We also used ts-migrate from Airbnb. It sounds like Patreon had one team, the platform team, handle the remaining 2000+ files. Whereas we had each product team tackle their domain as they replatformed. We also knew the goal and formed the strategy at the start of the project.
09.03.2026 00:15
π 1
π 0
π¬ 0
π 0
I just have to say at Target we migrated 1MM lines of JS to TS in 2 years. As in the dotcom site you all use. It wasn't a migration so much as a replatform, but certainly large swaths were ported directly. No AI at that time. Our situation was likely not the same as Patreon with 70+ engineers, tho.
09.03.2026 00:10
π 3
π 0
π¬ 1
π 0
This is like MS adding an "experimental" Copilot refactoring to add missing imports. It never worked, because it's an LLM. A perfectly good refactoring is ALREADY in the IDE and that always works, because it's not a hard problem to solve. Worst part? Couldn't use that if Copilot feature was active.
08.03.2026 01:59
π 0
π 0
π¬ 0
π 0
I'm sure the Ministry of Accuracy will make sure to keep things in check.
06.03.2026 17:26
π 0
π 0
π¬ 0
π 0
.NET and .NET Core official support policy | .NET
Learn about .NET and .NET Core support policies, which refer to several technologies including the runtime, ASP.NET Core and Entity Framework Core.
Yeah I like preview better, it's a little more accurate as well.
This is what .NET versioning does for example but it's usually to include servicing support, so there are Beta and RC releases too that have standard support:
dotnet.microsoft.com/en-us/platfo...
05.03.2026 21:47
π 1
π 0
π¬ 0
π 0
Community highlights β¨
- @webpro.nl shares Unbash: zero dep bash parser written in TS
- @chrisnicholas.dev explains decoders, a Zod alternative
- @voidzero.dev highlights What's New in ViteLand: February 2026
- @yoav.codes showcases Electrobun with WebGPU
- @kentcdodds.com announces new TS workshop
04.03.2026 15:42
π 5
π 0
π¬ 2
π 0
CSS shape() landed in Firefox 148. It's similar to path(), but allows the full expressiveness of CSSβ¦
02.03.2026 15:07
π 150
π 24
π¬ 8
π 3
And Cody has a spreadsheet/PDF with all 50 states' analysis for health insurance:
www.measuretwicemoney.com/choosefi
03.03.2026 18:12
π 0
π 0
π¬ 0
π 0
Navigating Health Insurance | With Cody Garrett | Ep 588
This is a pretty incredible episode and resource/spreadsheet.
open.spotify.com/episode/54TF...
03.03.2026 18:12
π 1
π 0
π¬ 1
π 0
somehow i cant stop watching
03.03.2026 05:33
π 1
π 0
π¬ 0
π 0
That's true. For myself, that's too risky a bet. I'd rather have my ducks in a row to continue to give my users a good service than risk being fined because I was hoping the paper pushers would never get to me. All it takes is one fine to put you under. Flying under the radar works until it doesn't.
02.03.2026 19:08
π 0
π 0
π¬ 0
π 0
And this is on top of the Big Ball of Privacy Law Mud.
This is one reason why some people are happy to be on platforms. The risk is you're tying your fate to them, but also they take care of all this stuff.
Conveniently left out of the "Build a seven-figure solo business" hustle webinars.
02.03.2026 19:01
π 0
π 0
π¬ 0
π 0
Yeah it sucks. I want to be in compliance but they make it really hard for small indie developers. Don't even talk to me about the new age assurance laws. π
02.03.2026 18:55
π 0
π 0
π¬ 2
π 0
(and Stripe Tax is basically a joke. It's extra PLUS you still have to set up tax merchant accounts in multiple countries.)
An MoR takes care of all that for you.
02.03.2026 18:51
π 1
π 0
π¬ 1
π 0
Stripe will happily take your payments without telling you that you've likely exceeded sales tax thresholds simply by having a customer in Germany pay you. Caveat emptor, indeed.
02.03.2026 18:50
π 1
π 0
π¬ 1
π 0
Not directed at me but re: last point, you want to learn more about Merchants of Record (MoR) providers like Paddle. They take care of sales taxes for SaaS and digital products. Highly recommend instead of plain Stripe (which is not an MoR).
02.03.2026 18:48
π 2
π 0
π¬ 1
π 0
Practical TypeScript: Foundations to Fluency
Learn TypeScript from the ground up. For beginners and devs filling gaps. Build real mental models, not just syntax. 5 hands-on workshops. Start understanding.
π’ Practical TypeScript: Foundations to Fluency is now live!
First two workshops are free, the whole bundle is $200. There's a 40% discount for one week only! Check out the home page to answer your burning questions like: "What? AI writes my code now..."
www.epicweb.dev/practical-ty...
02.03.2026 17:33
π 12
π 1
π¬ 2
π 0
Whatβs New in ViteLand: February 2026 Recap
Our February 2026 recap features Oxfmt Beta with 100% Prettier compatibility and 36x faster performance, Vite 8 devtools, Vitest async leak detection, and Rolldown performance gains.
β¨ What's New in ViteLand: February 2026 Recap
β Oxfmt Beta: 100% Prettier conformance
β @vite.dev 8 includes devtools
β @vitest.dev 4.1 beta adds aroundEach/aroundAll hooks
β @rolldown.rs becomes 9.6% faster
β Oxlint now supports 59/61 ts-eslint rules
β NestJS migrates to Vitest
β and much more π
02.03.2026 13:12
π 55
π 8
π¬ 0
π 1
Yep, that was the culprit. Jeez! @developers.google.com what gives!
Stripping `iss` before it hits origin fixes it. I can also force it to match the correct ISS value for validation, so it will continue to throw an error if it is a true mismatch (the intent of the parameter).
27.02.2026 15:23
π 1
π 0
π¬ 1
π 0
I haven't confirmed removing the param fixes it, it's just what I noticed that breaks both my local and prod versions. If removing it fixes the problem, that's pretty concerning. Google adopted an OPTIONAL RFC and didn't version the change on their authorization endpoint. WTF.
27.02.2026 15:09
π 1
π 0
π¬ 1
π 0
The symptom shows up as a redirect_uri_mismatch so that's what I was searching around to see if anyone else has been bitten by this. Worked for 13+ years and now just breaks!
27.02.2026 14:51
π 1
π 0
π¬ 1
π 0
I discovered it last night. I'm thinking I can mitigate it with Cloudflare in production so that's what I'm trying this morning.
The problem is the library strips known OAuth message params from the URI but it doesn't recognize `iss` as a message part. Maybe it's newer in the spec.
27.02.2026 14:50
π 1
π 0
π¬ 1
π 0
I can't find anyone else running into this, but Google OAuth sign in flow broke this or last month. It was very recent anyway, my logs only go back 7 days. The authentication code flow now adds the issuer (iss) query param on the redirect and it breaks my OAuth library which doesn't expect it. π
27.02.2026 14:50
π 1
π 0
π¬ 1
π 0