today at 5:23 PM
In my experience, C# and .NET are both "underappreciated" in a large part based on the legacy.
Some folks out there still think .NET is Windows only. Some folks out there haven't looked at C# since the .NET Framework days. Some folks out there are still salty about some of Microsoft's decision around OSS (debugger, hot reload fiasco, etc.)
But day to day working with C# and .NET for backends? It's really fantastic. AI seems to like it a lot and a lot of tools for agents to go HAM. Roslyn analyzer, source generators, workable hot reload (for a compiled language), tons of really good docs from Microsoft, generally low variance and high consistency from codebase to codebase (e.g. encountering Drizzle vs Prisma vs TypeORM vs Kysely vs ... in Node land).
My previous team switched from TS to C# and F#[0][1] and it was a blessing; a lot of the platform ops have been simplified. It's just one very solid tool chain and packages are generally stable (a lot less effort on maintaining platform config/tooling and more focus on just building). First party packages get consistent security updates and C# doesn't have nearly the same level of exposure to transitive packages that one encounters with JS/TS ecosystem.
C# feels like a really solid choice in the age of AI coding because of it's general stability, strong static analyzer, insistence on types, and really good tooling. Not right for all use cases (e.g. CLI, serverless functions), but fantastic for backend APIs.
If you're looking to do AI with/on C#, I wrote about some of my experience with making codebases successful for AI[2].
[0] https://engineering.usemotion.com/moving-off-of-typescript-e...
[1] https://typescript-is-like-csharp.chrlschn.dev/
[2] https://chrlschn.dev/blog/2026/08/the-unexpected-ai-stack-cs...
today at 6:44 PM
Being a Microsoft product is such a heavy baggage. Last time I checked the official debugger was still not open sourced and that alone was enough to make many people think twice. The fact that MSFT chose golang instead of C# for Typescript compiler, the state of Windows native GUI app dev... all erode C#'s reputation.
It's a good language though. The latest trend of hating OOP is unjustified/overreaction. I do write Rust too and I am not really convinced that traits are better than shallow inheritance + interface-oriented classes.
today at 7:38 PM
> Being a Microsoft product is such a heavy baggage
today at 9:41 PM
TypeScript has fewer competitors for the frontend than C# does for the backend
today at 9:45 PM
https://gist.github.com/matthiasak/c3c9c40d0f98ca91def1
JavaScript is one of the most common compile targets. "Who's the better JavaScript" was the most contested field a decade ago. These are not just toy languages - Facebook has Flow and Google has Dart. Both companies are far more web-adjacent than Microsoft. Yet Typescript won. It won so hard that we feel it has no competitors today.
today at 8:07 PM
The decision to use Go for TS7 compiler is even more baffling considering Go's low level memory management functionalities is nowhere near the level of C#'s Span<T> etc, especially since TS and C# are invented by the same person.
I got so worked up by it I built my own TyepScript compiler with my clankers in C# to prove that it can be done, my revival of the Common Language Runtime concept, project Copeland:
https://github.com/yuechen-li-dev/copeland
Currently, Copeland compiles a dialect of TypeScript to JavaScript, C#, WASM via Blazor WebAssembly, and NativeAOT, runs in csproj, and uses xUnit.NET for testing framework.
(The other reason is that Microsoft still refuses to do the obvious thing in .NET 11 and take the best part of Rust and let enums be `match`ed, despite the fact that both TypeScript and C# have weak enums historically, so I added payload enums to my TypeScript dialect because Microsoft won't.)
today at 9:51 PM
I love C# and F#, but every .NET shop I worked for was full of MS cool aid drinkers and enterprisey nonsense. I don’t miss the culture at all. The languages are excellent, though.
today at 10:00 PM
Exactly this. The workaround is to explicitly hire developers from other languages whose idiomatic use matches your style.
today at 8:18 PM
I've been writing C# for literally 25 years.
I still choose .NET as the go-to whenever I can. On Windows or Linux.
I think the baggage view was valid for a while - but now we have a new generation of developers who only know about .Net Core and may not associate the Framework years in quite the same way.
today at 7:28 PM
I recently started a new project on .NET, mainly for its batteries included approach, seemingly a bit more serious ecosystem and the tooling AI agents can leverage during development and ops. Really happy so far, but too early and too small to have meaningful conclusions yet.
I'm especially interested in exploring F# as a concise, non ambiguous way to communicate definitions between humans and agents, somewhere between writing code like the olden days of yesteryear and describing things in English prose.
today at 7:35 PM
If you're going heavy into AI coding with C# (like approaching 100% AI code), give my 5 part writeup a read. Battle tested configuration with a fast moving team that was pretty much 100% AI code with good code quality. The configuration for `csharprepl` in particular is a special kind of unlock; having the agent be able to reach into the running DI container, wrap/replace functions while app is running just unlocks a lot of capabilities for the agent. It's the difference between an ultrasound and reaching in and manipulating the organ.
today at 7:49 PM
Oh, thank you, this looks really helpful at this stage!
Yes, the plan is essentially 100% AI code with a small team of programmers and domain experts from other technical backgrounds. But still depends on a few things on the business end before I can really kick it off. More of an exploration until then.
today at 6:27 PM
It's certainly not a big topic of discussion online but it's powered all of the healthcare infra my career has been around. There's a lot of "dark matter" code written in C#.
today at 6:43 PM
The fact that when most devs look at C#, they can only see the silhouette of Microsoft is kind of making me mad sometimes, but I think it is understandable.
The corporate history of C# is undeniably strong, and even till this day in modern C#, they still have a lot of legacy components in it, like AppDomain, APM and MarshalByRefObject.
It is also unfortunately that C# has largely lost in the AI race, that it failed to gain tractions from AI developers despite heavy investments from Microsoft. Early AI stuff like Semantic Kernel actually included C# as its primary languages besides from Python. Nowadays it is not mentioned so much and the default is just Python but perhaps it really isn't a fair comparison due to the gravity of Python in AI as a whole.
Yet for another fact that even within Microsoft, Rust wins over C# for agentic development, is a clear indication that C# has lost a meaningful place in software development as its former parent started to berate it as well. TypeScript is rewritten in Go rather than C#, and consider the fact that Anders Hejlsberg made both C# and TypeScript...
