? Thanks.
Hello, first of all, I would like to thank you for these presentations. It's incredible to see Angular develop. My question is: why is the signal effect (() => ...) not recommended for use and what are its side effects? Thank you.
Supposing an app has many lazy loadable JS chunks. During SSR, if a few JS chunks are dynamically imported, is this info transferred to Browser, so those chunks are preloaded as soon as possible in the Browser?
Is It possible to define @defer/hydration triggers dynamically? (e.g. trigger specified in a CMS component’s data, but not hardcoded in HTML template)?
5 or 10 years down the line, what would you like the angular framework to become? Something that fades into the background, giving the dev a pure experience of web development. Or a fully featured language where devs don't write TS, they write Angular. Or something entirely different?
Are there any plans for creating an official component library (or moving Angular Material in that direction) that could live inside user’s codebase, like shadcn in React?
Can we expect future Resource integration with Router Resolvers, making the resolved data available inside the component as well - for example, to trigger a reload?
If GenKit and Gemini are so great products, why Angular core team is still spending so much time on GitHub issues, reviews, PRs. Why they can’t spend more time on new features and let AI fix bugs and maintenance issues?
Shall we use signal events or not? Some people say it's just a fallback to classic store and that's conceptually different than signal graph pattern. Personally I enjoy the redux pattern and it's separation of command from the whatever going to happen.
Would it be possible for resources to be “enabled” only when certain conditions are met? For example when 2 signals contain value. (Similar to tanstack query’s enabled property)
Everyone is talking about how amazing AI is, but how much does google and the angular team think about the COST? (i’ll intentionally leave the definition of cost vague).
What is your perspective on using a custom DI layer in a Clean Architecture setup for Angular applications? Given that Angular already provides a robust dependency injection system, does implementing manual injection add any real value?