Md. Redwan Hossain

@SeniorLazarus @anshuc @github Seems like the same scam where company owners fires up their own business just to claim insurance.

@anshuc @github This was set up. Just plain robbery for training purposes.

We are investigating unauthorized access to GitHub’s internal repositories. While we currently have no evidence of impact to customer information stored outside of GitHub’s internal repositories (such as our customers’ enterprises, organizations, and repositories), we are closely monitoring our infrastructure for follow-on activity.

@mjovanovictech Create, Update, Get, GetAll, Delete. Each one has now atleast 2-4 files. It was before just controller, service, repository (not mandatory), dto.

@redwan_ping How did you end up with 20 files? What are you even doing?

Vertical slices are not modules. This is where I see a lot of confusion. In a modular monolith, modules solve the macro architecture problem: - Where are the business boundaries? - Who owns the data? - How do modules communicate? - What public API does each module expose? Vertical slices solve a different problem. They help you organize code inside a module. A module is a bounded context. A vertical slice is a feature implementation pattern. For example, your `Ticketing` module might expose a public API and own its database schema. Inside that module, you can still organize each use case as a vertical slice: - request - response - validator - endpoint - handler - data access All in one folder. That gives you high cohesion at two levels: 1. The module keeps related business capabilities together 2. The vertical slice keeps everything for a feature together And the nice part? You don’t have to pick one internal architecture for the whole system. One module can use Clean Architecture. Another module can use vertical slices. A third module can use something simpler. The module boundary makes that safe. The real question is not "Clean Architecture or Vertical Slice Architecture?" It’s: What structure gives this module the most clarity? I wrote a full breakdown of where vertical slices fit inside a modular monolith, and how I think about macro vs micro architecture: milanjovanovic.tech/blog/where-ver…

@kristijan_kralj Write the entire codebase in Program.cs file. Easy right?

@hiwinit Never. I just hate the idea.

How many of you actually use PostgREST?

@shadcn Make it possible for angular.

How can I make shadcn/ui better for you? What do you need?

@catalinmpit That's why signal driven modern angular rocks for internal application. The core pattern remains the same for 10+ years.

FINALLY! I've been saying this for a while. Most apps would be just fine using React, Vite and TanStack Router/React Router. Sadly, people love to overcomplicate things.

p95 went from 2270 ms to 217 ms. Throughput went from 168 RPS to 1,238 RPS. I changed exactly one line of code. Thread.Sleep(200) ➜ await Task.Delay(200, ct) That's it. Same machine. Same 400 concurrent users. Same 200 ms of "work." The reason it matters this much: Thread.Sleep parks an OS thread. The thread pool only grows at 1–2 threads per second after MinThreads is exceeded. So with 400 VUs hammering a sync endpoint, requests aren't slow because the work is slow, they're slow because they're queued waiting for a thread that doesn't exist yet. await Task.Delay returns the thread to the pool while it waits. Suddenly, 16 threads can serve 400 concurrent requests instead of 16. I built a .NET 10 lab project that reproduces this on demand and produced the numbers above on a Ryzen 7 4800H. Code + k6 scripts + step-by-step breakdown. Will share it in 2h in my Newsletter. Join 20k+ engineers and read it: thecodeman.net

@cagataycivici Awesome. Can't wait to try out.

Official support for Angular Signal Forms ships with PrimeNG v22. ✅ #angular

@MrMatthewLayton @mjovanovictech There’s a reason. Exception types are optimized for throwing. Results types are usually based record struct in C# for optimization.

@redwan_ping @mjovanovictech Exactly, but most implementations I've seen don't return exceptions. They use their own Error types, instead of reusing Exception types.

C# 15 is getting union types. That’s a big deal. It means patterns like Result and Option finally have real language support. No more awkward wrappers. No more hacks. No more pretending exceptions are the best way to model failure. I’ve advocated for the Result pattern for years, but lack of native support always made it clunky. That changes with union types. Don’t want to wait? Here’s how to implement the Result pattern today: milanjovanovic.tech/blog/functiona… If you still prefer throwing exceptions, you can skip this.

@MikeTreit @KarenPayneMVP ORMs are great for DML.

@KarenPayneMVP ADO talking to stored procedures. No direct SQL. ORMs are for toy projects.

What is your preferred method to interact with data for C#: Microsoft data provider, EF Core, Dapper?

@systemdesignone Any type of engineering is complex. The CS field is obnoxiously bloated with non-engineering mindset and the outcome is this tweet.

Facts about Docker >Adds unnecessary overhead >Makes networking harder >Introduces security risks >Turns simple apps into complex ones >Increases operational complexity at scale >Makes debugging difficult >Bloats with inefficient images Why do people like this thing?

@MrMatthewLayton @mjovanovictech If you don't throw exception but returns it, that's just result pattern.

@mjovanovictech Exceptions can still be a good way to model failure, if you return them, instead of throwing them.

@freedomschamp @mjovanovictech what does you expect? It should not get updates?

#Angular21.2: New FormRoot Directive for Signal Forms 💥 The latest minor version introduces the FormRoot directive, which simplifies handling form submissions in #SignalForms. ➡️ Automatic Setup: Applying [formRoot] to a < f o r m > element automatically prevents the default browser submission behavior and adds the novalidate attribute. ➡️ Integrated Submission: You no longer need to manually call event.preventDefault(). When the form is submitted, it automatically triggers the submit() function on form. ➡️ Reset: You can now pass an initial state directly to the reset() method to clear both the interaction state and the model values simultaneously. What do you think of this addition? Let’s discuss below! 💬

@jfversluis SharpOutcome SharpServiceCollection SharpPersistence

Question for .NET developers 🤔 What's the one tool or NuGet package you install in every single project? For me it's CommunityToolkit.Mvvm. Can't live without source generators for observable properties.

@Jean__Meche @service Why not? Class in JS is just syntactic suger. Though in TS it is first class citizen. Still, for a frontend framework, functional components and service will be a lot easier. Just look at ngrx signal store, it rocks despite being functional.

@redwan_ping @service Are you sure about that ? 🙂

Angular will introduce a new decorator in v22: @Service - Root by default - inject() only It will stream line how we define services and reduce even more the boilerplate. github.com/angular/angula…

@kristijan_kralj So you always just used EF Core? An enterprise app can have EF Core, Dapper, SqlKata, and what not in the Persistent layer. Will you use all of these in the Application layer??

@iamevrenbal @cagataycivici angular is very stable even for experimental stuffs. Most of the time you can just find and replace. Also, now agentic AI can refactor code rapidly.

@cagataycivici Definitely the right decision. I'm extra careful with experimental features. I learned this the hard way with a Next.js experimental feature. Spent 2-3 days debugging just to find out the API changed after an update.

At PrimeNG we tried to support Signal Forms for v21 final but the API is way too experimental, will probably wait for the stable status before official support. As a library dev we must support three form apis of Angular in same codebase 😱 - template driven - reactive forms - signal forms