Sure, maybe Bing and a large part of Azure and Xbox services still have heavily relying on it, probably because of ASP.NET Core and Orleans, but outside of the two Microsoft greenhouses, it is a free for all battle royale where C# don't even have a chance of being picked at due to being too weak.
I even tried to contribute a C# SDK to the Restate project, but the maintainers actively rejected it due to them not being familiar and refused the handover. They said they will consider if there are commercial fundings backing it, then they would starting to consider (that's even with money attached they would still think about it). This also shows how unpopular C#/dotnet is among open source development and failed to build hype. The same Restate project even have a Java SDK for it, so it is not like they can't handle C#, they just didn't like to.
The last few thing C# is good at is just Unity I'm afraid, and that is also dying too as Unity is also not doing so well lately.
today at 6:51 PM
Rust and C# serve different purposes. For things like CLI tooling and the Copilot SDK (supports Go, Python, JS/TS, Java, .NET), Rust makes total sense.
For web API backends and MCP servers, I firmly believe C# is the better choice.
For AI use cases (as in make calls to LLMs), there's not much difference, IMO from Python and C# and TypeScript. At the end of the day, those are all just REST API calls and SDKs are mostly wrappers around it.
For LoB apps where the AI/LLM is just one small leg of the actual flow, C# is still a winner, IMO, because of how good the overall stack is (e.g. EF Core). Orleans (actor model framework) is also super interesting and I'm exploring some memory/communication system based around it.
today at 8:50 PM
It’s also awesome for SREs / SysOps / Whatever We Call Ourselves Today:
Opinionated in a good way, which means almost all C# apps have the same parameterisation with environment variables, command line, and config files. It’s all very systematic, consistent, but flexible and pluggable.
“Node has that too!” you’re about to say. Lol, no. It has bits and pieces the devs have to wire themselves, which means a fleet of vaguely related apps made by independent teams will have wildly different config styles. Good luck consistently making all of them pull API client certs from KMS or Key Vault, or whatever!
OpenTelemetry is now baked in to the core components and rapidly spreading through other libraries.
Excellent “pervasive” system APIs for spans, traces, metrics, and logs means that the sources are omnipresent and very rich, going back decades. Even very old third party libraries are often a breeze to instrument.
Debug snapshots from production crashes on App Service feels like cheating.
Built in container build outputs with no extra tooling.
Sane defaults for its package manager means far fewer nasty security surprises.
Far more first-party “coverage” of essential capabilities in the optional Microsoft packages so you don’t have to resort to NPM where your security-critical code is written by Sergei and support is “LOL Wut!?”
A GC that automatically responds to container limits instead of the host limits.
It uses thread pools for async tasks in most configurations so whole categories of production-time issues that occur with single threaded runtimes just doesn’t happen.
Etc, etc…
today at 6:23 PM
How does AI do with F#?
today at 6:50 PM
AIs write pretty good F#, but there simply isn't very much of a reason to use F# over C# since Microsoft basically uses F# as an experimental test bed to test .NET features before bringing them into C#. Plus, annoyingly, F# doesn't interop well with existing C# projects which there is even less of a reason for existing .NET shops to adopt F#.
today at 7:04 PM
I also agree with this. When we were internally finalizing on C#/F#, I showed the CTO that most of what he liked about F# was already possible in C# (though requiring third party libs like OneOf or Dunet) with the benefit that C# syntax is more familiar for most folks. With 15, I think they are even closer now.
today at 6:37 PM
Motion shifted to C# around August of last year. Part of the reason, I suspect, is that the CTO (Hi Chander!) is an OCAML fanboy and F# was the most practical platform alternative.
We shifted all new product backend build to C# (because at this time last year, we were still "AI assisted" and deemed F# too foreign for most devs to read well), but all internal tooling was built on F#.
By maybe May of this year? We had multiple F# backend codebases, fully AI generated. Only one person on the entire team had worked with F# by hand. Several things work in F#'s favor even though it has a lot less representation in training: 1) the people that write F# really love F# so the code out in the wild is going to be better quality (hunch), 2) F# is a bit more terse/dense than C# and less noise in context, 3) F# benefits from the same tooling stack, 4) F# type system is still a bit more expressive than C#.
I know C# deeply so I know what good looks like and I know how to get agents to write good C#. F# I'm less familiar with, but we had great success with it.
today at 7:16 PM
Built the massive enterprise backend of a super app fully in F#. From my experience, AI was able to write idiomatic F# code since Opus 4.8 (all models started doing very decent things around that period).
today at 7:22 PM
I think C# would be a great choice for both CLI apps and server less functions.
Other than that I totally agree with everything you said.
today at 7:31 PM
> I think C# would be a great choice for both CLI apps and server less functions.
today at 8:43 PM
I think by CLI apps they mean something more like Claude Code than ripgrep.
today at 4:13 PM
C# was my 2nd programming language after VB6. I remember in the early 2000s when I was 12, using my best friends computer to try and download VB .NET (he had faster internet) just because it was free, then being extremely disappointed when I couldnt just run my VB6 projects I wanted to run on it... Some years later, during 2007 ~ 2009 I tried out C# and that was life changing.
I wound up going to college after working retail at Disney, then after college, my first job was supposed to be C# but they didn't have the project approved, so I got my first exposure to Python - which I loved, then JavaScript - which I hated but have a love-hate with to this day, and finally C#, but it wasn't .NET 3.5 no, they started using .NET Core, which was... something else, lots of teeth pulling during the original .NET Core days.
So many wonderful projects made with .NET my only complaint is I wish Microsoft would make a cross-platform GUI stack they'd actually stick to, and make it work on Linux too. ;)
I know there's alternatives but nothing beats telling your coworkers "lets just use the built-in" when choosing a GUI stack.
today at 4:29 PM
We abandoned .Net years ago before our third major product rewrite prompted by withdrawal of various bits of tech they said was the next greatest thing. AppFabric, WWF, WCF and Silverlight hit us hard.
