👾

New places (no content yet): • @ashmind" target="_blank" rel="nofollow noopener">mastodon.nz/@ashmind • cohost.org/ashmind

@ereliuer_eteer I would say, don't focus your best skills on the goal of breaking it (just in case). But also don't be cautious with anything you try as part of normal usage, SL is expected to handle it.

@ashmind Am I encouraged to experiment? 😉 Or should I be cautious but report issues if something breaks down inadvertently?

@ashmind I carelessly pasted a small fragment of C# code with explosively recursive generic types to sharplab.io and I think the service is down now. I am really sorry. 😞

@ereliuer_eteer There are some built-in protections against type explosions btw: github.com/ashmind/SharpL… If you find something that still causes a problem, please let me know and I'll extend those further, the system is designed to be resilient against user input.

@ereliuer_eteer Thanks for reporting! However I don't think it's caused by your code — there is a nightly build and it seems latest main didn't build correctly, so likely just a timing coincidence.

@terrajobst I wonder where the tech people will go — keen to follow MS .NET ppl on whatever is the new service (hopefully not LinkedIn though).

@davidfowl Fair point. Even if it is an explicit opt-in, I would still pick global opt-in over per-endpoint config. But the question of opt-in vs opt-out and the startup cost is very interesting design problem, I'll think about it.

@ashmind The cost needs to be zero if you’re not using it. This filter does a pretty good job of that but there’s a startup cost. Source generators might be a solution but still needs to be opted into.

Which style do you prefer and why? #dotnet #aspnetcore #minimalapis

@davidfowl Yep looks like the best option, tbh could be a sensible framework default, without any call? (though could be a breaking change, not up to date on latest ASP .NET). Allowing individual endpoint to process its validation errors in a custom way feels like an edge case.

@ashmind This?

@PeledZohar @jaredpar @terrajobst We used it historically before LINQ expression trees existed. E.g. Where(Client.Age > 30) where > generated some kind of BooleanQueryExpression which was then converted into SQL. Not sure if there is much value in modern uses though.

@jaredpar @terrajobst Ok Yes, but the question is - can you think of a good reason why you might need it someday?

@KirillOsenkov No need for Roslyn fork, can just use IgnoresAccessChecksToGenerator: #L16-L18" target="_blank" rel="nofollow noopener">github.com/ashmind/mirror…

or you can have a fork of Roslyn that supports IgnoresAccessChecksTo so you can avoid having a fork of Roslyn with InternalsVisibleTo your project

@chriscoyier I might be a bit biased, being originally from Russia. But people do not choose what country to be born to. If the author of a library or technology personally chooses to promote heinous views, that can be a reason to drop the tech. Their birth country, not so much.

What is the fallout of Russia and web tech? Are there good reads on this? For instance BEM is a CSS methodology that came from Russia right? If you’re pro-Ukraine are you anti BEM or is that entirely unfair?

@davidfowl Great question, I had to think about it. I think still not OK, due to encapsulation. public IX Dependency { get; init; } — get breaks encapsulation public IX Dependency { private get; init; } — technically doesn't, but very weird/awkward pattern

Do C# 11 required properties change the game for dependency injection? Is property injection OK now? #dotnet #csharp

This works for method visits as well, e.g. here is a method that gets called two times. You can see what called it that time, and what return value was produced for that specific call.

Recently I was thinking about "time-travel" alternatives to traditional debugging for SharpLab's continuous Run. Here's one outcome. SharpLab used to dump all loop values at once, but that wasn't too useful. Now you can select a specific iteration and see values and flow within.

@sonergonul Thanks for raising! This is now resolved in newer branches (including main).

@ashmind Hi. Most of branches in sharplab.io I get this error. Do you know why? I just try to test it for UTF-8 string literals and unsigned right shift operator but no luck..

@davidfowl Makes sense — e.g. we have an app where migrating WebForms part is multiple years with no customer value. I've considered same, so we could reap immediate tech benefits (C# 10, Mac dev, Linux host), and leave fig rewrite (also simpler, no YARP) to future UX-oriented initiatives.

Someone did a webforms port to ASP.NET Core! #comment-15756" target="_blank" rel="nofollow noopener">devblogs.microsoft.com/dotnet/increme… #dotnet #aspnetcore

@davidfowl Yep hoping it evolves, last proposal I saw were sizes only: #heading=h.jhslnvomzqt1" target="_blank" rel="nofollow noopener">docs.google.com/document/d/1ck…

@ashmind For now. I remember the same when websockets first came out

One of our interns is also adding WebTransport support to Kestrel. github.com/dotnet/aspnetc… web.dev/webtransport/ Hoping to be one of the first frameworks to support this! #dotnet #aspnetcore

@benblo42 Thanks! It is definitely on my mind, but at the moment — not yet (and likely not this year, but we'll see). Though I would be open to people contributing it.

@ashmind kia ora! So, SharpLab is kinda awesome 😍. Any way to choose framework version, eg .NET 7.0? With System.Runtime.Experimental on top?

@OneWholesomeDev Yes, eventually, however it's not in my immediate backlog (need to do some maintenance and fixes first).