The techniques in the article right now have had around 95%-100% success at avoiding spam and take about 5 min. to implement. Your approach of putting an LLM in front of your inbox gives 97% accuracy, may have false positives (so you may not receive that account deletion email after all), requires to run inference and, I assume, would take at least an hour to setup.
Also, the two can be complementary, anyways, so I am not sure what your point is.
I noticed that, too. Technically I think this is a version of JS conversion. Interesting that he doesn't specifically mention XOR in the article. He does suggest combining methods though. I suspect this is effective.
It layers a pentatonic guitar melody with filter sweep, a saw/triangle bass, warm e-piano chords, TR-808 drums, and a sparse music box that drifts across the stereo field.
I'm blown away.
I do acknowledge the possiblity that it might be heavily plagiarized from an original composition in the training set - I wouldn't know.
> gave that task to Claude and within 15 minutes I had a working userscript
Hate to say it but you can just tell an LLM to make the calendar for you as an html artifact that includes a print view. It can also add a .ics export.
Of course you should go over the dates and holidays to see if it got them right.
Yea I've thought about it and looked into some pre-made calendar components too but then I realized this will still take me longer to get small things like margins, font etc. right and it will just be yet another coding project on my pile of projects ;)
Very true. While I believe there can be a place for a language tailored specifically for gamedev, GDScript certainly is not it.
C# support seems to have been added reluctantly, because the prospect of attracting unity devs was to good to pass on. But then they switch over and realize it is treated like the ugly stepchild and the hope is that you will just use their scripting language to avoid the friction.
And don't get me started on using GDExtension for other languages. Claiming C/C++ is available as a scripting languages in the same sentence as GDScript is a joke.
that's not been my experience at all. I've found that Godot works exceptionally well with C#, and I've felt zero pressure to use GDScript. It integrates really well with Rider too, which is the C# IDE I use. Even when there's places online that use GDScript examples, they tend to translate pretty much directly 1:1 to C#.
Visual studio 2026 offers a one time perpetual license for $500. It's not advertised at all and the hyperlink to it seems to shift constantly. I think if more people knew about this option, VS would have a significantly better reputation around here. $500 isn't cheap but it's a one time deal. If I hadn't discovered this I'd probably be looking for other options too.
Can you please elaborate why you think C# is not really well integrated into Godot? I'm using it myself and it seems fine to me. There's quite a community around Godot + C# check out https://chickensoft.games as well. FWIW the Slay the Spire 2 devs also seem to be happy with Godot + C#
For the record I would still choose C# out of all the options for a bigger game (GDscript with strict type-hinting might suffice for small ones). My main complaints that I can still remember:
- deprecated web export for C# with Godot 4. This is basically my main gripe and why C# has been a non-starter for me.
- Engine still has a separate binary for C# support. I think they are working on unifying them. But once you use the .NET one you can't export to web even when you don't use C#.
- setting up debugging+lsp was a pain and consistency was very flaky afterwards
Also a lot of the editor workflow is built around using GDScript with the built-in text-editor. I don't actually mind it that much, but you asked why it's not as well integrated and that's an obvious one for newbies.
If your experience was wildly different, please do tell.
Thanks for sharing your experience. Mostly very fair points.
Yes, exporting Godot + C# games to web is not possible right now and it seemingly takes the maintainers quite a lot of time to figure it out. It's apparently an official priority https://godotengine.org/priorities/#enable-users-to-export-t... but I wouldn't hold my breath for now.
Similarly for the new GDExtension based C# bindings https://github.com/raulsntos/godot-dotnet which have been cooking for quite a while now. Those would make it possible to only have a single Godot build and get rid of the .Net build. Tbf, the maintainers were teasing that godot-dotnet will soon™ be ready for early testing so who knows.
About setting up debugging+lsp, what were your struggles? Personally, I'm quite happy with JetBrains Rider Godot integration https://github.com/JetBrains/godot-support and JetBrains seems to be quite committed to supporting Godot https://godotengine.org/article/jetbrains-joins-dev-fund-sup... you even get support for Edit and Continue (EnC), which is what's easy to confuse with hot reload but EnC is actually the thing that makes it possible to change C# code while the game is running and the changes are reflected without losing the game state. Really handy. AFAIK you can't get this with open source C# tools like netcoredbg so I'd understand if you'd think that the Godot + C# DevEx is quite lacking if you use something like Neovim or Zed. I think VSCode has actually pretty good support but the really good tech behind the C# extension is proprietary and cannot be used by forks like Cursor or Windsurf.
But I totally agree, the GDScript integration is what Godot cares the most about and it shows in many places in the editor. So my experience is not wildly different in that sense. Just some different aspects I focus on maybe. But I'm always curious to see how people approach this stuff and what they feel about it.
I've been following the github issue for C# web export and let me tell you: It's not been a priority.
> EnC makes it possible to change C# code while the game is running and the changes are reflected without losing the game state.
sounds exactly like code hot reload to me? It would be very nice to have.
> About setting up debugging+lsp, what were your struggles?
> Godot + C# DevEx is quite lacking if you use something like Neovim or Zed
you got me. I tried integrating a Godot workflow using neovim and at some point with emacs. Both were subpar. This actually reminds of another pain point: gdshader. Also very integrated with the editor just like GDscript, which makes it another thing you have to work on getting integrated with your IDE of choice - or you bite the bullet and just use the editor for this. But shaders are such a massive shitshow that I can't really blame Godot for this.
FWIW, Rider has very good support for keybindings - I can't vouch for vim bindings directly but their emacs keybinding scheme is brilliant, IMO it's the next closest thing to working from Emacs itself. So I'm guessing the vim bindings support is similarly fantastic.
I spent ages trying to get Emacs to work well with C# stuff, because I can't stand using IDEs that don't have some sort of emacs-style keybinding support, but eventually I bit the bullet and started using Rider and honestly it's been amazing and worked seamlessly with everything I've thrown at it (especially Godot). Highly recommended.
Guests don't care about charisma, they care who your previous guests were. He early on got Elon Musk as a guest (AFAIK by writing a paper that was overly favorable to Tesla) and managed to snowball that into a big podcast.
Also guests agreeing to go on your show means they already want to talk about something, so in a way it's more important to shut up than ask good questions.
> I always assumed there was more to Org mode than "Markdown for Emacs", but this post makes it sound exactly like that.
That is because the post focuses on org-mode: the markup language. But there is also org-mode: the agenda tracker, org-mode: the literate programming system, org-mode: the spreadsheet, org-mode: the website publishing system, org-mode: the document writing system, org-mode the time-tracking system, org-mode: the personal wiki, etc.
Probably always be true, but also probably not effective in the wild. Researchers will train a version, see results are off, put guards against poisoned data, re-train and no damage been done to whatever they release.
Also, the two can be complementary, anyways, so I am not sure what your point is.
reply