Once burned, twice shy.
today at 4:33 PM
That's fair enough, but the dust has largely settled on dotnet core. The death of Silverlight was long ago. I think the only thing at risk of abandonment is the native UI situation, which is a mess.
today at 4:54 PM
Hard to say what's native in Windows anymore. I imagine Win32 APIs and more recently Windows App SDK?
The most supported UI frameworks from MS for Windows are WPF and WinUI 3. UWP is dead and WinForms (even though closest to something native to Windows) is probably going to end up in the land of forgotten (sorry nostalgic Windows devs, you have to face it eventually).
A "mess" is an understatement when the people who themselves develop Windows desktop apps have trouble explaining the landscape.
today at 6:01 PM
> Win32 APIs
Everything boils back down to Win32, including WPF, WinForms, WinUI2, and WinUI3.
today at 10:24 PM
Some of the more modern stuff is in a separate WinRT category that is a pain to use. E.g. requires "identity".
https://learn.microsoft.com/en-us/windows/apps/desktop/moder...
today at 5:32 PM
The most supported UI framework from Microsoft for Windows is .NET MAUI, released in 2022.
today at 6:18 PM
It's not quite that settled yet. Lots of third party libraries are getting abandoned pretty quickly. We are also stuck in purgatory due to a VSTO add-in which we can't rewrite in Office.js because it's nowhere near as powerful and there is no usable .Net Core integration.
today at 4:36 PM
I used to be in the anti .NET camp but for Enterprise applications .NET has been delightful and upgrading between versions has been trivial since .NET 6.
today at 7:13 PM
> upgrading between versions has been trivial since .NET 6
I don't understand this. I love C# as a language, but the idea that everyone is forced to upgrade every single library every 2 years is beyond unreasonable. I've recently run into a situation where some configuration behavior deep in ASP.NET just randomly changed between .NET 8 and .NET 10, and broke things. A runtime exception, not a compile-time one. What else is subtly broken in a DLL referenced by a reference of a reference? You may never know.
It seems like every software engineer has forgotten that code is ultimately math. All of the greats understood this, from the times of Alan Turing through the 1990s. Then all of a sudden it became some sort of a toy, a game for 20-year-olds to one-up each other, to move fast and break things. Somehow, as software development matured, it became less of a serious discipline. Now, we have an entire ecosystem where it's just accepted that any line of code you've ever written may just randomly break every 2 years. You can no longer just write a small C# library and expect to to continue to work forever. It's completely unacceptable in my view.
today at 5:10 PM
> Silverlight hit us hard.
So about that... when I started using .NET Core, it was bleeding edge and shiny, but it was to rewrite a Silverlight application, the only way to run it was Firefox ESR 54 (I think it was 54?) we thought "oh this is easy" only to find out that it was a colossal nightmare, EF didn't support some things, including Shapefiles, we had to force the support, get back a WKT from SQL Server, and send one back when we did updates, basically Entity Framework was mostly useless.
I also reverse scaffolded the existing database into EF compatible classes, so that saved us a ton of effort, the original idea was to be backwards compatible.
Best part? It never got used, when we found out they weren't using the project, my boss said, screw it, lets remake it in Python I know we can build it up quicker, and sure enough, it was done quicker than the C# rewrite, fastest rewrite I ever did, because he figured out the key things they actually needed, and trimmed the requirements to just those key things.
Sometimes rewrites don't need to be 1:1 especially if it's an internal only tool.
Anyway, lots of blood, sweat, tears, stress, etc from me as it was my first job in tech.
today at 4:56 PM
We've always stuck to the server-side web technologies: WebForms (now very much obsolete), MVC, Web API + Angular, Blazor.
today at 4:32 PM
Never chase the shiniest toy at work! Give it a year or two to settle down first.
today at 6:20 PM
I think you underestimate how long we left it to start with.
today at 4:31 PM
WCF, Silverlight... They were terrible. New .net stack is much better.
today at 5:19 PM
WCF had some good ideas and if you followed its best practices migrating to other networking stacks should have been straightforward. Delete your complex Bindings XML configs and just write implementations of your Contract interfaces and figure out where you need to dependency inject the new implementations. If you were deep enough into WCF you might have even had mock implementations you could bootstrap from.
That concept of "define an RPC interface and let the platform fill it" was ancient before WCF and continues to this day (I've got a bunch of tabs open for Cloudflare Durable Objects right now). The legacy problem with WCF was not but long term backward compatibility with the Bindings abstractions and XML config junk and SOAP Stack RPC. Without WCF native in the box in .NET 5+ it is harder (but not impossible) to build SOAP RPC monoliths. But so few businesses actually need SOAP and can't just switch to something more RESTful.
Silverlight was maybe ahead of its time given the growing popularity of Blazor WASM.
today at 5:57 PM
> Silverlight.
Boy do I have good news for you (I am half kidding). OpenSilver project lets you compile your Silverlight code to Web Assembly or iOS/Android/Windows/Mac.
You can dust off your code now.
https://github.com/OpenSilver/OpenSilver
today at 4:40 PM
>I wish Microsoft would make a cross-platform GUI stack
I know this may as well be beseeched of the King of Nod, but if they just made WPF cross platform I'd be delighted beyond words.
today at 5:11 PM
I'd be okay with this, I'd also be okay with MAUI (which... is a whole behemoth of how to make UIs), but anything, hell I would take them porting WinForms somehow into a cross-platform .NET thing that runs everywhere (Mono used to!), I just want anything at this point.
XAML was supposed to become an open spec, not sure why they have so many teams making so many different ways to make UIs. Then there's those "Windows 8" apps, or the "Universal Windows Apps" which are poorly named, you cannot run any of them on previous versions of Windows, if anything Win32 apps are universal apps.
today at 4:44 PM
Isn't this exactly the point of Avalonia?
today at 5:08 PM
Yes and Avalonia was created by not-Microsoft as Microsoft has repeatedly proven itself incapable of doing it.
today at 5:10 PM
VB6 clone written in Rust:
https://github.com/iBobX/RapidR
Not compatible, but you can use it in a cross-platform way.
today at 4:35 PM
> I wish Microsoft would make a cross-platform GUI stack they'd actually stick to
Sinofsky left a lasting damage. The way things go with LLMs, it might be now pure app-specific Rust instead of any common frameworks....
today at 5:10 PM
https://learn.microsoft.com/en-us/dotnet/core/compatibility/...
Be aware
(decimal)0.1
// .NET 10: 0.1
// .NET 11: 0.1000000000000000055511151231today at 5:23 PM
Thanks for that heads up, also:
(decimal)0.1f
0.100000001490116119384765625today at 7:22 PM
Wow, what a vicious footgun. I do a lot of work related to pricing, invoicing, exchange rates etc... where decimals do matter, so this kind of stresses me out. Thanks for the heads up.
today at 7:41 PM
You shouldn’t cast from double to decimal anyway. Just use decimal throughout.
today at 7:55 PM
Yeah, while I do agree this change in behavior could be a major footgun for a few people, the vast majority of use cases for decimal should be unaffected. Decimal is usually used for business applications to the exclusion of double/float, not alongside them (the only doubles I've seen in my company's applications are just storing Lat/Long values for addresses, everything else uses decimal values).
Hell, on the rare occasion I've needed to use a decimal literal in code, I just use a decimal literal: don't do (decimal)0.1, just use 0.1m.
today at 8:01 PM
Exactly. If you want an actual decimal literal you should always write it as 1.2345m rather than converting it with the (decimal) prefix. The compiler refuses to accept a plain 1.2345 assigned to a decimal type so it’s not such a footgun as it may appear to some; though it might offer a better DX if the type of the literal was inferred from the thing you are trying to assign it to, that would no doubt have unintended consequences.
today at 2:13 PM
Awesome stuff, as usual in .NET releases! The runtime async support is surprising to me... I get that stack traces are more friendly without state machines in the code, but I didn't expect they could switch to a runtime version and have equivalent performance. Cool!
today at 9:15 PM
Friendlier stack traces have an important upside: Cancellation is done by throwing exceptions, but you can't stop it from storing the full stack trace in a string. This can result in a noticeable hitch if the stack is very deep and complex. With a friendlier stack trace, I hope this issue is finally gone.
today at 2:29 PM
They did that after the learnings with experiment of having a Erlang, Go, Java like implementation for async, await.
today at 3:15 PM
Is this a joke, like "hen's teeth"? Those languages don't have async/await.
today at 3:50 PM
They don't need to mark functions as needing special compilation / rewriting to support asynchronous use since the runtime provides the necessary mechanism to pause and resume (e.g. stackful continuations). So 'async' is unnecessary and 'await' can be a library function (or remain a keyword of course).
today at 3:33 PM
No joke at all, not that I said "like implementation", not that they have async/await.
https://github.com/dotnet/runtimelab/issues/2398
Be sure to read all the way to the bottom,
today at 5:30 PM
Seems to be a gift of the modern CLR cycling faster with each release since .NET 5 versus the backwards compatibilities the runtime used to have to keep in .NET <= 4 because the same runtime would power multiple versions at a time of the compilers and BCL above it. In the old world "Framework" days after the CLR made the huge changes from 1.0 to 2.0 (for generics and other things) it seemed almost "afraid" to add features, especially for things that compilers could support themselves. The .NET 5+ CLR makes fewer "forward compatibility" guarantees as it now assumes it cycles directly with the compilers and that most deployment methods "bake" the CLR version with the application.
There have been smaller runtime improvements like this one in recent releases, but this one seems a big one to point to to prove that the runtime is truly growing at a faster and more interesting pace than the old days.
today at 6:06 PM
What IDE are people using for C# these days? Back in the day (circa 2008) Visual Studio was your only real option.
But I'm on linux now. If I wanted to pick up back up then what should I be using?
today at 7:54 PM
I switched to Mac and tried Rider for a while but ended up very happy doing everything in VS Code. I guess in the pre Agent days I would have put more effort into learning Rider but in modern times I just don’t need so many fancy tools for manual refactoring etc. But if you’re still very hands on, it’s an excellent - and possibly superior - Visual Studio replacement.
today at 6:20 PM
VSCode is mostly good enough and if you don't mind forking out for a license, Jetbrains Rider is better overall
today at 6:35 PM
Claude and codex ? Half joking, but can't remember last time I bothered booting up Rider - probably over 6 months - will unsubscribe next year. VSCode is good enough for exploring changes, LLMs are driving the CLI tooling anyway.
today at 6:17 PM
JetBrains Rider is top-tier.
today at 8:50 PM
I've been using Rider for about a decade, but honestly these days I just use Zed. I'm surprised how fast roslyn-language-server is now.
today at 10:11 PM
IDE? In the AI era?
today at 8:09 PM
Visual Studio / Rider / VS Code, depends how much RAM and CPU device has.
I prefer VS but for the performance, I use VS Code. Unable to work in Rider on 32 GB RAM.
today at 8:09 PM
Rider if you want a regular IDE. VS Code if reading code from an AI is enough.
today at 6:40 PM
I’d go with Rider myself. Vscode?
today at 6:13 PM
Surely VS Code?
today at 5:06 PM
> .NET 11 is a Standard Term Support (STS) release with two years of support under the .NET support policy. STS and Long Term Support (LTS) releases have the same quality and servicing policies; the difference is the length of support. .NET 10 remains a supported LTS option. Stay current with servicing updates whichever release you choose.
They burried the lede on this one.
- .NET 10 is end of life'd on Nov 14th, 2028 - .NET 11 is end of life'd on Nov 9th, 2028
Which means anyone who upgrades to 11 has fewer days of support. But the idea is that you'll upgrade to 12 as soon as it's RTM'd.
The problem with that is I have clients who refuse to budge from 2.1 (or even upgrade to the 2.1 re-release: 2.3).
today at 5:11 PM
Nothing new to comment on here. MS has been open about its long term and short term releases based on even and odd numbers. Unsurprisingly, there's a web page that explains it https://dotnet.microsoft.com/en-us/platform/support/policy
today at 5:24 PM
This whole thread is a bit of a tangent, but I'm not sure the STS/LTS distinction makes much sense for the end user. STS gets a relatively generous 2 years of support (cf. Java at 6 months) while LTS gets a relatively anemic 3 years of support (cf. Java at 8+ years [1]). Plus, with yearly releases, every LTS expires at the exact same time as its successor STS. Are STS releases generally less stable or more experimental than LTS releases? Otherwise, this feels like a valueless distinction.
[1]: https://www.oracle.com/java/technologies/java-se-support-roa...
today at 9:16 PM
Supporting anything for 8 years is a huge commitment. That means 8 concurrent versions needing at least some work all the time.
I think Microsoft regretted those long (10 years) support guarantees they had for .net framework. In the end, it probably slows you down in terms of innovating
today at 5:34 PM
Yes, STS releases tend to be more experimental and the LTS shores up the gaps.
today at 5:46 PM
> The problem with that is I have clients who refuse to budge from 2.1 (or even upgrade to the 2.1 re-release: 2.3).
2.3 on .NET Framework reaches end of life on 2027-04-13, so upgrades have to happen soon.
today at 8:44 PM
If you mean ASP.NET Core 2.1 or ASP.NET Core 2.3, then they are actually .NET Framework despite the name, and are not the same lineage as .NET 10 and .NET 11.
It's still in support, and led to one of the most confusing release announcements in history: https://devblogs.microsoft.com/dotnet/servicing-release-advi...
ASP.NET Core 2.3 does go end of life in April 2027 however: https://devblogs.microsoft.com/dotnet/aspnet-core-2-3-end-of...
today at 5:11 PM
>They burried the lede on this one.
Not really for any groups that commonly use .Net. Even numbers are LTS and odd numbers are not, or at least it's been this way for a while.
>I have clients who refuse to budge from 2.1
Give me their names, sounds like it would be easy to penetrate them again and again and again with AI.
today at 5:21 PM
Plus the even/odd LTS/short-term cycle matches Node.js and some others as a release pattern.
today at 5:18 PM
I feel like they've done a lot to resolve a lot of the confusing versioning and runtime compatibility problems that drove away non-Windows developers from using .NET.
However they still suffer from an incredible branding problem.
It's really too bad because as a runtime it's fairly incredible, and F# in particular is fun to work with. They also have good cross-platform support and tooling, which they didn't have a number of years ago.
Combine that with the fact that nobody really writes code anyway thanks to AI, and this could be a slam dunk runtime for an AI-native development environment.
But it will never happen.
today at 5:29 PM
> ...this could be a slam dunk runtime for an AI-native development environment.
- Roslyn analyzers (C# code analyzers written in C# that run on the C# compiler)
- Simple, stable tool chain
- Language has largely been stable and backwards compatible for the two and half decades of its existence
- Now have a very, very good type system that promotes good practices
- Very good corpus of docs
It's actually crazy what the agent can do with that.
today at 6:40 PM
I am actually quite convinced that having solid framework as a base will be a huge win for building business line applications with use of LLM agents.
today at 8:18 PM
I wonder what this means for custom async implementations in the long run. I was considering using my own AsyncBuilder to bring over a rudimentary "computation expression" capability. But that might be out of question now.
today at 2:06 PM
For the features that overlap between golang and .Net, can someone please let me know why people choose .Net over golang for non legacy systems? As far as my limited knowledge goes, go is much more efficient in syntax and memory.
today at 3:38 PM
The languages are very different. C# has much more powerful abstractions and is a much larger language in general. That is both an advantage and a disadvantage. Modern C# is also very fast, I wouldn't bet on Go winning that comparison.
My experience with Go is a while back, before it got generics. While I liked the overall idea of a language with little syntax and a focus on simplicity, I didn't really like Go in the end. It was a weird mix of too low-level and high level code. It exposed low level details that I don't need to think about in C#, and it is missing convenient abstractions like LINQ.
Personally, I like async/await in C# more than Go concurrency. I know this is heretical here, but I don't care about function coloring or any of the abstract advantages of other concurrency approaches. Async/Await in C# works for me (in non-legacy code) and is easy to understand.
today at 4:16 PM
the main problem with the .net async/await is the immediate execution of a call returning a Task, without control over where it may execute. F#'s asnyc was way saner and easier to understand in my opinion.
I'd beg to differ on the easy to understand... last time i worked with this stuff like ConfigureAwait only has cargo cults and no real good explanation, and wasn't covered in any recent version of eg. CLR via C#... Distributed tracing and thread management were totally bitten by async, also the way iterators and LINQ were broken... that was when I got off the train, as it broke the best abstractions for line of business code so that instead of designing an IO pipeline lazy developers could simple put Task.Run around blocking IO calls at the bottom of the codebas, and ... yeah, we are back at function coloring.
This brings me to: I winder for what kind of apps did you find it a delight, because for extensive data wrangling leveraging multiple calls it was a pain.
today at 4:21 PM
You don't need ConfigureAwait() at all in modern C# (unless you're writing desktop UI code, I don't know enough about that to comment there). In a modern ASP.NET Core web application you don't care about which thread your Task executes, there is no synchronization context at all.
I'm using C# for the backend of a web application, not legacy but .NET Core from the start.
today at 5:12 PM
> unless you're writing desktop UI code
Indeed, you need a crapload of ConfigureAwait() in desktop UI code, and it infects all underlying libraries that might have to run inside a desktop app. The worst thing they ever added to our beautiful language. It's a dream to work in ASP.NET without needing to pepper every await with a ConfigureAwait().
That said, I don't agree with the characterization of ConfigureAwait() as un-understandable. We understand it. We just hate it.
today at 6:46 PM
In case you were not aware, you can use Fody to automatically add ConfigureAwait everywhere it's needed at build time: https://github.com/fody/configureawait
today at 4:37 PM
Sometimes you do care where it executes, and then it starts to become a problem ;).
Something which was as easy as an IO call in the middle of a function running on a thread pool becomes a PITA if the IO call is cleverly async-ized (by wrapping the original sync IO method into a Task.Run(), but actually this is the better case!) ad then removed from the public API... (cannot recall the exact library, but it was a pain, as it was some core stuff, maybe vendored by MS themselves)
Also back then it caused us loosing correlation IDs as those were thread locals in traditional pre-async ASP.Net Core (think version 0.9-2.0 when I worked on that part, then up until core 4 and the async mania continuos broke readable calculation code in LINQ, TPL based performant data pipelines, etc...) and with async execution started to migrate between threads... If you needed high performance you already used threads, because it was not a classic web application, but a specialized optimization algorithm with a web api... and if you need more than CRUD, and use all the cores (and SIMD) to squeeze out all performance you have, you hit these problems right away. These all could be solved, but it actually broke stuff, a lot of stuff, as developers jumped on the fad, and it their sync code was turned into a .ConfigureAwait(false /* Microsoft blogs talk a lot about it without a clear explanation, so every cool kid does this */).GetAwaiter().GetResult()... then you some of your threaded code silently broken if you dare used Thread Locals... Faded memories, but that was when they f'ed up the C# language.
today at 4:41 PM
That's why I emphasized modern .NET. All I/O should be using async methods through the entire application. Wrapping sync code in async causes troubles, but that is entirely avoidable in new code. That's the price you have to pay to make this particular abstraction work well, everything I/O must be async.
today at 3:17 PM
Go is not much more efficient, compared to NativeAOT - though it can be faster in some areas. The big reason why someone would want to use C# is that it's a more expressive language. You can even write your DB queries with LINQ (C# lets you express code as data), making them type-safe.
today at 9:52 PM
For me, I use both languages. But there's a lot of things which are only like a couple lines or very easy to do in .NET but to do it in Go is a lot of work.
Some example.
Open API document generation using type handlers doing this in go was a pain in .net is comes out of the box.
A lot of people underestimate the power of extension methods.
.nets metadata system is compiled time safe where Go is not and this has burnt me a lot.
I really like go for command line tools Small binary sizes I lovely.
go's standard library is significantly more hackable than .nets there have been a few things where I've ended up using Go because to do it in .net would require me maintaining a fork of the standard library.
Go makes composing data insanely simple. where in .NET you can have an equivalency but it will be done via interfaces.
Go's Dark Type System is both great and really annoying.
C#'s stricter type system when it comes to interfaces can be really handy as that can be used as metadata to change how data is processed and enforce compile time checks.
today at 2:15 PM
Go isn't as much fun. And is usually slower for my work until I bump up the GC and then it uses roughly the same amount of memory.
I have also found c# async to be lighter than go's concurrency story, and dotnet11 is going to make that much more clear.
Now: I write software for fun, not for money.
today at 2:20 PM
So for a requirement where both golang and .net would suffice, would you choose .Net over golang?
today at 3:17 PM
Yes. Unless I really need CSP, but then I would much rather use f#+hopac because CML is the bee's knees.
For general concurrency where async doesn't suffice, go is nicer than C#, but hopac is nicer than go.
If go had proper concurrentML I am pretty sure I would use go a lot more. Again: this is for hobby projects where I get to choose freely.
today at 5:01 PM
Based Hopac enjoyer. I built the Spiral's language server using it. I have to admit though that these days I just use native .NET channels for all my message passing needs.
today at 3:41 PM
Modern .net is surprisingly performant while also being very high level if needed. So it is very close to being a "learn once" ecosystem. This means you use it to write code for a much broader range of usecases than others, you will have a broader / more mature ecosystem than others, and your potential hiring market is more vast.
today at 2:29 PM
I know it's not quite as important in the agentic world, but the .NET tooling (across Visual Studio, JetBrains Rider, and C# Dev Kit for VS Code) is just awesome. Some people might frown upon full-blown IDEs, but sometimes they're just incredibly handy.
today at 2:57 PM
Even the non-visual stuff is pretty awesome, claude is able to trace and profile .NET code without skills or MCPs.
today at 2:27 PM
That is an urban myth, CLR was designed with support for C++ in mind.
Also Go is at the stage of C# .NET 2.0, and still plenty of lacking language features.
I really think the creator of C# made a disservice to the .NET community by picking Go for the Typescript rewrite.
today at 4:43 PM
Go still has a better deployment story. Single static binary deployment are king. On .Net 8 I was hitting random compatibility issues with single binary application bundles, with System.Data.SQLite, and other libraries. (had to use that one and not he Microsoft one for some legacy compatibility reason, maybe around Guid handling maybe)
today at 6:20 PM
From your description you certainly weren't using AOT compilation, and forgetting about CGO.
Static linking as the only way in most platforms, we moved into other mechanisms for a reason.
In Go the only way to do plugins, without shipping everything as source code is to use microservices, the more fashionable name for multiple process with OS IPC.
today at 5:54 PM
Is cp -r/xcopy/robocopy a folder (with or without tar/zip wrappers) that much harder than single file deployment?
today at 7:08 PM
> still plenty of lacking language features
Fans of Go (myself included) prefer this. The small feature set is a feature, not a bug.
today at 7:27 PM
The history of generics and Go is enough to keep me away from that mess of a language forever.
Go seems actively hostile to it's users, especially in comparison to C#.
today at 9:58 PM
with Go, they have made one commitment. The language will always be backwards compatible. So code written in today's go should still be able to be compiled in 30 years using the compiler of the day. That means when adding new language features, they are very, very, very particular and stubborn about it.
today at 3:30 PM
The tooling and development UX.
Put things like their remote debugging experiences side-by-side and you will quickly learn why some of us go for the Microsoft tools rather than the others.
today at 4:43 PM
Exactly C# with .NET is batteries included type of ecosystem where most of stuff is done by Microsoft and if someone is experienced with the stack he can whip up enterprise ready app in no time.
I don’t have much experience with Go lang but what I have seen, there is lots of things one has to implement on his own. Which has its own merits, but to go from 0 to full fledged enterprise app you need more time.
today at 10:00 PM
Go is also very much a batteries included language when it comes to their standard library. Like it even comes with an ORM. Most people use the Http server that comes with standard library. even more so now that the router for it is a lot better.
today at 4:48 PM
I'm not a .NET person by any stretch and have been more of a Go person traditionally, but I think from what I've seen IRL, this is accurate.
Go is trying to be like C. Less language features (until recently...), so to solve problems you write more code. Intentionally simple language. Intentionally need a lot of code to implement something substantial. .NET seems more frameworks driven. Go traditionally has eschewed frameworks for doing stuff much like C tends to eschew framework type stuff and prefers smallest dependency reliance as possible.
today at 5:36 PM
Every new release of Go brings some of the features that have been available for twenty years in .NET (and every other language), like generics. Some features, like ergonomic error handling (whether or not you think unchecked exceptions are a good idea), Go has said will never be available. There isn't much to recommend about Go. It's fast to do the mediocre thing, faster than any language, but for a serious code-base intended to be maintained long-term, C# has better libraries, established patterns and frameworks, abstraction features, developer ergonomics, etc.
today at 3:55 PM
Compare the ecosystems of .NET vs golang. No comparison whatsover. IoT, mobile, web (front and backend), cloud, etc. I think Go lacks variety and is more focused on traditional greybeard areas.
today at 2:18 PM
Native interop is much better in .NET, that matters a lot for some systems.
today at 5:15 PM
There's one (usually minor) caveat here, which is around strings. In .NET, like Windows*, strings are UTF-16. If you call into native code that uses UTF-8 strings extensively, then they have to be marshalled (when passed to native code) and unmarshalled (when received from native code), which takes additional time and memory. I've run into this lately, which has led me to write special CStr/CString types (idea borrowed from Rust) to avoid unnecessary, especially round-trip, (un-)marshalling.
* = Technically, Windows now supports the UTF-8 "code page" where applications can use UTF-8 natively, but all NT system calls use UTF-16, so the conversion still has to happen somewhere
today at 6:01 PM
C# in .NET 11 adds support for UTF-8 string literals (`ReadOnlySpan<byte> str = "Some sample string."u8;`) which is a step to making such interop easier.
The Runtime has explored attempting "UTF-8 native" string storage as a platform feature especially for Linux FFI, but yeah Windows using UTF-16 is still a compatibility issue where platforms themselves just don't agree.
today at 4:49 PM
> go is much more efficient in syntax and memory
I'm going to talk about the second part first. .NET is often deployed with a JIT compiler runtime like Java. That means that hello-world scripts use a lot more memory than Go, but that difference becomes a lot smaller on real-world programs and .NET can use less memory than Go depending on what you're doing.
C#'s concurrency can also be more efficient than Go's since it's coroutine based async/await is a lightweight state machine rather than green threads. That's not to say Go is especially heavy or anything, but it is more overhead.
In terms of "efficient in syntax", Go has a lot fewer features. That's one of Go's biggest strengths and weaknesses.
Here's an example. With Go, you just have struct tags: strings of anything that can be attached to a field in a struct. Let's say you have a User struct and you want to make it so the JSON won't have the user's email in it. You do: Email string `json:"-"`. But what if you mess up and write "jsom" or `json:-` without the quotes? It's basically a meaningless string to the compiler. With C#/.NET, you have annotatons/attributes that are a part of the language and checked as types. You'd do: [JsonIgnore] public string Email. JsonIgnore is an attribute class so your IDE can help you.
There's a lot in Go where the answer kinda becomes "f-it, we'll just use strings instead of having something meaningful." Go talks about how important error handling is, but then you get an error from a function and it's basically just a string. If you're doing good error handling, you want to handle different errors differently, but that's not easy in Go. Frankly, it's not easy in most languages, but this is definitely an area where it felt like Go gave up on creating something good in order to get something out the door quickly enough that Google didn't cancel the project. In a language like Zig, errors are enums and you get exhaustiveness checking. With Go, someone could add another error state to something you're using and you'd never know - or they could change the phrasing of an existing error and you'd silently stop matching on it.
Go's lambda syntax is clunky. func(x MyType) string { x.ToString(); } vs (x) => x.ToString()
There's lots of stuff like that. I'm not arguing that Go is bad, but it has a ton less stuff and that's both a blessing and a curse.
today at 6:39 PM
Modern C# is a much more powerful and expressive language than Go: iterators, collection expressions, records, async await, source generator, much more powerful generics, low level memory management features like Span<T> and ref struct, etc. It is closer to C++ with opt out GC/opt in AoT than it is the old Java comparison days.
Go's advantage is mostly the fast compile speed, goroutines, profiling, and a very opinionated formatter. Neither has an overwhelming advantage over the other though.
today at 3:07 PM
What's go's equivalent of LINQ?
today at 6:54 PM
SQL strings
today at 3:15 PM
At former jobs, because people know it and switching from .Net <-> Go generally isn't worth the time to learn other language.
Sure, having a runtime is downside but difference in container size for your average CRUD App doesn't really matter. Performance IME is very similar.
Ops type here.
today at 6:29 PM
golang also has a runtime
today at 9:25 PM
Technically yes but since it ships with binary, effectively, most people don't care.
today at 5:45 PM
Well for me, that's like asking why don't you write everything in Erlang rather than C#.
Go is for me just a, hmm, concurrency programming language. Its popularity stems just from being pushed by Google, and the cosmetics of using braces. Despite what pundits say, it's completely and utterly not like C.
While C# is a successful and comfortable everything-language. It is blazing fast, has amazing tooling, you can get going to prototyping, then enterprise code, its serverside handling is excellent. It has many low level features and built-in data structures. It's I would say preferable for almost any kind of work except scripting. (I really don't like C# scripting.)
today at 6:10 PM
> except scripting
PowerShell says hello :)
today at 5:18 PM
.net has a much stronger ecosystem and more experienced developers.
Golang is a fantastic delightful language, but it's still missing a lot of stuff that's existed in C# for years.
However, do you *need* all those .net tricks?
Golang is usually enough for a small project you need to build in a weekend. Scaling it to a mono repo with 200 developers might be a bit difficult
today at 10:04 PM
One thing I really, really wish Go had was met a daughter that was checked at compile time.
me miss spelling a tag has burnt me so many times.
today at 4:12 PM
In addition to what the others have said, I also enjoy the extremely robust ringtone reflection system in .NET. I don't know of any other systems that provide more complete detail. Maybe Java, though it's been a while since I've used it. It makes reflection in other languages feel prehistoricaly primitive.
today at 6:37 PM
My first job out of college was at Microsoft, probably at the height of its powers. I was a developer on the .NET team and helped bring SQLCLR to life, the regrettable decision to let you run .NET code inside SQL Server. That job opened a lot of interesting doors for me. The weather sucked though.
today at 3:51 PM
To all the .NET outsiders, C# is indeed the .NET darling but you can also use F# if you feel like living in the future. Or VB if you feel nostalgic. So in .Net land, C# is not destiny. It is a choice
today at 4:18 PM
I feel like VB is a "better" Python for beginners and F# is a much better Python for data scientists.
The tooling ecosystem in Python is just so bad that I can barely stomach it after spending the last 20 years in .NET. At times, when supporting analysts at work wanting to write a little glue code, I've found their main complaint about C# is "curly braces" (which seems like a really odd thing to complain about, but ok) and I've found it easier to teach them a little VB than to figure out whatever the hell is going on with dependency and runtime version management in Python.
And then, for the folks who go deep on Python for ML tasks, I've mostly found them caring now about specific libraries than the language itself. I then show them how .NET has equivalent or better libraries available and that a lot of the code-golf stuff they do in Python has a much cleaner version in F#.
today at 7:36 PM
Ugh, I agree, Python tooling makes me want to pull my hair out every now and then. And then I weep for IronPython... it could have been a great blend, if it hadn't been abandoned. Though I can imagine that trying to "marry" Python to .NET became more and more of an uphill battle over the years. I think the last maintenance release was about two years ago.
today at 5:32 PM
What do you recommend for numpy, duckdb, pandas/polars and pytorch in python?
today at 6:04 PM
> C# is indeed the .NET darling but you can also use F# if you feel like living in the future. Or VB if you feel nostalgic
Or PowerShell if you want to hash out something really quick.
today at 2:01 PM
I am looking forward to async stack traces being more legible. It might encourage developers to actually read and use them.
today at 2:17 PM
Does runtime async mean stackful coroutines? It's not obvious what exactly the state machines are replaced by, if not automatically inserted yield() calls.
today at 2:40 PM
I took it to mean that coroutines are still stackless but that yielding is now an operation the runtime knows about instead of being compiled out by the compiler.
today at 7:32 PM
They have a section for breaking changes in .NET 11, containing a few dozen items. It starts with "If you're migrating an app to .NET 11", as if you'd ever have a choice. Microsoft forces you to upgrade every 2 years, and they ship breaking changes. It is completely unacceptable that these two coexist, and I'm confused why more people aren't angry about it.
Either you should not be forced to upgrade, or backward compatibility should be guaranteed -- not both. But Microsoft is happy to say "hey, we're going to completely abandon support for every version of .NET every 2 years, and if you can't keep up to date then everything that goes wrong is all your fault." Meanwhile, they have no problem breaking your code.
This essentially kills the possibility of just writing a bunch of small, useful libraries, and having them be useful forever. It is the antithesis of a healthy software ecosystem. Are you really expecting that every single open source library that is referenced by a reference of your references is going to be maintained in the face of breaking changes every 2 years? If not, then upgrading your version of .NET is just a gamble that things won't randomly break. But you must upgrade, or you're a bad developer who's leaving your apps vulnerable and every automated security checking system in the world will yell at you about how awful you are. You must upgrade everything you have ever written every two years, and hope that every library you've ever depended on does the same, or you are putting yourself at risk of some breaking change sneaking its way in, or leaving yourself vulnerable to a critical security issue that Microsoft refuses to back-apply.
I cannot understand why I'm the only one shouting from the rooftops about this. It's completely unacceptable to me.
today at 7:50 PM
Plenty of systems still use .NET 8/9 today. It's very likely that there are some websites that we use daily running on .NET 8 with ASP. Microsoft never barged in our server room and forced us to upgrade .NET, but YMMV.
today at 10:07 PM
Godot still defaults to .net 8
today at 7:20 PM
No mention of Avalonia. I remember it was chosen as a back-end for MAUI awhile ago.
today at 6:05 PM
C# is really good language with world class tooling
but job offers are kind of boring :/
today at 8:25 PM
I moved from PHP to C# jobs because thought it will be more fun.
Have to agree, PHP codebases also were nice organized, bigger repos, mostly web apps / company backoffice apps and so on. Personally, products were more interesting.
With today dotnet state, it really suits well for API development / microservices / integrations / we had many workers - jobs running, queue subscribers, K8S, good test coverage, open telemetry.
Even migrating from .NET Framework wasn't so big pain.
Job postings sucks nowadays for all stacks.
today at 6:34 PM
Must depend where you live. In the Midwest, like Wisconsin, there are plenty of jobs for .NET
today at 6:35 PM
Their may indeed be plenty of .NET jobs, but theyre likely very boring and ripe for automation.
today at 5:10 PM
C# is a really great language, and I hope I can keep using it. I like Microsoft’s ecosystem, but personally, I’d like to see its cross-platform support strengthened a bit more.
today at 6:40 PM
What's missing from cross-platform support? You can basically run it everywhere already?
today at 7:39 PM
MAUI is still not great, is it? I remember Linux wasn't supported either. I tried MAUI on macOS, but honestly it's hard to consider that supported. Many people use WinForms and WPF, but these don't support cross-platform, and honestly WinUI 3 is also lacking. On the server side, TypeScript is actually better (both are Microsoft languages).
In fact, desktop apps are mostly used for factories or Windows-only programs, which is a bit disappointing. It's a very good language for making games and the most widely used language for factory programming, but whether to build a product with it is a bit ambiguous. I'd use it when making a program for a specific customer, but for a product I'd probably use TypeScript.
today at 9:47 PM
If you're limiting yourself to Microsoft's options then I'm going to have to say that's a you problem. Avalonia is pretty great. What do you use when you're building apps in other languages?
> On the server side, TypeScript is actually better
I like TypeScript but I'd never pick it over C# on the server side unless I had to do React SSR for some reason.
today at 4:33 PM
C# finally has Sum types.
Not quite discriminated unions, but close. With an exhaustive checker with switch statements.
Now if only it made creating unique types easy, so I wouldn’t have to use a library like VOgen, the language would be nearly as good as F#
today at 7:47 PM
Hey there. C# language designer here. Working on that here: https://github.com/dotnet/csharplang/tree/main/proposals/dis...
You can start reading through it starting at: https://github.com/dotnet/csharplang/blob/main/proposals/dis...
Definitely want to move to get this in asap :)